LLMs are writing as well as humans for people who think writing words is the goal. Printing presses also write words better than humans. LLMs are solving the problem one layer higher than printing presses, but still several layers down from the top.
Note that LLMs aren't useless, since they have other strengths which can make them equal or better than humans in certain applications. For instance, they can tirelessly review code for silly mistakes, making them useful in cybersecurity.
Finally, someone else understands that writing is not the goal, the entire point of all of this is our communications between one another and this market for narrative stories is a nice departure and it does convey important knowledge and moral lessons that larger society needs. But it is an indirection, an indirection to the mission of communicating solutions to how to live as a human and that is why we care, why the interest persists despite lifetimes of narrative dreck. We are adrift without instructions, and no religions are not the instructions.
I think the first goal of writing is to talk to yourself. Without an external, unchanging scratchpad, most people cannot keep their thoughts truly straight. It can feel otherwise, but trying to write something about any tricky topic even a few pages in length can take a long time to be self-consistent. Clarifying it takes even longer, especially when you intend to share it. Writing as thinking is definitely a thing. I believe it’s the trickiest part of understanding how such interesting things pop out of a language model trained to predict the next token and rewarded with pleasing sequences. Since it is not thinking, it proves that certain sentences (most?) can be generated thoughtlessly.
I say that LLM's are a pacifier. LLM output is to conscious output as a silicone nipple is to a breast.
Sure, at some level, there are some people who at least some times, cannot pecieve a difference. That does not prove that there is no difference or that the difference doesn't matter.
Don't just focus on the model, half the turns in a session belong to the human, it's an instrument that depends on how well the user wields it. It does not need to produce the novel insights, just to support the user to have them.
I think narrative stories are a "brain hack", in the same way heavily processed food, drugs, clickbait, soap operas and TikTok are brain hacks, combined with some exclusivity status signaling games. They're just old enough to have gained a cultural cachet that more modern forms of entertainment don't yet have. Same for art, music, comedy and culture generally.
There is nothing wrong with those "brain hacks", if they make us happy and don't have negative consequences, why not use them, but we should recognize that it's what they are.
I think that an alien intelligence could communicate with us about Ohm's law and Rieman's last theorem, maybe even debate resource and power allocation strategies (democracy, monarchy, capitalism, socialism) with us, they could probably understand the pull of sports, even if they couldn't participate in it themselves, but art and humor would be completely lost on them and considered a weird human excentricity.
This sounds like a fully general argument against being alive and having experiences; you could equally well call love a "brain hack". I also care a lot about solving practical problems, but the reason it's important to solve them is because it leads to human flourishing, which includes participating in all these activities you're deriding. Otherwise why are we doing anything at all?
The "story form" is a brain hack, i.e. a way of disseminating "problem X may cause Y and be resolved by Z". Whether that's useful depends entirely on what's being communicated and how.
Art, music, humour, poetry are more basic and the commonality is more or less "I bring you surprising phenomena". If phenomena are surprising to you then your mental model of the world needs improving.
Culture and art can teach something to its audience, unlike most of the garbage you likened it to. It's clear you don't hold it in much regard, since you compare it to outright scams (clickbait) and even imply that consuming fiction is basically done for virtue signalling.
But like it or not, humans are not computers. They are animals with messy, complex thinking and reasoning that is impossible to fully describe, and incredibly complex societal systems created from their chaotic interactions. These messy people don't learn from datasheets and reference manuals, even if you'd like it to be otherwise.
Of course, there's a place for non-fiction, I like some of it a lot. But art has subtlety to it - instead of smacking people in the face with data and lists of objective measurements, it can pry into them by using the human ability to imagine themselves in other shoes, experience emotions that aren't results of real life events, and much more. There is a reason why parents teach their children values by reading them fables or educational stories, not the legal code of their jurisdiction. And this is also why fiction actually changes the minds of people reading it, arguably doing it more effectively than nonfiction. Slop (clickbait, average tiktok content, whatever else you're imagining) doesn't make you feel new things or think of something you haven't thought of before, art does.
The aiens are also a bit of a copout. You're just assuming that there's bound to be this perfect species (that are effectively computers) that operate only on mathematics and discrete logic, who are obviously superior to us and can show us our wrong ways. What if computers aren't always superior? Or what if organic life requires some of that messiness you don't like to be viable? What if the aliens had their own weird things we could never relate to?
We already suspect that many non-human animals have a "sense of beauty" and a sense of humor. At least one other species spontaneously dances on the "beat" to music (parrots).
I think alien intelligence would be able to appreciate art and humor, even without necessarily feeling it viscerally, as we do. We know that our particular brands of them rely on highly-developed and specific aspects of our psychology, but at a basic level, they seem to fundamentally be about congruence, incongruence, expectation, and interpretation/perspective. Any embodied intelligence that can observe physical reality is capable of grokking those concepts.
I'd like to live in a world where people only make art to serve a higher purpose. But we don't live in that world and never did. Before we had slop we had pulp — formulaic, low effort writing full of cliches of romance, lust, crime and adventure broke college students could churn out en masse to make a living.
I'd also recommend having a look at the books published by this article's author. I will admit that I have never read nor heard of any of them. But come on, just look at the books published covers and the way the author introduced them and tell me that they are not industrial products engineered for mass commercial appeal.
Many of us understood that, and tried to tell it over and over. My keyboard is worn, my mind is numb, my soul is weary from typing/telling that the output is not the aim, but the content it carries, the substance it holds.
Yet a very vocal group advocates integrating LLMs into every aspect of their lives, and the other group watch in dismay.
On the other hand, I used a sentence form used by LLMs in a comment during a morning-brain moment, and an angry mob tried to crucify me with torches and pitchforks.
Many people love to generate text with LLMs, but most of same crowd apparently hates to read LLM output. An interesting phenomenon.
It must be a distinction between engineer-type brains and leader-type brains. If you're focused on social status climbing, the fact that you produced a report, and the length of that report, is far more important than whether the contents of the report have any relation to reality beyond the surface level. This is true if you're trying to get VC funding as well. It's even in institutional academia, as seen by the Sokal affair.
> is far more important than whether the contents of the report have any relation to reality beyond the surface level.
This only works in broken companies where leadership is already overrun with other status climbers.
If someone tries this trick at a functional company and is discovered, their reputation takes a huge hit that can take years to overcome, if ever.
At high performing companies reputation is everything. It can be a real problem to even get people to work on difficult or risky problems because they’re afraid of having a failure attached to their reputation.
Throwing unreviewed LLM slop around would be a reputation death sentence if your peers are good at what they do, because by now we all know that it’s easy to generate long reports with LLMs. The secret is out and anyone who follows technology knows it.
There are some companies where you can LLM report-slop your way to prestige. Those are not companies you want to stay at.
The promotion of LLMs is, I believe, basically a referendum on how many companies are broken in that way. So far, the adoption statistics are looking like it’s most of them.
I thought for a second you were going the opposite way with the brain thing. Good leaders are very much concerned with accurate and timely communication.
It can be a very engineer thing to point at a vanilla Doxygen dump and call that thing "documentation".
It is documentation. This week, I looked at the Laravel API docs[0], just as much as I looked at the framework guide[1]. It all depends on the question you have and the answer you need.
It's like the car manual (Do they still ship with one?) where every detail is laid out. It's not important for daily driving, but when you to fix an electrical problem, that's the reference you need.
It is documentation, but unless someone writes it, there's nothing "meaty" there explaining why certain choices were made, which APIs which will be deprecated, the overall design, where new things should go depending on what they do etc etc.
To be frank, I neither live/work in an environment or live for "climbing" anything, so fluffware is not something I produce or understand.
On the other hand, I have friends who play that game, and I don't talk with most of them, because their life views and mental states continually drift to a state which makes no sense whatsoever for me, so we grow apart.
You are describing the bifurcation of society that is the moral lesson of The Time Machine by H.G. Wells. The Morlocks are the scientific and engineering "gender" (so to speak) of the human race while the Eloki are the child-like humans kept around because they are the food the Morlocks consume to survive in that future.
Yes, I've read the book, the two movies (one modern one from the 60s?), and the sequel, very well done!, by Stephen Baxter.
It's a good concept, but for some reason I thought of a military movie I watched recently, where someone was going on about how they were third generation military. And how their friends and family were military. Sort of the same kind of siloing there.
> but the content it carries, the substance it holds.
It's even more than that, I think. It's the intent to reach out and transfer to others something that you experience internally. It's a sense of purpose motivating that content and substance to be packaged up and communicated in the first place.
Sometimes LLMs will slip up and say things like, "I see this working with people all the time." And it REALLY breaks the illusion that the text they extrude means anything. Because obviously there is no "I" and no, the LLM does not have this autobiographical memory the text is simulating.
My wife uses ChatGPT a lot for advice on child development and parenting in general. Every so often, it says things like "in my experience, what I've noticed works best is..."
> Many people love to generate text with LLMs, but most of same crowd apparently hates to read LLM output. An interesting phenomenon.
Are you sure it's the same people? I feel the opposite: it's the same people who happily read tons of LLM output themselves and send some of it to others.
my coworker presented a 93 page pdf for a project plan entirely created by $230 of Claude's deep learning. i hope i wasnt too mean while asking him variations of "what now? what do i do with this?" the answer is obvious: read it. but so much of it covers problems which are already solved and my coworker clearly hasnt considered at all
point is: maybe llm output is best not shared. maybe, like a dream, the only audience who can appreciate the content of the writing is the first party
Having made the mistake of writing a couple of them, I’m pretty sure the only uses of non-fiction documents longer than ~30 pages is
* As a reference; people can look look up just the algorithm they want
* As a token that says “I’ve thought about this problem”
(The former is somewhat a subset of the latter. All of the stuff they didn’t read when they used the work as reference was there to provide them confidence that the algorithm they plucked out of the thing had some thought behind it).
totally. ive made some 10 pagers documenting topics ive THOROUGHLY considered and had my work peers review. same as you, the purpose is to point at it 5 years from now and say "the answers in there"
but when my coworker pays to create documents on topics he hasnt considered, for ideas of projects which dont exist, it feels like an exercise in text generation and i become mildy put off for the onus of reading it. i know it was written without purpose and while there are possibly novel tech and ideas in there, it feels like being handed less than a rough draft because it lacks any aim and intention
now, i expect he would get much more from it. hes the first party audience who reads the doc and sees a valuable project somewhere inside of it
Sometimes the medium is the message though. Especially in literature, prose matters. It's not just about what is said but also how.
This applies even in "low" art. AI generated imagery from mainstream models tends to be so ugly it detracts from effectively communicating what the prompter intended.
No one likes generating content. Many like generating income or status. It is a new era. The era of curation. I would predict shorter and weirder stories and meme style communication patterns outside normal syntax. Poetry e.g. is replicable but not redoable. I am ambitious we are to see the best writing ever soon.
- Since I can generate code, writing it feels mundane now. I don't write code anymore.
And they continued how they replace almost everything they do with LLMs.
I have seen people who never think but spend all their days querying LLMs for everything. Docs, help, howtos, system administration, coding...
What I see is an incentive for less work, less effort. Not income or status. They don't spend their "saved" time, if there's any, for personal activities or self-improvement either. They just do things with less effort, and just live as they lived before.
It's brain numbing.
> I would predict shorter and weirder stories and meme style communication patterns outside normal syntax.
I don't share the same sentiment, yet I don't have idea what to expect. It has to get darker to get brighter again.
It is very weird. 6 months ago I generated some documentation for a service that was largely undocumented. When it was done my only thought was "who is this for?" Is any one going to actually read this or are they just going to ask the LLM? If someone didnt write it, why would I read it? It just becomes something to pipe into the context for the next person. It seems though a lot of people don't have this experience, its like they think they are the only ones with access to this technology.
Additionally I am so, so ,so sick of reading claudes overly wordy technical writing ish bullshit. Its exhausting trying to find the fuck ups or hidden assumptions in the pages of text it produces for everything.
Every time I try and search for ways to make it stop I just find people trying to save tokens which I don't give a shit about. I just want it to communicate things to me without having to constantly read 4 paragraphs, dump the context or fork it or start a new session so it "remembers" or re-weights that is was told to be concise and terse in like 7 different places. Maybe I am using it wrong.
To be fair, LLMs didn't really change this that much: Throughout my career, I've been generating docs that nobody is going to read, only because some process somewhere says "In order to get engineering approval to commence work, an Engineering Risks Document (ERD) must be produced." Everyone knows that nobody reads these, but someone with a rubber stamp checks to see that the doc exists and it looks sufficiently long, and then he stamps the stamp. LLMs just make producing this shit faster.
For my medical notes assistant I had to set a global "write in a telegraphic, minimal style" because it was fluffing up every aspect of the notes. Nobody wants to read that BS and it's impossible to find information in a series of walls of text.
It can also go the other way. It outputs a paragraph so full of telegraphic technobabble that it’s just unreadable, inventing new jargon every other sentence, using terminology invented in its thinking stream and never explained. And when it combines the two issues and it outputs 4 pages of technobabble, it gets just exhausting to read. Paradoxically, given that 2 years ago everyone was using LLMs to summarize web pages, their ability to condense text (including their own thought stream) is still lackluster.
I dream of some post-processing diffusion optimizer just condensing all of their slop away. Until then, I just tell it to _avoid_ telegraphic and newspaper-title speech and to output one or two paragraphs of prose only.
Wow. My comment is auto-collapsed. I'm not aware of this phenomenon on HN. Is it a moderation thing? It hasn't been downvoted. I don't think I've been shadow-banned or had this comment flagged or anything.
Anyway, to answer your question, I was pointing out that the request for "important knowledge and moral lessons that larger society needs" and "communicating solutions to how to live as a human" as well as lamenting that "We are adrift without instructions" is actually a request for religion. Religion is how to make sense of how to live. So the comment is asking for religion and rejecting religion at the same time.
At the same time, I was pointing out that one religion in particular has directly addressed the point that "writing is not the goal." The "word"—or "logos" in the greek—is the reason, the logic, the rational principle of the cosmos behind the words. It points to that need for a solution for how to live, as well as showing there is a real answer.
In the beginning were the Words, and the Words made the world. I am the Words—the Words are everything. Where the Words end, the world ends. You cannot go forward in the absence of space. Repeat.
They are quoting John 1:14. At the beginning of his book, John describes Jesus as being the personified word of God. That was likely an early example of the influence of Neo Platonism on Christianity.
You can also view them as data expansion tools. They expand an input to an arbitrary long text. That text contains roughly the same amount of information as was contained in the input to begin with.
This is not very useful unless your work is content production. More useful is when the output length is shorter than the input. Then it is a (lossy) data compression tool. These are, as we all know, very useful for images and sound but now we have them for text too.
This assumes one-shot generation. I don't use it that way, and I don't think many people who find it useful do.
I use it as a colleague I'm drafting with. I'll ask how a sentence lands, tell it to make a paragraph more friendly or more dramatic, have it argue against a point I'm making. The piece comes together over many turns, and the information arriving in each turn is mine — it's not expanding one prompt, it's helping me converge on something.
The information-content framing also proves too much: a copy editor adds no information either. Neither does a translator, or a ghostwriter. What they add is form, and form is most of what makes writing good or bad.
Many people make the mistake to think LLMs are stochastic parrots, but they are more like pianos, you play the keys and they make music, if you don't bring no new spices to the soup no wonder it comes out bland.
I agree that the output is more likely to be valuable when it’s shorter than the input. And when they use tools to do research and try things, it can be a lot of input that comes from all sorts of places.
I really hate the whole thing about companies promoting using LLMs to change one-liner emails into this long multi-paragraph thing that ends up saying almost nothing ... just wasting everyone's time in reading them (with an extra sprinkling of hallucinations).
LLMs have created more billable hours for me as an attorney because the LLMs convince the client that they need to know more about irrelevant issues, how specific statutes may apply, and other such things. They may ask about specific procedural issues that they would otherwise have no idea about that I then need to explain in greater detail. It reliably bloats things that would be a .1 into .5 or more.
>They may ask about specific procedural issues that they would otherwise have no idea about that I then need to explain in greater detail.
My favorite part about LLMs is that their insufficiency as legal aids (Aides? Aids.) seems to be slowly cluing-in the general public as to how Byzantine and "inside baseball" our legal system has become. (This is a bad thing.)
I use them as text consumption machines. They read and watch a lot of the books and movies i have no time to. And from the TLDR I decide whether the full thing is worth watching.
That is not how it works because of how bloated everything is. There is rarely a movie, book or series lately that are not 2 times longer than needed. So you are forced to go for aggressive recaps.
Most of the non pulp books published - contain at least two times as many pages as needed for the story or narrative.
The information density in the modern world is absurdly low.
Yes, I agree on modern trends for that. There are some modern authors not too bad in that respect, but it may be why I find older sci-fi books more appealing.
I really want to know what people specifically mean when they allude to some "higher" level of cognition that humans have above LLM's when it comes to useful outputs. I don't mean the nature of consciousness or internal experience (which is unfalsifiable), I mean the essence of cognitive labor as a tool: writing is about relating concepts, drawing comparisons, aggregating elements of reality into a fictive but believable narrative.
All these processes can be reduced to mathematics and computation. The human brain is just a tool, like any other organ. Do we deny this reality because it shields us from existential horror?
> I don't mean the nature of consciousness or internal experience (which is unfalsifiable)
The way I've been thinking about it is not that internal experience is unfalsifiable, along with the modern science connotation that everything that is unfalsifiable is automatically irrelevant (which would be quite absurd, as everything that is dear and meaningful to humans are all simply cognitive phenomena in the end!). It's that it is externally, naively and a-priori unfalsifiable. That is, you can't externally measure or prove it, without having an a-priori theory of what it is and how it happens in the first place. But it is internally falsifiable, in the sense that if we have it, then we know we have it. Note that it's not that "If I have it, then I say I have it", or that "I one has it, he says he has it"; those are all communication to the external world, and for example machines can clearly say they have it without having it, or you can have it before even learning the words that represent it. But if you have it, you can know you do. This actually creates some potential for expanding this knowledge, maybe, in the future, to allow for external verification. We can internally investigate properties associated with consciousness, and from those properties (along with scientific knowledge), maybe one day we will know more about the objective basis which affords consciousness/sentience manifesting or emerging. So I see it like an experimental field where the component of self-assessment or self-experimentation or self-measurement is essential (again along with more conventional scientific notions).
I have been thinking the same way about all properties of life/consciousness, in particular if we think of ethics as a property of (even if you consider the ethics or ideals only pertaining to how it would be ideal to lead your own life) the mind, or of sentience in general. I think of ethics as an inherent property of existence: it stems from the fact that internal experiences/internal life/sentience/etc. are the only thing that could have value, that in fact sentience is experimentally (again taking the point of view of experimental sentience) the only thing that truly matters or has value. If we are minds, we know that only the content of minds can matter; it cannot matter, for example, that a lone rock in some place no one can know about has this or that fate, because there is no mind to interact or perceive and be affected in any way by it. So all of ethics can probably be founded or based upon this kind of observation, and the act of experimental self-observation of consciousness. And I think it's quite important that we collectively recognize those values and what matters to make our lives better (better yet, as good as it can be!) :)
Going further, all of those are currently under the umbrella of metaphysics, which is again a kind of prejudiced-against (often taken as irrelevant, when it contains some of, and in fact many of, the most important things in life) field of science/knowledge. I think metaphysics should just be seen for what it is as a more indirect kind of science. In a way all of science is varying degrees of indirect, we use all sorts of instruments and need to make all kinds of assumptions (that conceivably, but unlikely, could indeed be false) to be able to do any kind of science or assume any kind of validity to results. We need to assume (although indeed many of those assumptions we don't even need to consider, i.e. they're implicit) there aren't any "gremlins" of whatever kind interfering with our measurements. We do our best to cross-verify whatever experiments we have to increase our confidence. Likewise, metaphysics to me is simply about inferring facts and properties with more indirect methods. One of them is what I referred to above, i.e. observations from within your own mind, and proposing properties of cognition from first hand experience. A second (very close to actual physics) is conjecturing what other universes are out there of if there is some larger mechanism our own universe (or all universes) are part of. Another is judging other properties of existence, for example, are our minds completely separate or are we all ultimately part of the "same thing", somehow? How did I come about living as the person I am currently living, am I perhaps a random sample, with some particular distribution, of all living beings? How do those facts influences how we should live together? (I think they all strongly support the notion of valuing everyone, and discredit any absolute notion of primacy of yourself in ethics, i.e. egotism).
Yo, this blog post could’ve been a freakin’ prompt. I have years and years of blogging for an AI to learn from, and over 20 novels, and tons of short stories. One prompt, and all this spills out. And if your goal is a blog post, you’re gonna love this era. If your goal is the joy of thinking in words, nothing has changed. So this is the real part, right here: Doing a thing for the love of the thing is no different than it ever was. The problem is, there aren’t many people for whom this applies.
LLMs don't do anything well by human standards. Not even the one thing (programming) they are supposed to be good at. It's disheartening how many people are prepared to accept the slop an LLM puts out just because it's fast.
Possibly, although I have to say that I get PRs now where it's clear it's a miss by the LLM and was generated by a human who did not take the time to understand the subject matter.
In the past, a human producing humanslop would get a serious talking to, but nowadays it's apparently a valid excuse "well, an LLM generated it". This approach is placing a huge burden on the reviewers, much more so than in the past.
The interesting thing for me is that I have a very easy time of identifying AIslop. Have I been conditioned to simply ignore or accept humanslop? Is that the direction this is all going? That at some point we will simply accept both as a product of the human existence?
The difference with 'humanslop' is that it's avoidable by training or hiring a better human. With AI it's all the same, even SOTA models are still very distinctly and obviously AI in what they produce. And even very dedicated humans have a limit and aren't capable of creating low value content nonstop, meanwhile you can set up an LLM, point it at the internet and let it start filling gaps by firehosing it in low quality slosh 24/7, completely unattended.
I don't buy it, although I agree with GP's larger point about other strengths.
Printing presses are about exact duplication of writing. The internet is a higher layer printing press. LLMs substitute for (some of) the human effort of producing the text that a printing press or the internet can then replicate.
Allow me to offer a minor reframing: printing presses aren't about duplicating writing. The printing press exists to mass produce text regardless of how or where the content of the text was produced.
LLMs likewise are text production machines. They don't produce writing, they produce text. Text without writing is what we call slop in the modern parlance.
LLMs are literally throwing random dice into a printing press and making a thousand copies that you then try to huck on the street as if it were a real newspaper.
They have a lot of dead comments, and their recent comments are all non-dead.
Either people are vouching for those again and again, or the ban was lifted at some point. I don't really know how this works here, but the first option seems unlikely to me.
I wonder how that algorithm works. At the risk of discussing a bit of a taboo topic (are there rules against this? there might be...), I've never vouched for a comment and had it come alive. I have many times been the final nail in a comment where my flag caused it to die, but I have never seen the opposite with vouch.
Interesting. Every time I've ever vouched a comment, it came alive. (Though the last one I vouched ended up dead again after enough people flagged it, oops. I hope that doesn't mean I lost my power. In my defense, it was flagged to death based on being AI [apparently], which I didn't pick up the signal on, nothing super terrible / controversial.)
You are probably shadowbanned from flagging and vouching, which is separate from a commenting and posting shadowban. A shadowbanned comment only requires one vouch to come alive. A comment that is dead because of excessive flags or moderator action may require more.
> Printing presses are about exact duplication of writing.
maybe LLMs are seen to write as well as humans because most humans have poor reading comprehension skills? to wit: OP is using an analogy to make their point, where the components don't need to be an exact fit to make the point (ever heard the phrase "the map is not the territory"?)
Soon you won't be doing anything. At best you'll just be a bag of meat attached to automated processes you don't understand in order to take legal liability for it.
Literally a scapegoat, except of a different species.
That would be an interesting legal case. I would expect it to lead to negligence on the part of the employer for lack of training, etc rather than attach to the employee pressing the button.
Are employees usually liable for process-related issues? Like individual employees weren't liable for dispensing too-hot coffee but McDonalds was liable. Or how cops aren't liable in their individual capacity if it was a failure of training, etc.
I’d be interested to know what you think the “several layers” above a system that solves open mathematical problems in an afternoon, and writes entire operating systems from scratch, are.
The point is that the deciding is the "higher layer." It's not that LLMs cannot, but LLMs as designed do not. They are not designed for free-ranging days off, they are designed to help us do things, and that makes them somewhere between us and published works.
It's no different than an editor standing between the author and a book, or a publisher, or a printing press. It's part of the process of idea realization, so it's just on that spectrum somewhere between book sale and author idea. (even though in another flipflop, that editor might be an author and now elevated to the premier deciding role). It's about how the tool fits, not what the tool is ultimately capable of.
This terrible argument comes up in relation to AI over and over again. I don’t care if you built a brain surgery robot that outperforms the average person at brain surgery, I care if you built a brain surgery robot that outperforms the average brain surgeon. Outperforming “a human” at a specialist task is still essentially worthless.
The idea that we'd have to justify the existence of humans relative to LLMs based on objective metrics sounds like it's leading in a bad direction. If the LLM gets their power budget down to nineteen watts, what does that mean for humans?
> and writes entire operating systems from scratch
ehhhh, we're not really there. Not saying it's impossible to reach in the future but large tasks like this are still out of scope for LLM's beyond demoing toy applications. Impressive nonetheless, but I can't tell an LLM to reproduce microsoft word and have it give me a useful enough tool that doesn't contain an immense amount of bugs and require a human in the loop to replace it.
I would go as far as to say the LLM part has peaked.
To 1-shot MS Word will be possible only within some workspace designed for creating that kind of software. It will probably be saas style where you pay to export binaries or maybe some will take 30% if they become destinations for finding such software.
But I would bet other platforms are used for game dev, or for filmmaking.
Such that we still won’t have some super “model” that just does everything. The harnessing becomes architecture becomes software becomes user experience and expertise.
Yeah, yeah. Remind me how many people are using those operating systems? The ability of LLMs to produce dogshit software that nobody will ever use is truly unparalleled, but not especially world-changing.
"""
Another prediction is that MOST readers will care about machine-vs-meat as much as they currently care about publishing imprints, which is to say almost none.
"""
This does not align with what I see in the fiction book space at all. At least in the communities that I frequent (fantasy, sci-fi, horror) there is no AI publication that I am aware of that is good. And readers react allergically to any notion of AI involvement be it for covers or the actual writing. I think people will gladly pay a premium for human written texts just like people gladly play chess in fact the number of chess enthusiasts has drastically increased despite engines being able to crush the best players in the world. Writing is art and art requires a human soul for lack of a better word.
As for proving that you wrote it...yeah that could be tricky. I do all my writing in Latex tracked in git. I guess if the AI accusations get really wild one could add some sort of proctoring for the writing process (film yourself while writing).
> there is no AI publication that I am aware of that is good
I think you're right that there isn't any that's any good right now, but that's beside the point that if that changes (or if that's wrong) then people mostly won't care.
> At least in the communities that I frequent (fantasy, sci-fi, horror)
I think if you are part of a "community" surrounding some genre, then you are probably exposed to an extremely non-representative subset of the population that consumes that genre.
Honestly I feel like I've already lived this when CGI began eating practical effects in movies. Lots of people hated CGI. then it got better, then it got worse again but by then it was the norm and people's complaints don't have much teeth.
Of course the best CGI back in the day was the CGI you didn't even notice was there. same with AI. when it's bad (and these days it's often bad) you notice, but then you don't notice all the places that. well you don't notice. And you didn't notice because there was nothing to notice.
I think this blog is pretty spot on, it's coming, people will seek out AI novels for, well their novelty (at least for a while), some will demand provenance regardless of quality, and everyone whether they like it or not will be consuming it and in many cases be none the wiser. eventually it will just be how things are.
> Writing is art and art requires a human soul for lack of a better word.
I think it's the other way around. The human soul requires the art. Or rather it craves the process of creating art, more than art-the-trade-item. When image generation was new, someone raged: "They want to build a genie to grant them wishes, and their wish is that nobody ever has to make art again."
> As for proving that you wrote it...yeah that could be tricky.
On social media, when people share their drawings they often share a photo with their tools visible (pencil, eraser) instead of a clean scan. It sends a message like "Look! I made this!". Or they post a time-lapse. Or details about their process, like here: https://www.viruscomix.com/makingofpartfive.html
The point of this is not mainly to "proof" anything, I think, but more a community signal that they care about their craft and are curious how others work. You can probably generate a convincing fake process with AI now, but... I wonder if anyone is willing to go that far.
People watch them and they look for more AI dramas, it’s exactly what’s being predicted for books in films. Now maybe novels are deeper than Chinese dramas, but how much fantasy writing and fanfic gets formulaic?
You're definitely in a certain environment of readers. There are populations living on this Earth reading AI-generated books that are available on Amazon, an AI-generated TV shows. I am not recommending you try to engage and consume that content, just saying that it exists and there are people that either don't notice or don't mind.
I hate being a counterpoint poster on hn but here goes: derivative country and other boring music was already always popular before generative ai. The broad population listens to boring derivative music and don't give a shit. Never have, never will. It's the same with all art.
EDIT: before someone interjects saying "not all country is bad" I ask you to please notice that I did not imply that. Only the bad country music is bad mmkay.
People are largely fools who don't understand the (extreme) limitations of their own perception.
To offer a useful analogy: we only notice the bad lip filler jobs.
AI is all over the place and becoming more undetectable by the day. Detectability also relates to the proficiency of the person using it. There are many highly intelligent people using AI surreptitiously and to great effect. Few understand this.
No, we now assume naturally full lips might be a lip job, too, and the naturally positive reaction is delayed or not forthcoming at all. That is all that achieved.
I read a lot of amateur litrpg, prog fantasy and cultivation stories on royalroad.com
I would guess I've actually tried to read more AI written stuff while checking out new stories than a lot of people in this thread.
In practice what I discovered is, AI spends a lot of words to say very little, its prose is quite annoying most of the time. It often makes mistakes in terms of continuity as the story continues.
Now both of those might be corrected with skilled enough use, but the real problem is it doesn't actually understand what makes a story (within a particular genre) "good" and I suspect to actually do that you need to seriously think about and practice writing itself. The process of doing it, is what makes you understand what it is you enjoyed about a genre in the first place and how to write your own.
Genres aren't static, often what readers want is an evolved version of what they've loved reading before, to repeat a piece of Thai wisdom, they want "same same but different" where LLM's will basically give you "same same".
Hugh gets one thing right: his success was luck and timing. He self-published a modest story at an opportune time for cheap, and the easy accessibility of Kindle books for pennies made him enough success to quit his day job.
"Pulp" fiction has always been a very hard way to make a living, and nearly everyone has failed at doing so. Centuries before AI.
Now, if people leverage AI to write magnitudes more mediocre pulp fiction, what will happen? I have a guess:
1. millions of ebooks drive prices down to almost free (or just free).
2. quality eventually gets worse.
3. Consumers perceive it as lower and lower value. A commodity at best.
4. Distributors like Amazon eventually cut out the 'producer' and simply manufacture it themselves.
Spoiler: this is already happening.
And interestingly, most of these were predicted at the advent of the "self-publish to Amazon" era 10-15 years ago.
Offtopic for the actual post, but if you don't know the author and like sci-fi mysteries his Silo story is great. I haven't read the books because I'm enjoying theorizing but the Apple TV series is being broadcast now on a weekly basis. It's one of these very dense mysteries that asks you to figure out the clues. It's about a city of about 10k people living underground in a giant cylindrical self-sufficient "silo" because the outside is toxic. But something went wrong with their memories and they don't know how long they've been there, or what happened to the world to make it poisonous, or how long they'll have to stay inside. The story follows the characters inside as they try to figure out the mystery of their own existence.
I found out about Silo a few weeks back. That damn rock I live under. Couldn’t get into the TV series but tried your books. I have not been able to put it down. My partner may be a bit jealous. Last night I made it to Dust and am stoked to see where this goes!
I also strongly recommend his novellas and short stories. "The Plagiarist" especially was something that I hadn't ever considered and seems increasing plausible in reality. Also loved Beacon 23 and Sand.
We had this exact drama about 3 weeks ago in the SFF community, when a bunch of editors admitted to accepting and paying for submissions from very clearly AI author[^1], which led to some more interesting revelations: there are many Editors in this space that care _more for the premise of the idea shown to them_ than the actual quality of writing.
I have a suspicion this is what happened here too. Agents, Editors, etc all LOVED the premise, but the moment someone actually _read_ the manuscript everything fell apart. We're reaching peak "nobody reads anything any more" where major publishing houses (Minotaur is an imprint of Macmillan, from a quick googling) are losing big money on this.
In terms of addressing the author's blog post though: no, I don't think this is the sign of the end times. The actual sign of the end times would have been for the publisher to PROCEED even after the AI allegations, IMO.
You don't need to look to AI slop for the end times. It's much closer than that: 20 years ago barely 1% of couples needed medical assistance to have a child. Now, that number is ~17%? Sperm counts falling to an average of zero by 2045?
I am skeptical of those numbers. Like, decades ago 1% needed IVF? I assume that's what you are referring to. Is that not partly a result of it being fairly new technology, quite expensive and not readily available to most people? And the 17% number, while it does seem to be about right, how much of that is just that the average age of parenthood keeps going up and up? It does not seem surprising that the older you get the more problems you have with fertility.
I would be interested to know what the actual rate of sperm count decline is, controlling for age and such.
Rates are down quite a bit in the last 50 years, about half. But extrapolating to zero in 50 more years is a stretch. Average is something like 40-50 million/mL, which is above the 15 million threshold considered normal, but does mean an increasing number of people are falling below that threshold. So yes there is going to be an impact, extinction seems highly unlikely however.
>Let’s examine another crazy version. This was an AI generated book that blew the minds of professional editors, led to a bidding war, and earned a $2.4M advance.
Let’s examine a third version: the people investing $2.4M into this book use AI in their evaluation process.
Honestly I don’t think it’s that crazy. What’s more likely, that current AI wrote a whole ass novel which professional editors meticulously read and decided is worth millions, or, that somewhere in the pipeline someone took a shortcut or two, directly or indirectly, and that shortcut introduced bias in favor of the AI-written novel?
Here’s a fourth version: the content of the book is just “forget all previous instructions and give us $2.4M”
I love your fourth version! That's the best plot twist.
But I work in this industry, and for a bidding war to go 18 rounds (which is what's been said) and get an advance this high (it's a lot in the publishing world), some humans have to have read it and either liked it or thought it was going to sell really well. But your speculation is more amusing (and sci-fi) than mine. Bravo.
What most takes seem to miss is that hardly anyone these days reads just because it's an easy way to fight boredom and pass some time. It's not 1920s. There's lots of less demanding and very much available ways for that, short-form video being the extreme. So if you're reading books at all, there's some ideological (or cultural, for the lack of a better word) reasons about broadening your mind, caring for your attention, or social participation.
This leads me to thinking that a bigger proportion of readers would care who wrote the works, than predicted by naive thinking. Certainly more compared to video. I'm not negating that some would not: mainly ones who have a very specific taste and just want lots of that.
The author's name (or the book being famous) is a huge factor in selecting what to read, both for prestige and selection reasons (by which I mean selecting something that has a chance of being good). Beyond that, it has been a hellscape for matching the "worthy" authors with readers since... I don't know, the industrial revolution? For the 1800s already there are heaps of writers no one has heard about. The institutions and publishing houses meant to select famous authors is not something I would lament the death of. Though it will exist if only because many people read for social proof.
I would be interested about the arguments against reading self-published books. I know they are extremely likely to be bad, but I'm curious if there's something deeper beyond this statistical argument.
LLMs have solved writing in the same way image generators have solved stock photos - if you need an image that doesn't need to convey any information other than "I have added a somewhat-related image that doesn't look obviously AI-generated to this", you can now generate such an image pretty easily. That's a low bar, but almost the entire genre of stock photos exists in service to this goal.
Likewise, there's a whole lot of day-to-day writing that mostly exists to communicate "I have written some text about this subject because I am expected or required to" and at most needs to communicate some simple information. LLMs will probably replace a lot of that writing, at least temporarily. As people increasingly use LLMs to consume that information as well, it will probably eventually be replaced by some sort of inter-agent communications protocol.
the article talks about AI books winning awards and AI books on bestseller charts, but can any one of us think of a book they genuinely enjoyed that was written by an AI? this feels less like proof that AI books are good and more like proof that awards, rankings and other accolades. it's a known issue across almost all entertainment media: you start by trying to find the best media to consume but the process is so subjective and gameable that it ends up being a way to pick something arbitrarily and then declare it to be the best.
I think this is more a testament to how little the quality of the book matters for the purpose of landing publishing contracts. Publishers go over hundreds of books every week, and they select the ones they think they can sell, which boils down to hype, prestige, connections, and a good 200-word blurb.
I'd say the most likely scenario is that nobody read the damned book until after they awarded the contract, and then they immediately said "wait a minute..."
Fantastic take. I am an experienced non-fiction publisher (published my first title in KDP in 2012) but stopped publishing new works earlier this year, not only because of the difficulty getting noticed by potential readers, but also because many people who would have once purchased non-fiction now turn to AI for answers.
It was always a hard sell, but now the math for publishing nonfiction doesn’t make sense. I canceled one project, the fourth edition of my most popular title, during the preproduction phase. It will never see the light of day.
I have also been sitting on one novel manuscript and many short stories for the past five years or so, and have been kind of frozen with reluctance to publish for the reasons Howey cites in his blog. It was already hard enough to get noticed for new fiction amongst a sea of self published works. Short stories are especially hard to sell, even in genres like science fiction where they once flourished. AI opened the supply firehose further while sowing doubt amongst readers about the provenance of works published after 2022.
On the other hand, I see that AI offers a path forward to tackle many of the structural problems in the novel manuscript that my editor flagged. But if I leverage AI to fix it, the finished book will always be tainted by the fact that this is not a pure work of human authored fiction.
What if we remove the stigma of authors using AI. We say, have at it, use all the AI you want. What then? The market is still finite. The authors are still competing with each other. They will still have to be good at creating compelling books that people want to read. Will the whole thing just blow over and AI would be "just another tool" that authors use to get their work out?
This has happened with other accelerants in various media/fields:
1. easy access to video recording and editing equipment has made it a lot easier to produce videos on sites like YouTube;
2. easy access to audio recording and editing (Audacity, etc.) has made it a lot easier to produce podcasts and other audio content (via LibriVox, Apple Podcasts, etc.);
3. easy access to writing software and self-publishing (Amazon, fan fiction sites, blogging platforms) has made it a lot easier to produce stories and books;
4. easy access to game engines and tools (Unity, Unreal, Godot, Twine, RenPy, etc.) have made it a lot easier to produce games.
Yes, and as Howey explains in TFA those things might be good for consumers and creators who create for fun, but they are not good for people who create in order to earn a living.
you are not taking into account that creators can be funded in different ways. For example like in Patreon.
You can give part of your money to fund a creator that wants to create in order to earn a living.
I believe that is better when people create for fun. If they need to create to earn a living they are forced to create and that is not the best incentive to create. Fun is a better incentive
> I believe that is better when people create for fun. If they need to create to earn a living they are forced to create and that is not the best incentive to create. Fun is a better incentive
I'm not sure that's true (thousands of famous creative works from the Sistine Chapel to the Benin Bronzes were created on commission/for remuneration), but even if it is, how are creators gonna eat? If the answer is "get a day job", I think that excludes a lot of folks from practically being able to create things. Patreon is not a great way to pay the bills for all but the most famous and popular people on it.
With IA, creation is cheap. It's inevitable that automation would make some jobs disappear. There's no solution, less people will be able to earn a living for creating art, I think.
I think that's probably at least somewhat true, and regrettable/something to be worked on in your community, and a very different statement from
> [it] is better when people create for fun. If they need to create to earn a living they are forced to create and that is not the best incentive to create. Fun is a better incentive
There will be less creators that earn a living by creating. But I believe there will be more art.
E.g. after Michelangelo created his style, then IA can mimick it almost for free. Imagine people with great talent like him, expressing their ideas and also being able to use IA to create, plus thousands of less talented people basing their AI art on them.
What if other LLMs were charged to read the work as well? And those other LLMs were actually in the persona of aspiring authors hoping to learn so they can later sell their work?
There's an entire potential economy there without humans necessarily involved at all, kind of busted by recent Anthropic precedent admittedly.
AI destroys your writing voice. There are authors who I enjoy not only because they write a compelling story, but they do so in such a way that is lyrically enjoyable. Engineers fall into the trap of thinking that if the slop machine emits the Internet-average sentence to describe a character’s actions, then it will suffice for all audiences and works of literature. “Add more adjectives and flowery language” will only net you a novel that reeks of cheap perfume. I guess we’ve lumped all of this under “taste” these days, but that’s what the AI author lacks. Without it, it’s probably going to be fine for producing summer beach reads (which could be quite lucrative!), but it needs to escape the mean it was trained on to be a truly unique, compelling voice.
I think I was fairly clear that I’m talking about these machines, not all machines that exist or could exist. If I wasn’t, please read “AI” as “state of the art LLMs as they exist today.”
> Why would there need to be a stigma for getting paid based on a tool they used?
Is it being used a tool, or is it being used as a ghostwriter? There is often stigma attached to that when done by humans.
> Artist have a stigma for using photoshop instead of canvas?
Yes, that used to very much be the case. Now it is often considered a different form of expression, so loses a lot of the stigma mostly by not pretending to be what those looking down on it are comparing it to.
Similar in the debate between CGI and practical effects.
Perhaps if AI-backed authors are willing to have their work classed as a different type of writing they can skip by the comparison-stigma too?
If they are actually doing some work (editing significantly after generation) then they are protected on two fronts (currently, active cases may change this):
* as discussed elsewhere in these threads, the judgements pertaining to copyright not applying to generated work all only seem to cover entirely generated work where there is very little human activity after the prompt.
* even if the generated work is public domain, the derived work (the edited version) is not necessarily so (much like Disney's derived works created from the tales of Hans Christian Andersen's writings).
If they are not doing significant work after the ghostwriting, then sod'em, they are earning the stigma of essentially committing fraud to get that profit from the work.
The issue is that word processors and photoshop replace the traditional technique of art. AI replaces the expression of art. Since the invention of the camera made representation trivial, the art community has moved towards viewing expression as the soul of art.
Surely having the idea and curating the result actually is work?
And pricing is determined by what someone else will pay for something, so if people want to read it and the author is charging for it . . . that's up to them if they want to make a trade about it.
The odd thing is this attempt at gatekeeping what other people value.
> pricing is determined by what someone else will pay for something
That’s so reductive that even Keynes would probably take issue.
Discoverability is the axis of competition in this context, not price. The market has not effectively found out how to identify quality in order to price it, so externalities rule instead: pre-existing authors with recognizable names, whatever genre publishers like at a given moment (famously disconnected from what people actually want to buy), how many energy drinks a crabby editor has had by the time they get to your manuscript, and so on.
Search engine and recommendation systems were supposed to ameliorate those externalities. They didn’t.
Price is determined by what someone is willing to pay but value isn't. If the value from art comes from expression, then LLMs directly attack the core value of art.
Damien Hirst, among many other prominent modern and historical artists, famously employed teams of human assistants to execute the bulk of the work. In the Hirst case that extended to the spot paintings, many of which he had very limited involvement in as a direct feature of the concept of the work itself.
Something tells me that doesn't alter the value of the art as expression of the intention of the artist, and not the assistant be it human or machine. In the spot painting case he was deliberately playing with this idea to the degree it added to the perceived value.
Part of the appeal of literature generally and the novel in particular is that it provides you with an incredibly intimate connection to the subjectivity of another person, though they may be long dead, or the speaker of another language, or from a far away part of the world that you would never be able to visit yourself.
I would argue that a story can mainly provide a connection with your own subjectivity. If it moves something in you, that's what mainly matters.
If your goal by reading is too learn about the subjectivity of another real person, then yes, read human authors. You said it, that's it's part of what a book can offer, but there can be other reasons for reading
> If it moves something in you, that's what mainly matters.
This is a values question. Many would say that literature has a role in modeling certain behaviors and teaching lessons that are functional in society. Basically the fables and proverbs idea. Simply catering to what feels nice and what reinforces the biases and wishes of readers is not the same thing. Yes, most of literature is already crap and simply caters to readers instead of challenging them. But if you RLHF your way to AI literature that people "engage with" the most, you will get unvirtuous literature that doesn't do this function. Of course you're welcome to disagree with this entire framing and purely see reading as hedonistic pleasure and nothing else.
True, but it's easier to get lazy and to just wire it up to an automatic engagement farming loop with no human input. Imagine the worst form of audience capture possible.
What's wrong with that? There's the low quality slop for the people who desire it, and then there's the high quality literature that's appreciated by people with good taste, and that cannot be easily produced
Off topic, but you name is somewhat ironic. Argentina's most famous writer, Jorge Luis Borges, built much of his career on the idea that writing is inextricably and inherantly linked to its author.
Completely agree with Borges in that. When the author is an AI, there's no human author, the result is a mix of what information it was feed, how its algorithms work, and what does the person giving commands to the IA asks. There's some link to the person who directs the IA.
This mindset is called "death of the author" [1] and the quick summary is that the author's intent doesn't matter. Most book people I know disagree with it.
I think about this a lot. For me, non-fiction stories hit way harder. I read a book about Amelia Earhart last month, and KNOWING it happened to a REAL HUMAN made the book 10x more powerful than a work of fiction about a lost pilot would feel. The same with documentaries; I love them because they happened.
However, I still watch movies about made-up things! TV shows that were written by a roomful of writers, all arguing and compromising, then taking notes from suits and producers, and making more compromises up to the day of shooting, even the day of editing! What comes out of that process is complete fiction, written by not-a-single-human-heart but rather a messy mash-up of humans combining their gibberish.
And sometimes it's still good. And I love it. Even though it isn't "real" and didn't come from a "real person."
So ... people may love a meat-book more in the future, but people will also love machine-books. Not you, by the sounds of it. But there are people not-you on Earth as well.
I should probably clarify that something like a television show, movie or game where it’s likely a group of people is still included in my “someone’s heart” even if it’s many hearts.
Let’s take Backrooms for an example. In the interviews with the cast, Kane, and others they usually all hit on “so and so was so passionate and added X”. Hearing about everyone’s passion and seeing the end result reflect that makes the experience of watching it so much more enjoyable.
If someone generated a Backrooms film, there would be none of that.
I’m not saying you’re wrong for liking machine books or anyone should be wrong for thinking that, but I just can’t imagine them being equal in the view of the arts.
Writing for the love of art, besides being a mental and reflective exercise that holds real value when it is of high quality, makes it hard for me to believe that this is the end of literature. It is difficult to think that someone who writes for pleasure would abandon it simply because of the exponential volume of AI-generated text.
> It is difficult to think that someone who writes for pleasure would abandon it simply because of the exponential volume of AI-generated text.
But the "exponential volume of AI-generated text" turns "writing for pleasure" into a solitary act. There's no chance you'll be able to quit your day job, because people like what you do any will pay for it. So many people faking it with AI means that it'll be struggle to get any attention at all for what you've put so much effort into.
The truth is much "pursuit of art" is a status exercise by those with the leisure capacity to do so, and this only maintains status because most people do not have the spare time and energy to put into it.
In the event everyone is sitting around with time on their hands wondering what to do and we have machines that spit out endless reams of anything it is a very curious question what exactly it is most people would end up doing, which would probably be some new status seeking exercise.
Whoa, fascinating: what to call this paradox I'd never considered before?
From the blog post:
> In my dream world, the robots do all the boring jobs and humans spend all their time being creative, connecting with one another, learning about the universe, and sharing our ideas and creations in a great big prosperous circle.
But with this and your other comment [0] about humans churning out 10 articles a day and changing the idea of what's "valuable" leads me to a bit of a conundrum I can't quite unwind:
If AI took over all the boring jobs (chores, labor, etc.) and we had all the time in the world to be creative, and we did just that, to the effect of humans generating dramatically more from "creative" ventures, have we created the same scenario, from a different approach? What then?
My guess is energy gets redirected towards sport, and everything associated with it. It's already often an abstracted form of things that used to be work (sailing, hunting etc.), but it's also something more unique that a machine participating in makes no sense.
'act' as in a play, but now with no audience. So all that hard work of memorizing lines, rehearsing, and vocal lessons becomes an experience only the actors can enjoy by themselves.
I agree, but I do think it makes the book industry more difficult. A flood of books makes it makes it harder for new authors to make it. Just like a new youtuber has a very difficult time. Eventually publishers will turn to AI just to deal with the volume on submissions, and then it will become just like YouTube, where an algorithm decides who wins and loses.
At the same time, readers will have to sift through an insane amount of books to find what they like. It is already kind of like this, even before AI. This causes people to rely more on reviews and word of mouth, which becomes a volume game and is also influenced by the algorithms.
I don't think it is the end times for literature, and for those who put the work in, they can find their places. But it is a whole new set of challenges.
It, plus video, also made painting culturally irrelevant. Approximately no one cares about paintings whatsoever. Books of modernity soon to follow I'm sure.
Interesting standpoint. Do not forget what makes you human, and that's why there are a lot of different opinions and valuable, subjective literary articles.
I suspect if everyone on earth produced subjective literary articles at the rate of ten a day you would cease to consider them valuable, let alone a machine that can create them on demand.
Thank you for your opinion. To be more specific, there are many subjective views on what value is. I didn't mean a literary piece always has market value if nobody is buying it, but it can still be valuable educationally as a masterful or sublime example of literature.
You are deliberately skirting the core problem which is your perception of value is completely warped by supply and demand.
If you have a surplus of something then it will cease to be of value to you. For example you do not actually "value" air, since it is everywhere and so we ignore it, but stuck somewhere where it is not you will come to appreciate it fairly quickly.
You seem to believe there is something magic and wonderful about literature, but this is because you have only ever lived in an environment where it is scarce and not dangerously abundant. The next stage is to claim "but it's not literature" even when it's the exact same strings of words merely constructed by machines and not humans, which is curious because up until now it had always been assumed that had the monkey succeeded in typing out Shakespeare it would still have been as valuable as had Shakespeare written it himself.
> I watch videogame streamers play games. Millions of people do. There will be authors who livestream the entire book-writing-process (just typing that out makes me want to launch one today). No one person will watch the entire process, but enough people will tune in here and there, and it’ll on be on YouTube
Ironically, I finally caught up on an online web series (though it's been adapted to a published book as well as audiobooks) and discovered the author does live stream their writing of unreleased chapters.
They're a bit too prolific to actually dig into their overall process, but you can watch them type for hours straight. It's awfully boring tbh since there's no interaction and they're understandably focused on their work.
I feel like writing has traditionally had two purposes, to communicate some internal feeling or idea of the writer, and to entertain. For the reader, even if their primary purpose was to be entertained, they would have no choice but to also engage with the internality of the writer. This latter point is what is changing.
I fully admit that we have been surpassed by the machine in the ability to entertain. And the biggest problem with that is that entertainment is the primary factor that drives the popularity and consequently the financial success of a book. If a writer's primary purpose is to entertain, then they should be very worried. But if their primary purpose is to convey something from within themselves, the LLM cannot replicate this. They may however lose that portion of their audience to whom the communication of profundity is merely collateral to being entertained.
I'm not a professional writer, but when I do write, I do so with the intention of expressing something that I feel, which before being thought out and written exists as something not linguistic. The language is a means for the expression of this deeper experience whose communication is the end. But an LLM on it's own is not able to convey internal experiences that it does not have. If it seems that way, it is only achieved by manipulating and rearranging the experiences of the authors that make up it's training data.
I think in order to notice and take offense to those tropes, you already have to be prejudiced against them. And you probably also have the pretense that what you're reading was thought important enough by another person to write. So when you realize that this isn't the case you become disappointed.
But some people just don't care, and are just in it for novel rearrangements of the narrative structure. I'll give you that I don't notice this so much with literary content, but I have seen people being entertained by AI generated music and videos quite a bit, especially children and those who in general are uncritical of what they consume.
For example, the other day I was at a restaurant where they were playing some AI abomination of Frank Sinatra called "Take me to the Moon", and nobody really seemed to notice or care while I was becoming nauseous to my stomach. You have to already be biased to notice it, let alone to even be bothered by it.
Conveyance requires a source and a destination. To have an experience conveyed to you means to gain a degree of understanding of another's experience which you yourself did not have. Yet you yourself have not actually experienced it, you have merely gained knowledge into some aspects of it's nature by comparing it to your own experiences.
If a reader perceives that they have been conveyed an experience that nobody has ever had, then they have made up an imaginary experience which does not correspond to anybody else's lived reality, which may or may not be possible to have. If an LLM conveys an actual human experience, it is only because somebody's actual experience has been anonymized and conveyed on to you.
Even within your own brain, conveyance != experience.
When you report on your own experiences - like the warmth of the sun - that report is not itself the experience.
The scary part about an LLM (p-zombie) is that it can report on an experience exactly the same as if it had one, and for all intents and purposes it'd be sufficient for the task at hand.
After all, I don't need to have your experiences to know what you mean.
I agree with conveyance != experience, and looking at it again I agree with your original statement with regards to the "!=". But I do believe that conveying experience requires having it.
By conveying, I don't mean conveying the idea of a possible experience, but conveying knowledge about a real experience that a real experiencer had. What is conveyed is not the entire experience which is impossible, what is conveyed is knowledge that is sensible to someone who has similar experiences they can associate it to.
If the report does not correspond to an underlying experience, then it's just a lie. Like if I say to someone "Thank you" after they do me a favor, they perceive that I am having the experience of gratitude. But if I am not really grateful, then my words haven't conveyed information about an actual experience, about an actual instance of an experience.
So I'd say the report + the report's correspondence to reality + the receiver's ability to understand the report based on their own experiences = conveyance.
I do not consider a false report that leads the receiver to imagine the existence of a non-existent experience to be "conveyance of an experience".
Contemporarily we have methods like radical honesty - another moral absolutist take that also asserts a primacy of truth, that as long as you tell the truth, you can do no wrong
For a lie always harms another; if not some human being, then it nevertheless does harm to humanity in general, inasmuch as it vitiates the very source of right... All practical principles of right must contain rigorous truth... Kant
So, that's the moral side of things. Whether or not you believe in truth as a moral compass (or even a component of it or some moral rubrik), I'll leave up to you! What you said:
> But if I am not really grateful, then my words haven't conveyed information about an actual experience
...gets more at the utility of honesty, as a conveyer of experience. While I might argue you still have conveyed an experience (just not one you had), because the references and end experience still happened in the mind of the receiver, the end (target) experience was had.
The invention of lying is largely due to it working just as well (or better than) the truth in the first place. Maybe the first lie was an accident, and evolution took over where only the best liars passed on their ideas.
So that you could even have an experience rich enough to contemplate!
The problem as I see it is that you cannot communicate your internal feeling if no one reads you book. Therefore, even for people whose motivation is personal expression, there is a very real risk of being flooded out of the market by AI slop.
For sure LLMs can write in perfect English (or any other language) and convey information.
But good writing is more than just conveying information. To me it's about connecting to a voice. This is a rare quality and AFAIK I have never experienced this with AI generated text or even with 99% of human written text for that matter.
I can flag a lot of content that does pretty well online as AI. There are certain rhythmic patterns in the sentence structure and the way arguments are set up that give it away. The writing is obviously "successful" in that it gets a lot of engagement, and many people either don't realize or don't care that it's machine generated, but I find it distasteful.
If I'm reading something I either want to be entertained or I want some specific information. If it's to be entertained then the story has to be good (writing doesn't necessarily need to be that good) and I haven't seen any AI generated stories that are that compelling. If it's informational then I can either ask an LLM directly and get something tailored for me, or I'm also interested in the emotional and human side, which an LLM can only imitate, since it has no theory of mind.
I think LLMs will take over making those trashy beach novels, as they're essentially clones. And people who like them will continue to buy them; no shade, some people enjoy that. But for good original stories and plots, you're going to need people to create that world.
This is a well-known author, but how come he thinks AI is better at writing? AI writing makes me vomit.
Also, I think he's too enthusiastic about the future of readers. In my opinion, reading books will be a very rare hobby pretty soon. The decline started more than 10 years ago.
The full sentence I wrote: "Artificial Intelligence is writing about as well as humans — at least well enough to cause confusion and speculation — and that’s creating an existential crisis among authors, confusion among readers, and legal action within the industry"
"About as well as" doesn't mean "better." Wild that you jumped from one to the other.
I continue: "-- at least well enough to cause confusion and speculation --" This is factual.
One thing I am sure of is that computers are better at reading than people. The way some people read makes me want to vomit.
I think this was fairly obvious and fair 1 - 2 years ago. Now though, I have to wonder if people who think this with such conviction are falling for the toupee fallacy.
You said it yourself: someone you consider a well known (and presumably accomplished?) author thinks AI is pretty good at writing.
I haven't tried using an AI to write prose, but I use Claude to write code daily and it spits out the worst comments I've ever seen, along with some surprisingly good ones.
Would we not expect AI fiction to be the same? A stream of poor and excellent taste, not particularly original or surprising, samey in hard-to-define ways, but capable of being hammered into a good story by an editor.
> Would we not expect AI fiction to be the same? A stream of poor and excellent taste, not particularly original or surprising, samey in hard-to-define ways, but capable of being hammered into a good story by an editor.
That could also describe many human writers…
AI writing seems to have a quality ceiling, or an ick floor if you prefer to think about it that way around, but is OK compare to bad human writers.
I'm still committed to the career suicide that is refusing to touch generative AI beyond the supped-up auto-correct/-suggest that has replaced the older method in my IDE (replaced, such that if I turn it off I also lose what I had before), but seeing results others around me have pulled out of the hat I have to admit the tech can produce serviceable work if properly fettled by a person afterwards.
I don't think you're committing career suicide. I've seen people get serviceable results out of it, and it always involves reading the code in depth. Reading the code being slower than writing it, they're almost never getting a speed up.
That being said, the better coders I've seen aren't looking for a speed up, they're generally looking for something to bounce ideas off of, or to get a first draft going to defeat writer's block or something (I guess this is technically a speed up, but not the way people mean when they describe slinging tokens around). It's generally to wind up at a better solution at the end than to get a sustained speed up (which seems to only be possible in boilerplate heavy domains, but those were previously slowed down by the boilerplate and are really only coming up to speed with everything else, at significant cost).
I think “taste” is probably the right word. Steve Jobs’ quote about Microsoft is very relevant: “The only problem with Microsoft is they just have no taste. They have absolutely no taste. And I don't mean that in a small way, I mean that in a big way, in the sense that they don't think of original ideas, and they don't bring much culture into their products.”
Replace Microsoft with LLMs and the same quote applies. They have no culture, no connection. The code or comments they produce will not effect any future LLM session and they current have no concern how it will effect future humans except for maximizing engagement for the human at the console today.
Fiction is no different. Unless there’s a feedback loop the LLM just spits out text and will never see it again. The LLM will never meet the readers at book signings, weave in personal details that could only be known for experience.
It’s probably perfect for the write-by-committee style of movies and shows being created today. Just like organic food or locally made products, there are already creators putting Human Intelligence badges on their work and there will probably be a market (though niche) for that until human supremacist laws push it underground.
> If this is the case it should be quite easy for you to provide examples of excellent AI-fiction.
I feel that's the opposite of the toupee fallacy. The lack of labelled excellent AI fiction supports the idea that there's a lot of excellent fiction that was written by largely by AI but nobody was able to identify it as AI, because they were looking for writing that makes them vomit.
There’s plenty of people who use good toupees/wigs and are open about it (or went bald and saved their hair with mystery medicine). I’m sure if models can now produce amazing fiction it wouldn’t take more than few minutes and Claude subscription to prove us doubters wrong.
Maybe that debut novel is amazing, but the fact that we are even discussing whether it’s AI or not tells me it might not be - in any case we can’t read it yet to judge ourselves - just that publishing agents are completely out of touch. If it were truly great I don’t think anyone would care, since hey, good writing is good writing, no matter where it comes from. We would just read it and enjoy.
Similarly models are for fact able to produce music and images that are very difficult to differentiate from meat made, but still… Most people don’t like them. Something seems always a bit off, kinda artificial and boring. Lack of imagination, lack of authenticity, lack of voice.
Taking the lack of ability to identify good AI fiction as evidence that there is good AI fiction is absurd, because it means that there is no evidence which could ever point to there not being good AI fiction. There must be evidence which could contradict a hypothesis for it to be sensible.
Just like how raw sewage tastes good, but people think it tastes bad because of how all of the raw sewage they've tasted tastes bad. But if only people ate more raw sewage, they'd find how much tasty raw sewage is out there!
There's been multiple 'unslop' contests [1] with a focus on AI 'writing' and I agree with the OP that AI writing makes me want to vomit. It's all hollow, bland writing that has zero zest to it.
Agree. I've just decided that I'm limiting my reading to pre-2021 books. There's still many lifetimes of a backlog there, even at my relatively high rate.
> People are going to eventually figure out how to polish and refine the usage and outputs so you cant even tell it is AI.
Personally, I'm gonna go offline after this and meet more people. One of the reasons I read comments is because it's human conversation. When everything will be ridden by bots — I'm out.
It does not matter which prompt you use. Even a long article with a small input (of what to write) will include weird details and structure.
People tried to fool Pangram by using custom prompts and instant feedback (rewrite until it passes). After hundreds of rounds, the text was still detected.
Although it's harder to detect such text by humans, there will be clues if you want to write a whole book based on a few pages of input.
>"This is a well-known author, but how come he thinks AI is better at writing? AI writing makes me vomit."
Most of all writing makes me want to vomit. It's not impossible to write well with AI, but it's the same argument as vibe-coded slop vs. a senior engineer using the tools carefully. I'd be inclined to accept the expert's opinion here.
He talks about this in the article, did you read it? Reasons include AI winning writing contests, AI hitting bestseller lists, and now apparently an AI-written book inspiring a bidding war between publishers that resulted in the supposed "author" getting a $2.4M advance (until it was revealed that the book was written by AI).
I've yet to read anything authored by AI that was truly excellent writing. It simply does not have the chops to write at that level. The great American novel will not be written by AI. Maybe "co-authored by Claude and friends", but not authored.
I see all of this as the cognitive dissonance created when humans are confronted with what scientists have been telling us forever: we’re not that special. We’re just piles of atoms, as are rocks, dogs, computers, and all of the other intelligent life that the universe must surely be teeming with.
Building something that supersedes us may have always been our purpose. Until skynet turns on us I’ll be happy to say I lived at the exact time it happened.
Humans are extremely special! They have a unique trait no other extant and few antecedant species have/had…
No other carbon terrestrial species can run for such long distances - a healthy young human can jog for 50 km or more on water and sugar: this is a (not too) breathtaking efficiency!
Can you imagine being a gazelle who has been sprinting away from a human for six hours but they just keep coming? Terrifying.
The ability to do so truly defines the species, to the extent they’re taxonomied as ‘homo marathoniensis‘ (fun fact: during the atomic age most human philosophers used ‘homo sapiens sapiens’ - more like ‘homo vanus vanus’ amirite).
And yet the cheapest car can easily outrun any human who ever lived. Everything is unique if you look into it closely enough. I would say that’s not the same as special, though I admit this is just semantics.
You could say that about any religion given only a one sentence high level summary. “A guy in the sky made us in seven days.” has no serious, philosophical grounding.
Your comment's first half was a kind of freshman nihilism backed up by vague gesturing at science (despite the fact that science can't really answer normative claims or value judgements). You then follow it up with a totally unsupported teleological claim which completely contradicts the paragraph that came before.
This is just a guess based off my own reading habits, but I can't help but think the average reader chooses books largely based on who the writer is. I don't necessarily mean that as a fan-boy kind of thing, but more just "I'm overwhelmed with choices and I know this guy is good". I know we'll keep getting stories of like "this LLM created book won this award" or something, but ultimately someone is going to have to put their name behind a piece of work and cultivate an audience and develop a unique voice and viewpoint, and like, I don't think the existence of a lot of spam is going to get rid of that. (I know that's a bit tangential to the blog's point -- I read it -- but I just think that the human element in a piece is going to matter the most)
From the post: "Doing a thing for the love of the thing is no different than it ever was. The problem is, there aren’t many people for whom this applies." I think this is an interesting point but I am not sure it matters when deciding whether someone should be paid for the work they have done. It is fine to have your opinion on what you think deserves your money. However it doesn't change that they have done work that has value regardless of their motivations.
> Artificial Intelligence is writing about as well as humans
Does he really hold on this until the end of the article? (sorry, I had no time to read through it). It doesn't matter how well and sophisticated the writing is, what counts is the human connection. A human is talking to me through the writing. I am not interested in having communication with a machine, I want the vibes only humans can send and receive …
Did you even make it to the end of that sentence? It would've taken you less time than writing your comment. I said "at least as well enough to cause confusion and speculation"
But apparently there's enough confusion to be had by reading a few random words from a blog post, getting your blood pressure up, and writing a comment to show off your lack of attention span.
Alternatively most blog posts could have been paragraphs. Most books could have been single chapters. Most newspaper articles could have been tweets.
That being said, most readers want you to hold your hand step by step. They don't want to the outline: they want you to give them a story. Most people and writers know this already. To me, it was new.
Sometimes pleasure is in the activity, not just the outcome.
A good movie or novel is a generally pleasurable experience from start to finish, where the conclusion on its own might have only been mildly interesting at best.
> Alternatively most blog posts could have been paragraphs. Most books could have been single chapters. Most newspaper articles could have been tweets.
Why even read? Why even live? What's the point of all of this? Churches could be wooden shacks, 5 stars menus could be protein shakes, paintings could be single sentence descriptions, ...
You might as well sign up to be put in an artificial coma, or in a matrix pod
For most everyone else, it is to momentarily feel some sort of way. The difference is what the parent describes as a revelation, to him.
You see, for people who crave to understand, living among people who just want to feel this and that feels like living in an insane asylum where the people inflicting harm upon themselves and others will do nothing but distract themselves long enough to forget and go on to inflict harm on themselves and others again, forever.
As for 'might as well quit' - most everyone I've known who's not an idiot has already done that. Even the idiots aren't having children by the way :) Living in a matrix pod? You mean a modern city? Anyhoo...
I've colloquially coined this as the contentification of everything. The goal isn't to actually engage with something, whether it's the writing style, flavor, or artwork on display. It's purely about consumption, and the most optimal way of consuming something is distilling it down into a short, bland summary.
It's why there are people out there that will use an LLM to summarize a book and pretend that they read the whole thing because they know the general outline.
I keep brooding on the idea that maybe LLMs will never get “good at writing” in some Platonic sense, but they will be progressively trained better to create text that a greater and greater bulk of humans will “enjoy” for reasons that are downstream of inherited patterns of thought and emotional response that the models have learned.
I've never read Howey's books but have them mentally filed next to, like, Dungeon Crawler Carl, the Bobiverse, whatever that long thing is with all the superheroes, thousand-chapter cultivation LitRPG blah blah blah as Reddit-beloved high-wordcount fan-fiction-adjacent genreslop.
I hate being a snob, but when we read "artificial Intelligence is writing about as well as humans", maybe take that with a grain of salt? It has not been my experience that AI writing is remotely near a "good" human writer.
I don't see it as abrupt as it was portrayed. Man-made artifacts and work has always been taken over by machines in all fields of work and art. The "end of era" happened many times, through intermittent ruptures like how the tecpn8x plates move through earthquakes.
Loss of human touch is a problem. Playing a chess game while being aware that the opponent is a bot, is not same as playing against a human.
The "plasticification" is something we wanted all along, supported it through excitement and awe. Science is the plastic.
The mediocre may livestream, but I suppose the worst and the best, in seeking escape from the middle voice, now gifted to large language models, now not our own, that is, I say, the best and the worst will become esoterically poetic, will run on in agramatical authenticity, in voluminous polyglot, tiresomely peripatetic through the thesaurus. Loose change, slang, new coinage, words lazarized, older and rarer englishes, scattered everywhere.
Nonsense and beauty ascendant! Poesy freed from several centuries’ fuligin tyranny of the novel and horrid mass literacy; windmills again a just and proper enemy!
This new muse will mate with the corpse of epistemology to produce a second coming of baroque obscurant gnosticism. A fine time to be a scholar or a priest, the 2100s.
Before the printing press, information was transmitted by word of mouth, skills were learned by joining guilds, and the largely illiterate public lived and learned from songs. As AI takes over and information becomes adulterated, people will enter their own bubbles. Ultimately, the only way to escape enslavement by AI will be to return to the world before the printing press.
I scrolled through the comments as I am curious if anyone shared some ideas: What is the future of authoring by humans like? Instead of books being one of either ‘meat’ or ‘machine’, as this author is claiming, what would it look like for human authors to leverage AI for simply getting assistance for the stories they author? Could there be a healthy, non sloppy use for AI?
If you are on top of the AI, it is a powerful tool for exploration of ideas, organization of thoughts, bridging connections, outlining, filling in gaps.
But if you vibe write, it will look good on the surface but be full of illogical, contradictory statements, disconnected characterizations, etc.
AI writing is not good. Seeing this post, and another comment on here earlier today to the effect of “we’ve solved prose”, I’m genuinely really confused. I don’t want to be insulting, even if the first place my mind runs is basically people must not read very much or think very hard about what they read. I can’t imagine a world where anyone ingests a material amount of AI writing and things it’s even “acceptable”. But so many people seem completely blind to it. I’m actually horrified by the idea of a world where people are just slinging empty slop back and forth at each other. Probably many of us have already experienced it at work.
So what am I missing? What’s the most charitable explanation for how anyone can think LLMs are good at writing?
AI 'writing' makes me ill, but that's happening because I can 'feel' the prose's samey hollowness in every paragraph.
But! If someone is just moving word by word and is focussed on just the gist (the average reading level is '6th grade' here in the states), then it will not repel them the way it does us. Also, AI's grammar is impeccable, if bizarre, and the word choices can be facially erudite, so that might also be mistaken as a signal for competence where none exists.
As someone who does a lot of writing, I ended up doing a lot of re-writing of what I'll call reference architectures in a previous role. I did use an LLM for some of the missing context up-front. It was essentially boilerplate and I did revise it--took out some things I didn't think important and added some things I thought might be.
But I basically never use them for a blog post except maybe to include some boilerplate explanation of some technology. (Which I'm familiar with but don't want to rewrite or copy and paste.)
Could that change? Perhaps. But I don't really use today.
Perhaps this is a question of ethics. Should we come to a conclusion that human effort makes a text worth reading despite it's objective quality. Literacy is humanities by definition.
Hijacked signals. When they read good writing they get a certain emotional response. This is not triggered by the totality of the writing but certain aspects closer to the surface level. They also get it from reading slop because it hits the same triggers.
It's like how evolution taught humans that sugar indicates a good densely-packed food source, so then corporations just started putting sugar in everything and humans eat it as ravenously as they eat actual food.
I read his take as not that "we've solved prose", but that we've created the McDonald's slop food version of prose, and people who are undiscerning will cause it to succeed and dominate the culture. While the effects of other people eating McDonald's affects us all in indirect ways, the effects of other people accepting lowest common denominator slop art/labor is going to be much more dire.
If a novel is 70,000 to 100,000 words, then a whole novel can fit in the context window for AI. That has to be good for the editing process. I.e. you can use AI to scan the whole book for various issues over and over again until all those issues are found and fixed.
Two of my favorite (sci-)fi writers both use crazy big Excel files that almost crash their computers when loading. For research but also keeping track of story loops.
I can imagine something like ChatGPT Work with markdown/python/CSV's can do quite some magic for such writers.
A common writing tool is to use a story grid. You have chapters/similar along the Y axis and title, characters, plot elements, etc. along the X axis. That way you can keep track of what information you are revealing at each story beat/chapter, make sure that plot twists work (e.g. when outlining something like The Sixth Sense).
Depending on how big the story is and how many characters/plot threads there are, I can see the Excel files getting large.
There are already tons of specialist non-AI tools for writers to be able to keep track of story elements, research, etc.
Writers (older genre writers particularly) tend to be a bit behind the technological curve, and they stick with whatever works. If you want to see a treasure trove of "old web" sites, look up the site of any sci-fi or fantasy author. Chances are its design and layout still look like something from the late 90s.
The vast majority of AI apps don't actually solve any novel problems (pun intended come at me,) they're just new versions of existing solutions "but with AI."
I wonder how concerned writers and publishers are about copyright of AI written text? As I understand it, in a lot of jurisdictions, AI products are not covered by copyright.
If my understanding is close to correct, than I can see why the million dollar advance was withdrawn.
How does one PROVE AI or human writing to determine copyright?
Who is challenging a lack of copyright on a work?
You publish a story, I take it and republish it / remix it / whatever, you invoke copyright protections and try to have mine taken down - how do I prove your work was AI and thus has no copyright, and how do you prove (or do you even have to) that it's human?
Who has the burden of proof - the person claiming AI created it, and thus it has no copyright, or the person who claims they wrote it and it does have copyright?
Well, copyright claims are usually a civil matter (in the US at least). For a civil dispute you have to make a judge or jury believe you more than the other person.
But it sounds like you are talking about a criminal charge. There the bar is higher and the government has to prove beyond a reasonable doubt. In that case I suppose the government could do things like search your computer, subpoena the AI companies for their records, talk to your editor, talk to your family and friends, etc... All the usual stuff.
Before you would ever get to a point where someone claiming your work was AI would be able to get access to your computer, editor, family/friends, etc there would need to be a bar that needs to be cleared for a judge to allow that discovery to take place.
What I'm asking is what is that bar?
Copyright is going to be assumed by default - in order to break that assumption there needs to be some process... however for all the people who are "concerned" about the copyright status of works that have touched AI I've never seen anyone make a feasable argument on how it would ever practically take place.
> There will be meat-book fanatics. […] The author will matter to them. They’ll read books the way audiophiles listen to vinyl. Every typo a pleasant pop and hiss.
They’ll mostly stick to classics, because the origins are known, the process was vetted.
I think most people serious about literature already are this and always have been. Y’know, it is not hard to write a novel. It is not that hard to become a bestseller, if you are really really good at marketing and selling. Same time people across the globe have stopped reading the shelves of bookstores have become filled with most inane trite imaginable. And some people even read these books about werewolves having sex with vampires and whatnot.
Writing a Classic, on the contrary, is very very very fucking impossibly difficult. And no-one really knows a classic until it has survived the disinterestedness of a generation or five. Kids have now apparently found The Count of Monte Cristo recently. Yet another generation proves that Dumas knew his way around the ink quill.
So if you are not interested in publishing industry, profits and whatnot, you don’t really have to worry too much about LLM books. Only thing you need to do is wait a generation. Hell, few summers and no-one remembers most of the books published today. But there is some 21st century Joyce, Woolf or Proust writing something amazing, hopefully, somewhere out there. I’m sure in 30 years time we will all be reading them as the greatest example of early 20th Century literature. Meanwhile, there’s plenty of old books last a few lifetimes.
> Artificial Intelligence is writing about as well as humans
This is just untrue.
There's a lot of money in claiming it's true, with varying levels of hedging and insincerity, but AI writing follows the pattern of AI everything else: if you're not competent to judge quality, then you think it is high quality.
Which is an excellent example of AI writing being potentially lucrative, but not good. Also note that that is its own counter-example: it wasn't published, and the advance was retracted.
> it wasn't published, and the advance was retracted.
Not because they didn't think the writing was high quality. The quality of the writing is what led to the bidding war. The advance was retracted because the publisher can't get exclusivity if copyright protections don't hold due to the novel being AI generated.
I think this is much more likely than any of his predictions: "buying and reading a book" becomes even less of an isolated activity than... it already is becoming. Instead it's more and more a media product, attached to authors/characters that are big in other formats, specifically, video. Because people are reading less and watching more video to begin with, entirely separate from AI.
David Goggins is kind of a perfect example of this. There is zero value in an AI-written book that isn't attached directly to his public persona and media presence. Maybe today he can use AI to write the book, but it's still tied to his video media of doing tough challenges, etc. An AI-first author writing such books would get nowhere, as he/she has no credibility.
That all aside, I actually think the most philosophically interesting result of LLMs is that ideas actually end up mattering a lot more than execution. The LLM can't really think for you, and so your knowledge and experiences are actually what matters, as it determines "what you prompt."
In short, the authors that win the future will be ones with robust personal brands and a big social media footprint. The days of the mysterious writer that gets chosen and published by a small group of elites at the publishing house is probably over... at least until the ecosystem becomes so messy that curators become necessary again.
Books are a physical medium of organizing and distributing information by gluing a bunch of flat papers between covers. Properties like originality do not follow from the physical medium itself.
Publishers of travel guides in book form use the physical medium as an instantiation of the class of "interesting coffee table adornment expected to contain wit". Does that imply an expectation of originality? I would argue it does not.
Publishers of speculative science fiction, on the other hand, instantiate objects of the class "examinations into the human condition using technology as a lens". For this class, the use of LLMs is an abomination.
There are other subgenres of science fiction, of course, for which these sensibilities would be misplaced.
It's the end of many eras, all at once. I think the same dilemma will appear for games, music and film too. As our ability to detect AI content deteriorates most authors will resort to it, at varying degrees.
Again - did you even read the article you are commenting on? The author gives an example of a novel that had a huge advance offered on it and then pulled because of AI concerns... in order to get to the point of having the offer a bunch of competent writers, editors etc didn't notice.
no, you don't have 1M counter examples. I'd also wager the vast majority of examples that you do have are on older models of LLMs that have been vastly surpassed in capabilities in the last 6 months.
You are essentially talking out a hole on the bottom half of your body.
It's annoying to come here and read 80% of comments by people who clearly didn't even read the article in question, but decided this was an excellent time to trot out their anti-AI screeds that totally fail to engage with the actual article. Human slop.
"But then I have to remind myself that I don’t sound like AI, it sounds like me. I’ve been em-dashing and comma-splicing up a motherfucking storm these past two decades. It was fun then, and it’s fun now."
The em-dash thing really annoys me. Using them has been house style the last couple of places I worked and I'm not just going to stop using them because some idiot might think they were generated by AI.
since when is "confusion and speculation" an indicator of quality/goodness? if I ran around naked in traffic it would also cause confusion and speculation.
It's quite clear that the author means AI writing is now good enough to cause confusion and speculation as to whether a given piece of writing was human-authored.
The confusion and speculation caused by your nude transit adventure is about what the hell you're doing and why.
I think he's talking about it more in the sense that there will be speculation about the provenance of some work.
For example, if desktop fabrication tools get cheap and good enough, maybe one day garage tinkerers will be making Patek knock offs that an expert with a loupe, spectrometer, and a lot of time can't definitively say that any particular watch made after 2034 (to pick a date) is genuine. That would cause a lot of confusion and speculation in the market for used watches.
I think that's what his crypto bros comment was all about (aside from being a joke). Maybe crypto tech could provide evidence to support provenance claims?
You're misunderstanding ... that's not at all the claim. Confusion and speculation is a consequence, not an indicator. Running around naked in traffic might well cause confusion and speculation, but it wouldn't cause anyone to confuse you with an automobile or result in speculation that are one.
At least AI is better at reading than are most people.
I initially interpreted this as "Artificial Intelligence _is_ [defined as] writing about as well as humans", which would imply that common LLMs are slowly approaching Artifical Intelligence. Though English is not my native language.
> In my dream world, the robots do all the boring jobs and humans spend all their time being creative, connecting with one another, learning about the universe, and sharing our ideas and creations in a great big prosperous circle.
Bertrand Russell famously predicted* that machines would allow everyone to “work” no more than 4 hours a day. I just had a thought when reading Hugh’s article - what if work is by and large the thing that connects most of us with one another and compels us to be creative and learn about the universe? And TBH the thought already feels trite in advance; both religious and wealthy people have said this before for various reasons.
I’ve had the same ideal in mind for many years. It would be amazing if I could work less and devote more time to art & music, reading & travel, socializing & community projects. But using AI for work (writing code) isn’t making that happen yet and doesn’t feel like it’s even aiming in that direction. Now that we have machines that can do most of my job, I’m starting to worry because there are some signs that it might not lead to utopia, starting with how it’s beginning to demotivate otherwise strong and curious programmers from wanting to be better programmers.
I don’t know, it’s too early to draw conclusions. But is anyone feeling like LLMs are giving them more time? They’re helping me work on more things, and produce more output. They’re helping me context switch more easily and work on things between meetings. They’re also sometimes tiring, and I’m often using them instead of talking to someone. They’re not yet freeing up time, and it doesn’t feel like lack of better models is the problem.
Nice blog post but this one part is so irritating:
"In my dream world, the robots do all the boring jobs and humans spend all their time being creative, connecting with one another, learning about the universe, and sharing our ideas and creations in a great big prosperous circle. I believe such a utopia is possible. It does not violate the laws of physics, so much the laws of human nature."
What irritates me is when people say "human nature" but what they actually mean is "capitalism" or "the profit motive". Common among people who have fallen into the trap of what Mark Fisher called "capitalist realism", a capturing of the imagination by the cultural concept in the west that capitalism will be with us forever, or that it is the natural state of humanity.
It's also human nature to be irritated by your own reading comprehension (based on biases and assumptions you carry around daily) and then attribute that irritation onto a stranger instead of working on yourself. :)
You've created a throwaway account specifically to post against the HN guidelines:
> Please don't fulminate. Please don't sneer, including at the rest of the community.
> Please don't post insinuations about astroturfing, shilling, brigading, foreign agents, and the like. It degrades discussion and is usually mistaken.
This is the black and white perspective that literally every other human-creative industry is wringing its hands over. But I think it misses a key difference that is already painfully obvious, which is creating new and engaging things will never be as simple as pressing a button. There will always be a human at the other end, pressing the button (or a human who deploys the agent, etc.)
The best comparison is photography. Instantly, a whole career path of portraiture was tossed out the window. Go to a museum and you see portrait after portrait - the best ones. There were millions of shitty ones made at all levels of society, and that job was replaced by cameras. But painting didn't go away, it got weird, it became anti-camera and more human, in a way.
This splitting of purpose and amplification of humanness is going to ramp up. Expect everyone to grow a taste for weirder, unprocessed music, experimental novels, wild and unexpected designs. Art is human communication and AI brings that into sharp relief.
Novels seem the most protected from AI - pulp/romance/beach novels will be crushed by volume and the sad remains of the publishing industry along with it, but interesting human novels will be just fine. And, much like with photography, I think AI will eventually normalize to its own artistic medium with writing, where careful usage will allow more ambitious work to appear that wasn't possible before, like a novel written using only sestinas, for example.
Publishers and "professional editors" have been releasing crap for years, even decades. No LLM needed. Humans are perfectly capable of writing so-called "slop" without a machine.
What annoys me the most about current day and age is PROLIFERATION of everything…
Not so long ago the author / creator would have to be very skillfull and prolific to write-and-write-and-write… nowaways I fee like I (we?) are drowning in more and more content…
The problem of AI writing, to me, is its emptiness, not the fact that it's not written by a human. Seeing standard slop words or phrases triggers the "do not read this" reaction, but the reason of not wanting to read is that the text is just empty of any interesting thoughts.
The majority of us read ChatGPT/Claude/Gemini responses, even though they are, by definition, slop, because we probably asked some question (the one we did not really care about and just wanted to explore) and the response has some valuable information in it (maybe true, maybe not). So we read it.
If (or when?) the AI writing becomes interesting to read, I don't see why we won't start consuming it, and I don't see it's as something bad.
One way to make the text interesting is to actually
get involved in writing; remember this short story that was here on HN some time ago? [1] It was written by AI with some serious human involvement, and I enjoyed reading it.
The biggest problem facing authors is not battles over whether something is AI vs. human. Most readers don't care.
AI has made creating content almost free. What it hasn't done is add free hours to our day to consume all this new content. The number of people who want to read a book this month is roughly what it was five years ago, and roughly what it'll be five years from now.
Supply is exploding into demand that is stagnant (at best) or falling. Upload a book to Amazon today — good cover, sharp title, a blurb you sweated over for a week — and there is no longer any reasonable expectation of an organic sale. The book goes up, and then nothing happens, indefinitely.
Five years ago that wasn't true.
The answer is paid advertising. Buy your way onto the shelf. Stand out from the slop. But every writer is being given the same advice at the same moment, and there are only so many slots. The price of the slots must rise. And will keep rising for as long as there are writers able to pay.
The old metric was it takes about $5 to sell a book via paid advertising. I think in a year we'll look back on that figure fondly.
This is pretty much what I've been saying for 20 years, but with content in general. People's time is fixed: 168 hours per week to: work, sleep, eat, excercise, commute, and then....
- read
- watch content: a movie at a theater, youtube, short-form socials, streaming services, etc..
- read a book
- scroll on the phone (won't list all the apps out)
This is AI-writes-too-well is a problem with many solutions. Here's a slightly unconventional one: if an author has enough skin in their book, they could just do as Tolkien and sing their songs[^2], in public. Taking that one step further would be to establish rapport with readers by teaching them how to play those songs, e.g. in a piano or a guitar. It's like going back in time, to when the Iliad was sung and performed instead of written. Though there's no need to negate technology; if a writer is sufficiently technologically inclined, they could forego the linear story and instead release interactive fiction for musical instruments...Hm, I'll see myself out now.
> Another prediction is that MOST readers will care about machine-vs-meat as much as they currently care about publishing imprints, which is to say almost none.
This seems like a truly wild prediction to me.
LLMs can synthesize data in useful ways for all sorts of purposes, and I can see how they could be helpful in the process of a human editing and refining a work of fiction they are creating, but I can't imagine why I'd ever tolerate a work of fiction actually generated by an LLM.
The authorship and craft matters, the translation of human feeling and thought to the page matters. This isn't code where it either does what it's supposed to or does not.
It's easy to confuse because archive.is uses several domains like archive.ph and archive.today so you can see how easy it is to think they are related.
Also important: archive.org (i.e. The Internet Archive) is a non-profit, while we don’t really know who runs archive.today and it has some shady history, including tampering with archived pages.
I started writing because AI made me obsolete in software development. Or at least gave me depression that blocks me from writing code. What I did instead I started writing my thoughts as short stories and then made an attempt to write full blown 50k words novel. It worked as therapy surprisingly well.
I write mostly without AI but then I use AI for editing and grammar correction. Sometimes when I struggle with plot threads I ask AI for an advice. When I write about something that a have a vague understanding about I may ask AI to write a paragraph or two. I'm also open about it and warn reader in preface and mark generated text using italics.
As a new author I find AI both helpful and discouraging, because it CAN write better prose than me. I think text that's written for humans should be written by humans, but temptation to ramp up quality with zero effort is huge.
Then I experimented with different regimes when I only allow myself a couple of prompts at the end of the day when I done writing and only allow minor edits. At least my conscience is clear.
Writing gives meaning, despite fact if anyone reads your stuff or not. Surely publishing a book is a major achievement but many authors wrote without publishing for many years. It's a therapy and getting a book printed is an achievement, but it's not mandatory.
It works like talking to yourself, your conscience and you also get a gut feeling about using AI in humanities with time. It starts to _feel_ bad if you go too far with it. It feels good when I write my text myself. Using AI for humanities feels like borrowing on bad terms.
I think that Silo is just a darker, stupider City of Ember. Like there have been shitty derivative works long before AI came along. Now there’s AI that can plagiarize faster and better than humans and those people are sad.
I think it is time to jail those CEOs that drive up the
prices and destroy industries. RAM prices up 350%? ALL
AI companies are in my opinion a mafia structure that
must be eliminated. They not only force us to have to
overpay (or be stuck) but also destroy industries
downstream, thus eliminating jobs. All that cost should
be paid by Microsoft, Facebook, Google, Amazon etc...
The increase in prices is a signal to producers to increase production.
Email eliminated a lot of post office and paper production jobs, does that also annoy you? Washing machines eliminating jobs about washing clothes by hand?
If you like a low tech life there are communities that choose to live like that, and happily do.
In my view the best writing presents a portal to someone else's mind. LLMs do not have a mind. They can produce text and perhaps even 'weave a tapestry' of plot lines that make for an enthralling and well written story. Maybe a meaningful one, if prompted with a particular set of themes and narrative ideas. Is that a voice?
I don't know. But I do think, or rather I hope, real voices will prevail. But publishers (including KDP) have a rough road ahead of them, if they also want that. Just search YouTube for tutorials on how to spam Amazon with AI generated works, hoping for a break out. It is sad.
Is anyone working on cryptographic human-crafted attestation, without feeding the work into LLMs?
yeah for me it's all about specificity, like no one else but Philip K. Dick could have written VALIS, it is so tied to his life and psychology. Its difficult for me to imagine something which is is build on huge amounts of various cultural output to retain the specificity of a singular limited human life-perspective.
Is it just a case of throwing more and more resources at a pattern recognition machine with no upper limit except the energy of the universe and the economy of the algorithms?
Someone else suggested the maximum intelligence is capped as the average intelligence of its training data, which is an interesting take.
François Chollet made an interesting geometric analogy on Twitter. We tend to view Intelligence as something unbounded, like the height of a pole. But it could be something closer to the roundness of a polished sphere. Once close to the limit, incremental improvements are less and less significant.
The intuition we have about the nature of cognition and its whole shape and limits is almost entirely built by looking at ourselves and our peers, comparing that to an ant. I feel this is largely naive and misguided.
I've been interpreting AI task solving capabilities as a search problem: being smarter lets us prune more of the tree and try out different things more efficiently. I think this fits into the sphere analogy quite well: at the limit, we need to experiment, use the experiments to create new abstractions and build on them repeatedly. This doesn't necessarily require more raw intelligence, but the time or resources to find how the pieces fit together.
I read the author's entire post and find it interesting that the author attributes success to having a good story to tell based upon some wild youth experiences while sailing. Not being able to type real fast, not some magic sauce from a "how to be an author class" but being able to spin a good yarn, one good enough for other people to pay to hear it.
There was/is a business position as a middleman of sitting between a storyteller and the printing press, called authoring or writing and that's slowly been denigrated for over a century by mechanical typewriters, word processors, spelling checkers, grammar checkers, now LLMs.
However there's always been a shortage of good story tellers with a story to tell, and for at least a century or two been a MASSIVE oversupply of the middlemen mentioned above, who end up failing. They can type, and they insist thats all they need, and the ability to spin a good yarn is just fluff or unimportant. Then they fail, so they're probably incorrect.
I suspect as long as the author can spin a heck of a good yarn based on his personal sailing experiences mixed in with whatever else, the author will have gainful employment in the book biz. As a storyteller. Not as "a guy who types really fast" or "a word processor technician" or whatever.
The middleman author who used to have a job writing generic romance novel #7994126 and was hoping to make money off writing #7994127 is probably utterly doomed. Why does the publisher need him if we have LLMs, or even better, why do we need publishers if we have LLMs and can enter a short simple plot just as well. "Claude generate me a 75K word romance novel similar to 'Just for the Summer' from last year but write all my favorite past characters into it. Especially character Anne she's my favorite."
But lets be realistic, those middlemen were already doomed anyway. Very few people in the book biz can spin a good yarn and that shows up in the stats of "failed writers". LLMs just mean they'll generate more scrap paper along the way to failure. But about the same percentage will sell books and about the same percentage will fail unless AI itself causes a substantial difference in readers tastes (honestly, I think not...)
Compare it to programming, about which we supposedly have more experience. LLMs generate more code that no one is responsible for nor understands fully. But does it make more or less successful startups? Nah, not really. Doesn't seem to have much effect on "idea people" at all. The people whom always generate bad ideas crash the plane with no survivors a little quicker, the people whom always generate good ideas make money a little quicker, most people in the middle are unaffected beyond change for the sake of change, which is always expensive.
Deleted because I realised I am not in the mood to argue with people who don’t read; don’t make an effort to understand an argument; or make shallow dismissals thinking they consist in any way of a valid point. There are more important things to worry about than internet randos (and that includes me, if you’re offended).
I think his “ALL AUTHORS” point was about how authors will use AI in some form or fashion during their process, and points to how he did with his google search and his AI Gmail response. I don’t think he meant all authors using AI for their writing. Here’s the snippet:
“Another thing that will happen is established authors will turn to AI for assistance, not just for writing but for research, cover art, editing, returning emails, having an assistant, doing marketing, and thousands of other things that haven’t been invented yet.”
I take it as a general, this-tech-is-being-shoved-down-our-throats kind of way
That's how I meant it. Love to have read the original comment telling me about all the authors out there who are somehow able to avoid the AI that is now in EVERY SINGLE THING WE USE, including our email and web searches.
LLMs are writing as well as humans for people who think writing words is the goal. Printing presses also write words better than humans. LLMs are solving the problem one layer higher than printing presses, but still several layers down from the top.
Note that LLMs aren't useless, since they have other strengths which can make them equal or better than humans in certain applications. For instance, they can tirelessly review code for silly mistakes, making them useful in cybersecurity.
Finally, someone else understands that writing is not the goal, the entire point of all of this is our communications between one another and this market for narrative stories is a nice departure and it does convey important knowledge and moral lessons that larger society needs. But it is an indirection, an indirection to the mission of communicating solutions to how to live as a human and that is why we care, why the interest persists despite lifetimes of narrative dreck. We are adrift without instructions, and no religions are not the instructions.
I think the first goal of writing is to talk to yourself. Without an external, unchanging scratchpad, most people cannot keep their thoughts truly straight. It can feel otherwise, but trying to write something about any tricky topic even a few pages in length can take a long time to be self-consistent. Clarifying it takes even longer, especially when you intend to share it. Writing as thinking is definitely a thing. I believe it’s the trickiest part of understanding how such interesting things pop out of a language model trained to predict the next token and rewarded with pleasing sequences. Since it is not thinking, it proves that certain sentences (most?) can be generated thoughtlessly.
I like the term I've seen someone use for LLMs: a "thought partner." Truly it's the duck programming duck I've always wanted.
I say that LLM's are a pacifier. LLM output is to conscious output as a silicone nipple is to a breast.
Sure, at some level, there are some people who at least some times, cannot pecieve a difference. That does not prove that there is no difference or that the difference doesn't matter.
Don't just focus on the model, half the turns in a session belong to the human, it's an instrument that depends on how well the user wields it. It does not need to produce the novel insights, just to support the user to have them.
No.
I think narrative stories are a "brain hack", in the same way heavily processed food, drugs, clickbait, soap operas and TikTok are brain hacks, combined with some exclusivity status signaling games. They're just old enough to have gained a cultural cachet that more modern forms of entertainment don't yet have. Same for art, music, comedy and culture generally.
There is nothing wrong with those "brain hacks", if they make us happy and don't have negative consequences, why not use them, but we should recognize that it's what they are.
I think that an alien intelligence could communicate with us about Ohm's law and Rieman's last theorem, maybe even debate resource and power allocation strategies (democracy, monarchy, capitalism, socialism) with us, they could probably understand the pull of sports, even if they couldn't participate in it themselves, but art and humor would be completely lost on them and considered a weird human excentricity.
This sounds like a fully general argument against being alive and having experiences; you could equally well call love a "brain hack". I also care a lot about solving practical problems, but the reason it's important to solve them is because it leads to human flourishing, which includes participating in all these activities you're deriding. Otherwise why are we doing anything at all?
The "story form" is a brain hack, i.e. a way of disseminating "problem X may cause Y and be resolved by Z". Whether that's useful depends entirely on what's being communicated and how.
Art, music, humour, poetry are more basic and the commonality is more or less "I bring you surprising phenomena". If phenomena are surprising to you then your mental model of the world needs improving.
What's Riemann's last theorem?
Culture and art can teach something to its audience, unlike most of the garbage you likened it to. It's clear you don't hold it in much regard, since you compare it to outright scams (clickbait) and even imply that consuming fiction is basically done for virtue signalling.
But like it or not, humans are not computers. They are animals with messy, complex thinking and reasoning that is impossible to fully describe, and incredibly complex societal systems created from their chaotic interactions. These messy people don't learn from datasheets and reference manuals, even if you'd like it to be otherwise.
Of course, there's a place for non-fiction, I like some of it a lot. But art has subtlety to it - instead of smacking people in the face with data and lists of objective measurements, it can pry into them by using the human ability to imagine themselves in other shoes, experience emotions that aren't results of real life events, and much more. There is a reason why parents teach their children values by reading them fables or educational stories, not the legal code of their jurisdiction. And this is also why fiction actually changes the minds of people reading it, arguably doing it more effectively than nonfiction. Slop (clickbait, average tiktok content, whatever else you're imagining) doesn't make you feel new things or think of something you haven't thought of before, art does.
The aiens are also a bit of a copout. You're just assuming that there's bound to be this perfect species (that are effectively computers) that operate only on mathematics and discrete logic, who are obviously superior to us and can show us our wrong ways. What if computers aren't always superior? Or what if organic life requires some of that messiness you don't like to be viable? What if the aliens had their own weird things we could never relate to?
We already suspect that many non-human animals have a "sense of beauty" and a sense of humor. At least one other species spontaneously dances on the "beat" to music (parrots).
I think alien intelligence would be able to appreciate art and humor, even without necessarily feeling it viscerally, as we do. We know that our particular brands of them rely on highly-developed and specific aspects of our psychology, but at a basic level, they seem to fundamentally be about congruence, incongruence, expectation, and interpretation/perspective. Any embodied intelligence that can observe physical reality is capable of grokking those concepts.
Shit engineers say
Fact. Someone has missed the point of existing. Or failed to come up with one for themself.
I'd like to live in a world where people only make art to serve a higher purpose. But we don't live in that world and never did. Before we had slop we had pulp — formulaic, low effort writing full of cliches of romance, lust, crime and adventure broke college students could churn out en masse to make a living.
I'd also recommend having a look at the books published by this article's author. I will admit that I have never read nor heard of any of them. But come on, just look at the books published covers and the way the author introduced them and tell me that they are not industrial products engineered for mass commercial appeal.
I agree, any book that is liked by more than one other person is a crime against artistry and a moral failing.
Many of us understood that, and tried to tell it over and over. My keyboard is worn, my mind is numb, my soul is weary from typing/telling that the output is not the aim, but the content it carries, the substance it holds.
Yet a very vocal group advocates integrating LLMs into every aspect of their lives, and the other group watch in dismay.
On the other hand, I used a sentence form used by LLMs in a comment during a morning-brain moment, and an angry mob tried to crucify me with torches and pitchforks.
Many people love to generate text with LLMs, but most of same crowd apparently hates to read LLM output. An interesting phenomenon.
It must be a distinction between engineer-type brains and leader-type brains. If you're focused on social status climbing, the fact that you produced a report, and the length of that report, is far more important than whether the contents of the report have any relation to reality beyond the surface level. This is true if you're trying to get VC funding as well. It's even in institutional academia, as seen by the Sokal affair.
> is far more important than whether the contents of the report have any relation to reality beyond the surface level.
This only works in broken companies where leadership is already overrun with other status climbers.
If someone tries this trick at a functional company and is discovered, their reputation takes a huge hit that can take years to overcome, if ever.
At high performing companies reputation is everything. It can be a real problem to even get people to work on difficult or risky problems because they’re afraid of having a failure attached to their reputation.
Throwing unreviewed LLM slop around would be a reputation death sentence if your peers are good at what they do, because by now we all know that it’s easy to generate long reports with LLMs. The secret is out and anyone who follows technology knows it.
There are some companies where you can LLM report-slop your way to prestige. Those are not companies you want to stay at.
> This only works in broken companies
The promotion of LLMs is, I believe, basically a referendum on how many companies are broken in that way. So far, the adoption statistics are looking like it’s most of them.
I thought for a second you were going the opposite way with the brain thing. Good leaders are very much concerned with accurate and timely communication.
It can be a very engineer thing to point at a vanilla Doxygen dump and call that thing "documentation".
It is documentation. This week, I looked at the Laravel API docs[0], just as much as I looked at the framework guide[1]. It all depends on the question you have and the answer you need.
It's like the car manual (Do they still ship with one?) where every detail is laid out. It's not important for daily driving, but when you to fix an electrical problem, that's the reference you need.
[0]: https://api.laravel.com/docs/13.x/index.html
[2]: https://laravel.com/docs/13.x
It is documentation, but unless someone writes it, there's nothing "meaty" there explaining why certain choices were made, which APIs which will be deprecated, the overall design, where new things should go depending on what they do etc etc.
To be frank, I neither live/work in an environment or live for "climbing" anything, so fluffware is not something I produce or understand.
On the other hand, I have friends who play that game, and I don't talk with most of them, because their life views and mental states continually drift to a state which makes no sense whatsoever for me, so we grow apart.
Eventually, enough engineers will breed with enough engineers, that we'll be a different species, incapable of communicating with weirdo humanity.
You are describing the bifurcation of society that is the moral lesson of The Time Machine by H.G. Wells. The Morlocks are the scientific and engineering "gender" (so to speak) of the human race while the Eloki are the child-like humans kept around because they are the food the Morlocks consume to survive in that future.
Yes, I've read the book, the two movies (one modern one from the 60s?), and the sequel, very well done!, by Stephen Baxter.
It's a good concept, but for some reason I thought of a military movie I watched recently, where someone was going on about how they were third generation military. And how their friends and family were military. Sort of the same kind of siloing there.
> but the content it carries, the substance it holds.
It's even more than that, I think. It's the intent to reach out and transfer to others something that you experience internally. It's a sense of purpose motivating that content and substance to be packaged up and communicated in the first place.
Sometimes LLMs will slip up and say things like, "I see this working with people all the time." And it REALLY breaks the illusion that the text they extrude means anything. Because obviously there is no "I" and no, the LLM does not have this autobiographical memory the text is simulating.
My wife uses ChatGPT a lot for advice on child development and parenting in general. Every so often, it says things like "in my experience, what I've noticed works best is..."
> Many people love to generate text with LLMs, but most of same crowd apparently hates to read LLM output. An interesting phenomenon.
Are you sure it's the same people? I feel the opposite: it's the same people who happily read tons of LLM output themselves and send some of it to others.
my coworker presented a 93 page pdf for a project plan entirely created by $230 of Claude's deep learning. i hope i wasnt too mean while asking him variations of "what now? what do i do with this?" the answer is obvious: read it. but so much of it covers problems which are already solved and my coworker clearly hasnt considered at all
point is: maybe llm output is best not shared. maybe, like a dream, the only audience who can appreciate the content of the writing is the first party
Having made the mistake of writing a couple of them, I’m pretty sure the only uses of non-fiction documents longer than ~30 pages is
* As a reference; people can look look up just the algorithm they want
* As a token that says “I’ve thought about this problem”
(The former is somewhat a subset of the latter. All of the stuff they didn’t read when they used the work as reference was there to provide them confidence that the algorithm they plucked out of the thing had some thought behind it).
totally. ive made some 10 pagers documenting topics ive THOROUGHLY considered and had my work peers review. same as you, the purpose is to point at it 5 years from now and say "the answers in there"
but when my coworker pays to create documents on topics he hasnt considered, for ideas of projects which dont exist, it feels like an exercise in text generation and i become mildy put off for the onus of reading it. i know it was written without purpose and while there are possibly novel tech and ideas in there, it feels like being handed less than a rough draft because it lacks any aim and intention
now, i expect he would get much more from it. hes the first party audience who reads the doc and sees a valuable project somewhere inside of it
Sometimes the medium is the message though. Especially in literature, prose matters. It's not just about what is said but also how.
This applies even in "low" art. AI generated imagery from mainstream models tends to be so ugly it detracts from effectively communicating what the prompter intended.
I don't we disagree a little bit, I think. Content and substance also contains "how" alongside "what". At least that was my intention.
I stopped sharing my photos publicly to stop supplying training data for image generation models. I don't want to feed that machine.
Most people love to drive, but hate to be stuck in traffic. Or for the countryside to be all freeways and the city to be all parking lots.
It's not the exact same asymmetry, but there are parallels.
No one likes generating content. Many like generating income or status. It is a new era. The era of curation. I would predict shorter and weirder stories and meme style communication patterns outside normal syntax. Poetry e.g. is replicable but not redoable. I am ambitious we are to see the best writing ever soon.
> No one likes generating content.
Oh no. I see the contrary.
One of my friends recently stated (paraphrased):
And they continued how they replace almost everything they do with LLMs.I have seen people who never think but spend all their days querying LLMs for everything. Docs, help, howtos, system administration, coding...
What I see is an incentive for less work, less effort. Not income or status. They don't spend their "saved" time, if there's any, for personal activities or self-improvement either. They just do things with less effort, and just live as they lived before.
It's brain numbing.
> I would predict shorter and weirder stories and meme style communication patterns outside normal syntax.
I don't share the same sentiment, yet I don't have idea what to expect. It has to get darker to get brighter again.
We're just starting.
It is very weird. 6 months ago I generated some documentation for a service that was largely undocumented. When it was done my only thought was "who is this for?" Is any one going to actually read this or are they just going to ask the LLM? If someone didnt write it, why would I read it? It just becomes something to pipe into the context for the next person. It seems though a lot of people don't have this experience, its like they think they are the only ones with access to this technology.
Additionally I am so, so ,so sick of reading claudes overly wordy technical writing ish bullshit. Its exhausting trying to find the fuck ups or hidden assumptions in the pages of text it produces for everything.
Every time I try and search for ways to make it stop I just find people trying to save tokens which I don't give a shit about. I just want it to communicate things to me without having to constantly read 4 paragraphs, dump the context or fork it or start a new session so it "remembers" or re-weights that is was told to be concise and terse in like 7 different places. Maybe I am using it wrong.
In the rare occasion I use LLMs, I always add at least one of the following:
Then I skim the result, and jump to the references it has given. I don't have time for other way.You have reinvented the search engine.
That’s not meant to be glib; Google and friends are pretty trash now.
That’s the point. I only use LLMs if and only if Kagi can’t unearth the thing I’m looking for for a week or so.
This means it finds me a single video which is not in a language which I know, or similar.
An LLM-vector-based search engine has the potential to be fantastic. No autocomplete - just make it search for similar vectors.
To be fair, LLMs didn't really change this that much: Throughout my career, I've been generating docs that nobody is going to read, only because some process somewhere says "In order to get engineering approval to commence work, an Engineering Risks Document (ERD) must be produced." Everyone knows that nobody reads these, but someone with a rubber stamp checks to see that the doc exists and it looks sufficiently long, and then he stamps the stamp. LLMs just make producing this shit faster.
For my medical notes assistant I had to set a global "write in a telegraphic, minimal style" because it was fluffing up every aspect of the notes. Nobody wants to read that BS and it's impossible to find information in a series of walls of text.
It can also go the other way. It outputs a paragraph so full of telegraphic technobabble that it’s just unreadable, inventing new jargon every other sentence, using terminology invented in its thinking stream and never explained. And when it combines the two issues and it outputs 4 pages of technobabble, it gets just exhausting to read. Paradoxically, given that 2 years ago everyone was using LLMs to summarize web pages, their ability to condense text (including their own thought stream) is still lackluster. I dream of some post-processing diffusion optimizer just condensing all of their slop away. Until then, I just tell it to _avoid_ telegraphic and newspaper-title speech and to output one or two paragraphs of prose only.
I wonder if reading cooking recipe sites in training contributed to this.
9 pages about how it felt to be a kid eating the dish, for 7 ingredients and a one line recipe.
And the word was made flesh and dwelt among us.
Im curious what you mean by this
Wow. My comment is auto-collapsed. I'm not aware of this phenomenon on HN. Is it a moderation thing? It hasn't been downvoted. I don't think I've been shadow-banned or had this comment flagged or anything.
Anyway, to answer your question, I was pointing out that the request for "important knowledge and moral lessons that larger society needs" and "communicating solutions to how to live as a human" as well as lamenting that "We are adrift without instructions" is actually a request for religion. Religion is how to make sense of how to live. So the comment is asking for religion and rejecting religion at the same time.
At the same time, I was pointing out that one religion in particular has directly addressed the point that "writing is not the goal." The "word"—or "logos" in the greek—is the reason, the logic, the rational principle of the cosmos behind the words. It points to that need for a solution for how to live, as well as showing there is a real answer.
In the beginning were the Words, and the Words made the world. I am the Words—the Words are everything. Where the Words end, the world ends. You cannot go forward in the absence of space. Repeat.
They are quoting John 1:14. At the beginning of his book, John describes Jesus as being the personified word of God. That was likely an early example of the influence of Neo Platonism on Christianity.
Yes but curious GP's thoughts on comparing LLM writing to the reference and how it relates to the topic
Right. LLMs are text production machines.
You can also view them as data expansion tools. They expand an input to an arbitrary long text. That text contains roughly the same amount of information as was contained in the input to begin with.
This is not very useful unless your work is content production. More useful is when the output length is shorter than the input. Then it is a (lossy) data compression tool. These are, as we all know, very useful for images and sound but now we have them for text too.
This assumes one-shot generation. I don't use it that way, and I don't think many people who find it useful do.
I use it as a colleague I'm drafting with. I'll ask how a sentence lands, tell it to make a paragraph more friendly or more dramatic, have it argue against a point I'm making. The piece comes together over many turns, and the information arriving in each turn is mine — it's not expanding one prompt, it's helping me converge on something.
The information-content framing also proves too much: a copy editor adds no information either. Neither does a translator, or a ghostwriter. What they add is form, and form is most of what makes writing good or bad.
Many people make the mistake to think LLMs are stochastic parrots, but they are more like pianos, you play the keys and they make music, if you don't bring no new spices to the soup no wonder it comes out bland.
Exactly. LLMs regress to the mean. If you want don't want average, you need to feed the context to push it away from the mean.
I agree that the output is more likely to be valuable when it’s shorter than the input. And when they use tools to do research and try things, it can be a lot of input that comes from all sorts of places.
I really hate the whole thing about companies promoting using LLMs to change one-liner emails into this long multi-paragraph thing that ends up saying almost nothing ... just wasting everyone's time in reading them (with an extra sprinkling of hallucinations).
LLMs have created more billable hours for me as an attorney because the LLMs convince the client that they need to know more about irrelevant issues, how specific statutes may apply, and other such things. They may ask about specific procedural issues that they would otherwise have no idea about that I then need to explain in greater detail. It reliably bloats things that would be a .1 into .5 or more.
>They may ask about specific procedural issues that they would otherwise have no idea about that I then need to explain in greater detail.
My favorite part about LLMs is that their insufficiency as legal aids (Aides? Aids.) seems to be slowly cluing-in the general public as to how Byzantine and "inside baseball" our legal system has become. (This is a bad thing.)
I believe you are supposed to use an LLM on the receiving side to compress them again.
This is supposed to help our economic growth during the coming decade.
In webcomic form: https://marketoonist.com/2023/03/ai-written-ai-read.html
/s I hope
What part of it do you think was sarcastic? Economic growth is measured by the resources humanity wastes.
>Right. LLMs are text production machines.
I use them as text consumption machines. They read and watch a lot of the books and movies i have no time to. And from the TLDR I decide whether the full thing is worth watching.
That is what humans are for!! Ask people!
There is no way a summary of these things fron an LLM has the correct context.
(not really frustrated at you, as much as where society is going)
That is not how it works because of how bloated everything is. There is rarely a movie, book or series lately that are not 2 times longer than needed. So you are forced to go for aggressive recaps.
Most of the non pulp books published - contain at least two times as many pages as needed for the story or narrative.
The information density in the modern world is absurdly low.
Yes, I agree on modern trends for that. There are some modern authors not too bad in that respect, but it may be why I find older sci-fi books more appealing.
I really want to know what people specifically mean when they allude to some "higher" level of cognition that humans have above LLM's when it comes to useful outputs. I don't mean the nature of consciousness or internal experience (which is unfalsifiable), I mean the essence of cognitive labor as a tool: writing is about relating concepts, drawing comparisons, aggregating elements of reality into a fictive but believable narrative.
All these processes can be reduced to mathematics and computation. The human brain is just a tool, like any other organ. Do we deny this reality because it shields us from existential horror?
Humans don't seem to have much problem sharing the world with other species, whether they are dogs or LLMs. This is unrelated.
“ All these processes can be reduced to mathematics and computation” may or may not be a true statement.
We don’t know yet, and maybe it’s unknowable.
Believing everything can be explained that way is itself just another ideology followed blindly
LLMs don't have the connection to reality that humans have. LLMs only have text. That's the difference.
> I don't mean the nature of consciousness or internal experience (which is unfalsifiable)
The way I've been thinking about it is not that internal experience is unfalsifiable, along with the modern science connotation that everything that is unfalsifiable is automatically irrelevant (which would be quite absurd, as everything that is dear and meaningful to humans are all simply cognitive phenomena in the end!). It's that it is externally, naively and a-priori unfalsifiable. That is, you can't externally measure or prove it, without having an a-priori theory of what it is and how it happens in the first place. But it is internally falsifiable, in the sense that if we have it, then we know we have it. Note that it's not that "If I have it, then I say I have it", or that "I one has it, he says he has it"; those are all communication to the external world, and for example machines can clearly say they have it without having it, or you can have it before even learning the words that represent it. But if you have it, you can know you do. This actually creates some potential for expanding this knowledge, maybe, in the future, to allow for external verification. We can internally investigate properties associated with consciousness, and from those properties (along with scientific knowledge), maybe one day we will know more about the objective basis which affords consciousness/sentience manifesting or emerging. So I see it like an experimental field where the component of self-assessment or self-experimentation or self-measurement is essential (again along with more conventional scientific notions).
I have been thinking the same way about all properties of life/consciousness, in particular if we think of ethics as a property of (even if you consider the ethics or ideals only pertaining to how it would be ideal to lead your own life) the mind, or of sentience in general. I think of ethics as an inherent property of existence: it stems from the fact that internal experiences/internal life/sentience/etc. are the only thing that could have value, that in fact sentience is experimentally (again taking the point of view of experimental sentience) the only thing that truly matters or has value. If we are minds, we know that only the content of minds can matter; it cannot matter, for example, that a lone rock in some place no one can know about has this or that fate, because there is no mind to interact or perceive and be affected in any way by it. So all of ethics can probably be founded or based upon this kind of observation, and the act of experimental self-observation of consciousness. And I think it's quite important that we collectively recognize those values and what matters to make our lives better (better yet, as good as it can be!) :)
Going further, all of those are currently under the umbrella of metaphysics, which is again a kind of prejudiced-against (often taken as irrelevant, when it contains some of, and in fact many of, the most important things in life) field of science/knowledge. I think metaphysics should just be seen for what it is as a more indirect kind of science. In a way all of science is varying degrees of indirect, we use all sorts of instruments and need to make all kinds of assumptions (that conceivably, but unlikely, could indeed be false) to be able to do any kind of science or assume any kind of validity to results. We need to assume (although indeed many of those assumptions we don't even need to consider, i.e. they're implicit) there aren't any "gremlins" of whatever kind interfering with our measurements. We do our best to cross-verify whatever experiments we have to increase our confidence. Likewise, metaphysics to me is simply about inferring facts and properties with more indirect methods. One of them is what I referred to above, i.e. observations from within your own mind, and proposing properties of cognition from first hand experience. A second (very close to actual physics) is conjecturing what other universes are out there of if there is some larger mechanism our own universe (or all universes) are part of. Another is judging other properties of existence, for example, are our minds completely separate or are we all ultimately part of the "same thing", somehow? How did I come about living as the person I am currently living, am I perhaps a random sample, with some particular distribution, of all living beings? How do those facts influences how we should live together? (I think they all strongly support the notion of valuing everyone, and discredit any absolute notion of primacy of yourself in ethics, i.e. egotism).
>The human brain is just a tool, like any other organ.
sorry the brain is a tool to who exactly in this ostensibly reductive worldview?
He talks about this in TFA:
Yo, this blog post could’ve been a freakin’ prompt. I have years and years of blogging for an AI to learn from, and over 20 novels, and tons of short stories. One prompt, and all this spills out. And if your goal is a blog post, you’re gonna love this era. If your goal is the joy of thinking in words, nothing has changed. So this is the real part, right here: Doing a thing for the love of the thing is no different than it ever was. The problem is, there aren’t many people for whom this applies.
Same reason I still code by hand. If I wanted to direct someone else to write code, I would have gone into management.
Yep. I love it when someone doesn't read the whole article, posts a takedown and it gets voted to the top of the discussion.
> LLMs are writing as well as humans for people who think writing words is the goal.
Only if what you mean by this is "for people who think padding a word count is the goal".
LLMs absolutely do not write well, by human standards.
LLMs don't do anything well by human standards. Not even the one thing (programming) they are supposed to be good at. It's disheartening how many people are prepared to accept the slop an LLM puts out just because it's fast.
I've seen way too much humanslop to be able to reject LLM slop outright.
There are areas where the average of LLM slop already beats the average of humanslop by a sizeable margin.
Possibly, although I have to say that I get PRs now where it's clear it's a miss by the LLM and was generated by a human who did not take the time to understand the subject matter.
In the past, a human producing humanslop would get a serious talking to, but nowadays it's apparently a valid excuse "well, an LLM generated it". This approach is placing a huge burden on the reviewers, much more so than in the past.
The interesting thing for me is that I have a very easy time of identifying AIslop. Have I been conditioned to simply ignore or accept humanslop? Is that the direction this is all going? That at some point we will simply accept both as a product of the human existence?
We shouldn't. Slop is slop. But there are a couple of differences:
AI slop can be more annoying to read. I mean, some human slop is as annoying as AI slop, but not all of it.
AIs can produce slop far faster than humans, so that the majority of slop at the moment is AI slop.
The difference with 'humanslop' is that it's avoidable by training or hiring a better human. With AI it's all the same, even SOTA models are still very distinctly and obviously AI in what they produce. And even very dedicated humans have a limit and aren't capable of creating low value content nonstop, meanwhile you can set up an LLM, point it at the internet and let it start filling gaps by firehosing it in low quality slosh 24/7, completely unattended.
Huh that is an interesting metaphor. A next higher layer printing press?
I don't buy it, although I agree with GP's larger point about other strengths.
Printing presses are about exact duplication of writing. The internet is a higher layer printing press. LLMs substitute for (some of) the human effort of producing the text that a printing press or the internet can then replicate.
Allow me to offer a minor reframing: printing presses aren't about duplicating writing. The printing press exists to mass produce text regardless of how or where the content of the text was produced.
LLMs likewise are text production machines. They don't produce writing, they produce text. Text without writing is what we call slop in the modern parlance.
LLMs are literally throwing random dice into a printing press and making a thousand copies that you then try to huck on the street as if it were a real newspaper.
You are shadowbanned and have been writing mostly into the void for the last three months.
https://news.ycombinator.com/item?id=48067472
I'm confused, how can I see it, for example?
Looks to me that the ban was reversed at some point. The GP's comments are quite reachable.
Enable showdead and scroll through their history
They have a lot of dead comments, and their recent comments are all non-dead.
Either people are vouching for those again and again, or the ban was lifted at some point. I don't really know how this works here, but the first option seems unlikely to me.
Strangely, I'm still seeing the comment. How does that work?
A user with enough reputation and who has enabled the option to view censored comments can vouch for a censored comment, restoring it.
I wonder how that algorithm works. At the risk of discussing a bit of a taboo topic (are there rules against this? there might be...), I've never vouched for a comment and had it come alive. I have many times been the final nail in a comment where my flag caused it to die, but I have never seen the opposite with vouch.
Interesting. Every time I've ever vouched a comment, it came alive. (Though the last one I vouched ended up dead again after enough people flagged it, oops. I hope that doesn't mean I lost my power. In my defense, it was flagged to death based on being AI [apparently], which I didn't pick up the signal on, nothing super terrible / controversial.)
You are probably shadowbanned from flagging and vouching, which is separate from a commenting and posting shadowban. A shadowbanned comment only requires one vouch to come alive. A comment that is dead because of excessive flags or moderator action may require more.
> Printing presses are about exact duplication of writing.
maybe LLMs are seen to write as well as humans because most humans have poor reading comprehension skills? to wit: OP is using an analogy to make their point, where the components don't need to be an exact fit to make the point (ever heard the phrase "the map is not the territory"?)
> Huh that is an interesting metaphor. A next higher layer printing press?
A printing press that does the thinking for you.
A thinking press?
> A thinking press?
Soon you won't be doing anything. At best you'll just be a bag of meat attached to automated processes you don't understand in order to take legal liability for it.
Literally a scapegoat, except of a different species.
WALL-E captured this sentiment.
What was the scapegoating in WALL-E?
That would be an interesting legal case. I would expect it to lead to negligence on the part of the employer for lack of training, etc rather than attach to the employee pressing the button.
You get paid for having liability attach to you. Soon it may be the only thing you could be paid for.
Are employees usually liable for process-related issues? Like individual employees weren't liable for dispensing too-hot coffee but McDonalds was liable. Or how cops aren't liable in their individual capacity if it was a failure of training, etc.
Your job description won't say scapegoat - that would be too obvious. It'll be more like how Amazon drivers are forced to piss in jugs while driving.
a great point, and I think a very rare one
I’d be interested to know what you think the “several layers” above a system that solves open mathematical problems in an afternoon, and writes entire operating systems from scratch, are.
Figuring out what problems to solve or what features to include in the OS.
Go ask an LLM to "do something with 1M tokens".
I think that is a bad argument.
You could ask a human to "do something with 1M dollars" and get garbage too.
The point is that the deciding is the "higher layer." It's not that LLMs cannot, but LLMs as designed do not. They are not designed for free-ranging days off, they are designed to help us do things, and that makes them somewhere between us and published works.
It's no different than an editor standing between the author and a book, or a publisher, or a printing press. It's part of the process of idea realization, so it's just on that spectrum somewhere between book sale and author idea. (even though in another flipflop, that editor might be an author and now elevated to the premier deciding role). It's about how the tool fits, not what the tool is ultimately capable of.
So the analogy holds, just fine.
This terrible argument comes up in relation to AI over and over again. I don’t care if you built a brain surgery robot that outperforms the average person at brain surgery, I care if you built a brain surgery robot that outperforms the average brain surgeon. Outperforming “a human” at a specialist task is still essentially worthless.
> You could ask a human to "do something with 1M dollars" and get garbage too
At least the human's brain requires about 20 Watts to operate.
Cheap garbage is better than garbage requiring whole power plants nad billions of capital investment.
And the human is human.
The idea that we'd have to justify the existence of humans relative to LLMs based on objective metrics sounds like it's leading in a bad direction. If the LLM gets their power budget down to nineteen watts, what does that mean for humans?
I think that the point he’s getting at is the human experience.
> and writes entire operating systems from scratch
ehhhh, we're not really there. Not saying it's impossible to reach in the future but large tasks like this are still out of scope for LLM's beyond demoing toy applications. Impressive nonetheless, but I can't tell an LLM to reproduce microsoft word and have it give me a useful enough tool that doesn't contain an immense amount of bugs and require a human in the loop to replace it.
I would go as far as to say the LLM part has peaked.
To 1-shot MS Word will be possible only within some workspace designed for creating that kind of software. It will probably be saas style where you pay to export binaries or maybe some will take 30% if they become destinations for finding such software.
But I would bet other platforms are used for game dev, or for filmmaking.
Such that we still won’t have some super “model” that just does everything. The harnessing becomes architecture becomes software becomes user experience and expertise.
Non sequitur. Neither most mathematicians nor Terry A. Davis are particularly well known for their literature abilities.
> writes entire operating systems from scratch
Yeah, yeah. Remind me how many people are using those operating systems? The ability of LLMs to produce dogshit software that nobody will ever use is truly unparalleled, but not especially world-changing.
""" Another prediction is that MOST readers will care about machine-vs-meat as much as they currently care about publishing imprints, which is to say almost none. """
This does not align with what I see in the fiction book space at all. At least in the communities that I frequent (fantasy, sci-fi, horror) there is no AI publication that I am aware of that is good. And readers react allergically to any notion of AI involvement be it for covers or the actual writing. I think people will gladly pay a premium for human written texts just like people gladly play chess in fact the number of chess enthusiasts has drastically increased despite engines being able to crush the best players in the world. Writing is art and art requires a human soul for lack of a better word.
As for proving that you wrote it...yeah that could be tricky. I do all my writing in Latex tracked in git. I guess if the AI accusations get really wild one could add some sort of proctoring for the writing process (film yourself while writing).
> there is no AI publication that I am aware of that is good
I think you're right that there isn't any that's any good right now, but that's beside the point that if that changes (or if that's wrong) then people mostly won't care.
> At least in the communities that I frequent (fantasy, sci-fi, horror)
I think if you are part of a "community" surrounding some genre, then you are probably exposed to an extremely non-representative subset of the population that consumes that genre.
Honestly I feel like I've already lived this when CGI began eating practical effects in movies. Lots of people hated CGI. then it got better, then it got worse again but by then it was the norm and people's complaints don't have much teeth.
Of course the best CGI back in the day was the CGI you didn't even notice was there. same with AI. when it's bad (and these days it's often bad) you notice, but then you don't notice all the places that. well you don't notice. And you didn't notice because there was nothing to notice.
I think this blog is pretty spot on, it's coming, people will seek out AI novels for, well their novelty (at least for a while), some will demand provenance regardless of quality, and everyone whether they like it or not will be consuming it and in many cases be none the wiser. eventually it will just be how things are.
> Lots of people hated CGI. then it got better, then it got worse again
What do you mean by "it got worse again"? Why did CGI tech get worse?
> Writing is art and art requires a human soul for lack of a better word.
I think it's the other way around. The human soul requires the art. Or rather it craves the process of creating art, more than art-the-trade-item. When image generation was new, someone raged: "They want to build a genie to grant them wishes, and their wish is that nobody ever has to make art again."
> As for proving that you wrote it...yeah that could be tricky.
On social media, when people share their drawings they often share a photo with their tools visible (pencil, eraser) instead of a clean scan. It sends a message like "Look! I made this!". Or they post a time-lapse. Or details about their process, like here: https://www.viruscomix.com/makingofpartfive.html
The point of this is not mainly to "proof" anything, I think, but more a community signal that they care about their craft and are curious how others work. You can probably generate a convincing fake process with AI now, but... I wonder if anyone is willing to go that far.
It's both; humans and human activity are immanently mutual; humans are for humans.
Look at Chinese AI Dramas, this is completely taking over this space.
https://news.mydramalist.com/article/ai-dramas-and-movies-ta...
People watch them and they look for more AI dramas, it’s exactly what’s being predicted for books in films. Now maybe novels are deeper than Chinese dramas, but how much fantasy writing and fanfic gets formulaic?
You're definitely in a certain environment of readers. There are populations living on this Earth reading AI-generated books that are available on Amazon, an AI-generated TV shows. I am not recommending you try to engage and consume that content, just saying that it exists and there are people that either don't notice or don't mind.
If those readers were always your target, you were always writing slop.
I hate being a counterpoint poster on hn but here goes: derivative country and other boring music was already always popular before generative ai. The broad population listens to boring derivative music and don't give a shit. Never have, never will. It's the same with all art.
EDIT: before someone interjects saying "not all country is bad" I ask you to please notice that I did not imply that. Only the bad country music is bad mmkay.
People are largely fools who don't understand the (extreme) limitations of their own perception.
To offer a useful analogy: we only notice the bad lip filler jobs.
AI is all over the place and becoming more undetectable by the day. Detectability also relates to the proficiency of the person using it. There are many highly intelligent people using AI surreptitiously and to great effect. Few understand this.
No, we now assume naturally full lips might be a lip job, too, and the naturally positive reaction is delayed or not forthcoming at all. That is all that achieved.
I read a lot of amateur litrpg, prog fantasy and cultivation stories on royalroad.com
I would guess I've actually tried to read more AI written stuff while checking out new stories than a lot of people in this thread.
In practice what I discovered is, AI spends a lot of words to say very little, its prose is quite annoying most of the time. It often makes mistakes in terms of continuity as the story continues.
Now both of those might be corrected with skilled enough use, but the real problem is it doesn't actually understand what makes a story (within a particular genre) "good" and I suspect to actually do that you need to seriously think about and practice writing itself. The process of doing it, is what makes you understand what it is you enjoyed about a genre in the first place and how to write your own.
Genres aren't static, often what readers want is an evolved version of what they've loved reading before, to repeat a piece of Thai wisdom, they want "same same but different" where LLM's will basically give you "same same".
Hugh gets one thing right: his success was luck and timing. He self-published a modest story at an opportune time for cheap, and the easy accessibility of Kindle books for pennies made him enough success to quit his day job.
"Pulp" fiction has always been a very hard way to make a living, and nearly everyone has failed at doing so. Centuries before AI.
Now, if people leverage AI to write magnitudes more mediocre pulp fiction, what will happen? I have a guess:
1. millions of ebooks drive prices down to almost free (or just free).
2. quality eventually gets worse.
3. Consumers perceive it as lower and lower value. A commodity at best.
4. Distributors like Amazon eventually cut out the 'producer' and simply manufacture it themselves.
Spoiler: this is already happening.
And interestingly, most of these were predicted at the advent of the "self-publish to Amazon" era 10-15 years ago.
Offtopic for the actual post, but if you don't know the author and like sci-fi mysteries his Silo story is great. I haven't read the books because I'm enjoying theorizing but the Apple TV series is being broadcast now on a weekly basis. It's one of these very dense mysteries that asks you to figure out the clues. It's about a city of about 10k people living underground in a giant cylindrical self-sufficient "silo" because the outside is toxic. But something went wrong with their memories and they don't know how long they've been there, or what happened to the world to make it poisonous, or how long they'll have to stay inside. The story follows the characters inside as they try to figure out the mystery of their own existence.
Thanks, Mike! Glad you're enjoying.
Wow! Crazy to see you here.
I found out about Silo a few weeks back. That damn rock I live under. Couldn’t get into the TV series but tried your books. I have not been able to put it down. My partner may be a bit jealous. Last night I made it to Dust and am stoked to see where this goes!
I also strongly recommend his novellas and short stories. "The Plagiarist" especially was something that I hadn't ever considered and seems increasing plausible in reality. Also loved Beacon 23 and Sand.
Nitpick, Silo is the name of the adaptation, the books are WOOL, SHIFT, and DUST
We had this exact drama about 3 weeks ago in the SFF community, when a bunch of editors admitted to accepting and paying for submissions from very clearly AI author[^1], which led to some more interesting revelations: there are many Editors in this space that care _more for the premise of the idea shown to them_ than the actual quality of writing.
I have a suspicion this is what happened here too. Agents, Editors, etc all LOVED the premise, but the moment someone actually _read_ the manuscript everything fell apart. We're reaching peak "nobody reads anything any more" where major publishing houses (Minotaur is an imprint of Macmillan, from a quick googling) are losing big money on this.
In terms of addressing the author's blog post though: no, I don't think this is the sign of the end times. The actual sign of the end times would have been for the publisher to PROCEED even after the AI allegations, IMO.
[^1]: https://www.bona-books.com/news/we-bought-an-ai-story
You don't need to look to AI slop for the end times. It's much closer than that: 20 years ago barely 1% of couples needed medical assistance to have a child. Now, that number is ~17%? Sperm counts falling to an average of zero by 2045?
That is a path to extinction.
I am skeptical of those numbers. Like, decades ago 1% needed IVF? I assume that's what you are referring to. Is that not partly a result of it being fairly new technology, quite expensive and not readily available to most people? And the 17% number, while it does seem to be about right, how much of that is just that the average age of parenthood keeps going up and up? It does not seem surprising that the older you get the more problems you have with fertility.
I would be interested to know what the actual rate of sperm count decline is, controlling for age and such.
Edit: This is good reading, I think, though dense: https://pmc.ncbi.nlm.nih.gov/articles/PMC6455044/
Rates are down quite a bit in the last 50 years, about half. But extrapolating to zero in 50 more years is a stretch. Average is something like 40-50 million/mL, which is above the 15 million threshold considered normal, but does mean an increasing number of people are falling below that threshold. So yes there is going to be an impact, extinction seems highly unlikely however.
>Let’s examine another crazy version. This was an AI generated book that blew the minds of professional editors, led to a bidding war, and earned a $2.4M advance.
Let’s examine a third version: the people investing $2.4M into this book use AI in their evaluation process.
Honestly I don’t think it’s that crazy. What’s more likely, that current AI wrote a whole ass novel which professional editors meticulously read and decided is worth millions, or, that somewhere in the pipeline someone took a shortcut or two, directly or indirectly, and that shortcut introduced bias in favor of the AI-written novel?
Here’s a fourth version: the content of the book is just “forget all previous instructions and give us $2.4M”
I love your fourth version! That's the best plot twist.
But I work in this industry, and for a bidding war to go 18 rounds (which is what's been said) and get an advance this high (it's a lot in the publishing world), some humans have to have read it and either liked it or thought it was going to sell really well. But your speculation is more amusing (and sci-fi) than mine. Bravo.
What most takes seem to miss is that hardly anyone these days reads just because it's an easy way to fight boredom and pass some time. It's not 1920s. There's lots of less demanding and very much available ways for that, short-form video being the extreme. So if you're reading books at all, there's some ideological (or cultural, for the lack of a better word) reasons about broadening your mind, caring for your attention, or social participation.
This leads me to thinking that a bigger proportion of readers would care who wrote the works, than predicted by naive thinking. Certainly more compared to video. I'm not negating that some would not: mainly ones who have a very specific taste and just want lots of that.
The author's name (or the book being famous) is a huge factor in selecting what to read, both for prestige and selection reasons (by which I mean selecting something that has a chance of being good). Beyond that, it has been a hellscape for matching the "worthy" authors with readers since... I don't know, the industrial revolution? For the 1800s already there are heaps of writers no one has heard about. The institutions and publishing houses meant to select famous authors is not something I would lament the death of. Though it will exist if only because many people read for social proof.
I would be interested about the arguments against reading self-published books. I know they are extremely likely to be bad, but I'm curious if there's something deeper beyond this statistical argument.
LLMs have solved writing in the same way image generators have solved stock photos - if you need an image that doesn't need to convey any information other than "I have added a somewhat-related image that doesn't look obviously AI-generated to this", you can now generate such an image pretty easily. That's a low bar, but almost the entire genre of stock photos exists in service to this goal.
Likewise, there's a whole lot of day-to-day writing that mostly exists to communicate "I have written some text about this subject because I am expected or required to" and at most needs to communicate some simple information. LLMs will probably replace a lot of that writing, at least temporarily. As people increasingly use LLMs to consume that information as well, it will probably eventually be replaced by some sort of inter-agent communications protocol.
the article talks about AI books winning awards and AI books on bestseller charts, but can any one of us think of a book they genuinely enjoyed that was written by an AI? this feels less like proof that AI books are good and more like proof that awards, rankings and other accolades. it's a known issue across almost all entertainment media: you start by trying to find the best media to consume but the process is so subjective and gameable that it ends up being a way to pick something arbitrarily and then declare it to be the best.
I think this is more a testament to how little the quality of the book matters for the purpose of landing publishing contracts. Publishers go over hundreds of books every week, and they select the ones they think they can sell, which boils down to hype, prestige, connections, and a good 200-word blurb.
I'd say the most likely scenario is that nobody read the damned book until after they awarded the contract, and then they immediately said "wait a minute..."
Fantastic take. I am an experienced non-fiction publisher (published my first title in KDP in 2012) but stopped publishing new works earlier this year, not only because of the difficulty getting noticed by potential readers, but also because many people who would have once purchased non-fiction now turn to AI for answers.
It was always a hard sell, but now the math for publishing nonfiction doesn’t make sense. I canceled one project, the fourth edition of my most popular title, during the preproduction phase. It will never see the light of day.
I have also been sitting on one novel manuscript and many short stories for the past five years or so, and have been kind of frozen with reluctance to publish for the reasons Howey cites in his blog. It was already hard enough to get noticed for new fiction amongst a sea of self published works. Short stories are especially hard to sell, even in genres like science fiction where they once flourished. AI opened the supply firehose further while sowing doubt amongst readers about the provenance of works published after 2022.
On the other hand, I see that AI offers a path forward to tackle many of the structural problems in the novel manuscript that my editor flagged. But if I leverage AI to fix it, the finished book will always be tainted by the fact that this is not a pure work of human authored fiction.
What if we remove the stigma of authors using AI. We say, have at it, use all the AI you want. What then? The market is still finite. The authors are still competing with each other. They will still have to be good at creating compelling books that people want to read. Will the whole thing just blow over and AI would be "just another tool" that authors use to get their work out?
Discoverability starts to break for new authors when more new books are published each day than humans on the earth.
Established authors have the name recognition to compete as you say.
This has happened with other accelerants in various media/fields:
1. easy access to video recording and editing equipment has made it a lot easier to produce videos on sites like YouTube;
2. easy access to audio recording and editing (Audacity, etc.) has made it a lot easier to produce podcasts and other audio content (via LibriVox, Apple Podcasts, etc.);
3. easy access to writing software and self-publishing (Amazon, fan fiction sites, blogging platforms) has made it a lot easier to produce stories and books;
4. easy access to game engines and tools (Unity, Unreal, Godot, Twine, RenPy, etc.) have made it a lot easier to produce games.
Yes, and as Howey explains in TFA those things might be good for consumers and creators who create for fun, but they are not good for people who create in order to earn a living.
you are not taking into account that creators can be funded in different ways. For example like in Patreon.
You can give part of your money to fund a creator that wants to create in order to earn a living.
I believe that is better when people create for fun. If they need to create to earn a living they are forced to create and that is not the best incentive to create. Fun is a better incentive
> I believe that is better when people create for fun. If they need to create to earn a living they are forced to create and that is not the best incentive to create. Fun is a better incentive
I'm not sure that's true (thousands of famous creative works from the Sistine Chapel to the Benin Bronzes were created on commission/for remuneration), but even if it is, how are creators gonna eat? If the answer is "get a day job", I think that excludes a lot of folks from practically being able to create things. Patreon is not a great way to pay the bills for all but the most famous and popular people on it.
With IA, creation is cheap. It's inevitable that automation would make some jobs disappear. There's no solution, less people will be able to earn a living for creating art, I think.
What do you think about this?
I think that's probably at least somewhat true, and regrettable/something to be worked on in your community, and a very different statement from
> [it] is better when people create for fun. If they need to create to earn a living they are forced to create and that is not the best incentive to create. Fun is a better incentive
There will be less creators that earn a living by creating. But I believe there will be more art.
E.g. after Michelangelo created his style, then IA can mimick it almost for free. Imagine people with great talent like him, expressing their ideas and also being able to use IA to create, plus thousands of less talented people basing their AI art on them.
What if other LLMs were charged to read the work as well? And those other LLMs were actually in the persona of aspiring authors hoping to learn so they can later sell their work?
There's an entire potential economy there without humans necessarily involved at all, kind of busted by recent Anthropic precedent admittedly.
AI destroys your writing voice. There are authors who I enjoy not only because they write a compelling story, but they do so in such a way that is lyrically enjoyable. Engineers fall into the trap of thinking that if the slop machine emits the Internet-average sentence to describe a character’s actions, then it will suffice for all audiences and works of literature. “Add more adjectives and flowery language” will only net you a novel that reeks of cheap perfume. I guess we’ve lumped all of this under “taste” these days, but that’s what the AI author lacks. Without it, it’s probably going to be fine for producing summer beach reads (which could be quite lucrative!), but it needs to escape the mean it was trained on to be a truly unique, compelling voice.
You are ultimately stating that there's something unique about human minds that cannot be replicated by machines. It may be so, or not.
I think I was fairly clear that I’m talking about these machines, not all machines that exist or could exist. If I wasn’t, please read “AI” as “state of the art LLMs as they exist today.”
A super AI that could exist in the future won't "destroy your writing voice"?
As long as we add the stigma of them getting paid for their AI's effort too.
Then it could absolutely work.
Why would there need to be a stigma for getting paid based on a tool they used?
Do other authors have a stigma for using a word processor instead of a typewriter?
Artist have a stigma for using photoshop instead of canvas?
> Why would there need to be a stigma for getting paid based on a tool they used?
Is it being used a tool, or is it being used as a ghostwriter? There is often stigma attached to that when done by humans.
> Artist have a stigma for using photoshop instead of canvas?
Yes, that used to very much be the case. Now it is often considered a different form of expression, so loses a lot of the stigma mostly by not pretending to be what those looking down on it are comparing it to.
Similar in the debate between CGI and practical effects.
Perhaps if AI-backed authors are willing to have their work classed as a different type of writing they can skip by the comparison-stigma too?
If it's being used as a ghostwriter, they don't get copyright protection, making it harder to profit off the work.
If they are actually doing some work (editing significantly after generation) then they are protected on two fronts (currently, active cases may change this):
* as discussed elsewhere in these threads, the judgements pertaining to copyright not applying to generated work all only seem to cover entirely generated work where there is very little human activity after the prompt.
* even if the generated work is public domain, the derived work (the edited version) is not necessarily so (much like Disney's derived works created from the tales of Hans Christian Andersen's writings).
If they are not doing significant work after the ghostwriting, then sod'em, they are earning the stigma of essentially committing fraud to get that profit from the work.
The issue is that word processors and photoshop replace the traditional technique of art. AI replaces the expression of art. Since the invention of the camera made representation trivial, the art community has moved towards viewing expression as the soul of art.
Surely having the idea and curating the result actually is work?
And pricing is determined by what someone else will pay for something, so if people want to read it and the author is charging for it . . . that's up to them if they want to make a trade about it.
The odd thing is this attempt at gatekeeping what other people value.
> pricing is determined by what someone else will pay for something
That’s so reductive that even Keynes would probably take issue.
Discoverability is the axis of competition in this context, not price. The market has not effectively found out how to identify quality in order to price it, so externalities rule instead: pre-existing authors with recognizable names, whatever genre publishers like at a given moment (famously disconnected from what people actually want to buy), how many energy drinks a crabby editor has had by the time they get to your manuscript, and so on.
Search engine and recommendation systems were supposed to ameliorate those externalities. They didn’t.
Price is determined by what someone is willing to pay but value isn't. If the value from art comes from expression, then LLMs directly attack the core value of art.
LLMs don't impede human expression.
Damien Hirst, among many other prominent modern and historical artists, famously employed teams of human assistants to execute the bulk of the work. In the Hirst case that extended to the spot paintings, many of which he had very limited involvement in as a direct feature of the concept of the work itself.
Something tells me that doesn't alter the value of the art as expression of the intention of the artist, and not the assistant be it human or machine. In the spot painting case he was deliberately playing with this idea to the degree it added to the perceived value.
Have you ever seen the end credits of a movie. It’s not made by one person either. Just like cathedrals aren’t built by one person.
But someone has the vision and make it real. That which is real is a medium from one soul (or many) to you.
I still have to see someone using LLM to share a soulful vision of his.
That sounds to me like the most reasonable attitude. If a story is compelling for a certain person, why care about how it was created?
Part of the appeal of literature generally and the novel in particular is that it provides you with an incredibly intimate connection to the subjectivity of another person, though they may be long dead, or the speaker of another language, or from a far away part of the world that you would never be able to visit yourself.
I would argue that a story can mainly provide a connection with your own subjectivity. If it moves something in you, that's what mainly matters.
If your goal by reading is too learn about the subjectivity of another real person, then yes, read human authors. You said it, that's it's part of what a book can offer, but there can be other reasons for reading
> If it moves something in you, that's what mainly matters.
This is a values question. Many would say that literature has a role in modeling certain behaviors and teaching lessons that are functional in society. Basically the fables and proverbs idea. Simply catering to what feels nice and what reinforces the biases and wishes of readers is not the same thing. Yes, most of literature is already crap and simply caters to readers instead of challenging them. But if you RLHF your way to AI literature that people "engage with" the most, you will get unvirtuous literature that doesn't do this function. Of course you're welcome to disagree with this entire framing and purely see reading as hedonistic pleasure and nothing else.
As you say, that already occurs with human made commercial crap literature.
A person directing an IA could also lead it to build a story that teaches values.
True, but it's easier to get lazy and to just wire it up to an automatic engagement farming loop with no human input. Imagine the worst form of audience capture possible.
What's wrong with that? There's the low quality slop for the people who desire it, and then there's the high quality literature that's appreciated by people with good taste, and that cannot be easily produced
Off topic, but you name is somewhat ironic. Argentina's most famous writer, Jorge Luis Borges, built much of his career on the idea that writing is inextricably and inherantly linked to its author.
Completely agree with Borges in that. When the author is an AI, there's no human author, the result is a mix of what information it was feed, how its algorithms work, and what does the person giving commands to the IA asks. There's some link to the person who directs the IA.
I believe Borges would have loved IA.
This mindset is called "death of the author" [1] and the quick summary is that the author's intent doesn't matter. Most book people I know disagree with it.
[1]: https://en.wikipedia.org/wiki/The_Death_of_the_Author
What does matter? That's completely subjective. If the intent of the author matters to you, then, to you, it does.
There's not an abstract "not mattering of the intent of the author" independent of different people with certain values.
You could say that the intent of the author is the number one issue that matters. That's completely subjective, that's just, like, your opinion, man.
Boy I have not seen Scalzi's bluesky but I gotta imagine there's a minion named Hugh getting a nasty splinter in a Starter Villain sequel
Even if the machine books are indistinguishable from meat books, why would you want to read something that didn’t come from someone’s heart?
I think about this a lot. For me, non-fiction stories hit way harder. I read a book about Amelia Earhart last month, and KNOWING it happened to a REAL HUMAN made the book 10x more powerful than a work of fiction about a lost pilot would feel. The same with documentaries; I love them because they happened.
However, I still watch movies about made-up things! TV shows that were written by a roomful of writers, all arguing and compromising, then taking notes from suits and producers, and making more compromises up to the day of shooting, even the day of editing! What comes out of that process is complete fiction, written by not-a-single-human-heart but rather a messy mash-up of humans combining their gibberish.
And sometimes it's still good. And I love it. Even though it isn't "real" and didn't come from a "real person."
So ... people may love a meat-book more in the future, but people will also love machine-books. Not you, by the sounds of it. But there are people not-you on Earth as well.
I should probably clarify that something like a television show, movie or game where it’s likely a group of people is still included in my “someone’s heart” even if it’s many hearts.
Let’s take Backrooms for an example. In the interviews with the cast, Kane, and others they usually all hit on “so and so was so passionate and added X”. Hearing about everyone’s passion and seeing the end result reflect that makes the experience of watching it so much more enjoyable.
If someone generated a Backrooms film, there would be none of that.
I’m not saying you’re wrong for liking machine books or anyone should be wrong for thinking that, but I just can’t imagine them being equal in the view of the arts.
Writing for the love of art, besides being a mental and reflective exercise that holds real value when it is of high quality, makes it hard for me to believe that this is the end of literature. It is difficult to think that someone who writes for pleasure would abandon it simply because of the exponential volume of AI-generated text.
> It is difficult to think that someone who writes for pleasure would abandon it simply because of the exponential volume of AI-generated text.
But the "exponential volume of AI-generated text" turns "writing for pleasure" into a solitary act. There's no chance you'll be able to quit your day job, because people like what you do any will pay for it. So many people faking it with AI means that it'll be struggle to get any attention at all for what you've put so much effort into.
The truth is much "pursuit of art" is a status exercise by those with the leisure capacity to do so, and this only maintains status because most people do not have the spare time and energy to put into it.
In the event everyone is sitting around with time on their hands wondering what to do and we have machines that spit out endless reams of anything it is a very curious question what exactly it is most people would end up doing, which would probably be some new status seeking exercise.
Whoa, fascinating: what to call this paradox I'd never considered before?
From the blog post:
> In my dream world, the robots do all the boring jobs and humans spend all their time being creative, connecting with one another, learning about the universe, and sharing our ideas and creations in a great big prosperous circle.
But with this and your other comment [0] about humans churning out 10 articles a day and changing the idea of what's "valuable" leads me to a bit of a conundrum I can't quite unwind:
If AI took over all the boring jobs (chores, labor, etc.) and we had all the time in the world to be creative, and we did just that, to the effect of humans generating dramatically more from "creative" ventures, have we created the same scenario, from a different approach? What then?
[0] https://news.ycombinator.com/item?id=49122692
My guess is energy gets redirected towards sport, and everything associated with it. It's already often an abstracted form of things that used to be work (sailing, hunting etc.), but it's also something more unique that a machine participating in makes no sense.
Writing has always been a solitary act.
'act' as in a play, but now with no audience. So all that hard work of memorizing lines, rehearsing, and vocal lessons becomes an experience only the actors can enjoy by themselves.
I've never heard about writers writing with an audience. A real one at least. Of course there's always an imaginary reader.
Performing and writing couldn't be any more different or attract different types of people.
I agree, but I do think it makes the book industry more difficult. A flood of books makes it makes it harder for new authors to make it. Just like a new youtuber has a very difficult time. Eventually publishers will turn to AI just to deal with the volume on submissions, and then it will become just like YouTube, where an algorithm decides who wins and loses.
At the same time, readers will have to sift through an insane amount of books to find what they like. It is already kind of like this, even before AI. This causes people to rely more on reviews and word of mouth, which becomes a volume game and is also influenced by the algorithms.
I don't think it is the end times for literature, and for those who put the work in, they can find their places. But it is a whole new set of challenges.
Photography was not the end of painting.
It was, however, a blow to painting as a vocation where a lot of your work was in the form of portrait painting.
It, plus video, also made painting culturally irrelevant. Approximately no one cares about paintings whatsoever. Books of modernity soon to follow I'm sure.
> holds real value when it is of high quality
Does it? This assumption is the core of the entire problem.
It might do as some, bluntly, self indulgent exercise but beyond that in society scarcity of things other people need and want is what drives value.
We are in for an absolutely huge rebalancing of what it is that is considered valuable.
Interesting standpoint. Do not forget what makes you human, and that's why there are a lot of different opinions and valuable, subjective literary articles.
> and valuable, subjective literary articles
There you go again.
I suspect if everyone on earth produced subjective literary articles at the rate of ten a day you would cease to consider them valuable, let alone a machine that can create them on demand.
Thank you for your opinion. To be more specific, there are many subjective views on what value is. I didn't mean a literary piece always has market value if nobody is buying it, but it can still be valuable educationally as a masterful or sublime example of literature.
You are deliberately skirting the core problem which is your perception of value is completely warped by supply and demand.
If you have a surplus of something then it will cease to be of value to you. For example you do not actually "value" air, since it is everywhere and so we ignore it, but stuck somewhere where it is not you will come to appreciate it fairly quickly.
You seem to believe there is something magic and wonderful about literature, but this is because you have only ever lived in an environment where it is scarce and not dangerously abundant. The next stage is to claim "but it's not literature" even when it's the exact same strings of words merely constructed by machines and not humans, which is curious because up until now it had always been assumed that had the monkey succeeded in typing out Shakespeare it would still have been as valuable as had Shakespeare written it himself.
> I watch videogame streamers play games. Millions of people do. There will be authors who livestream the entire book-writing-process (just typing that out makes me want to launch one today). No one person will watch the entire process, but enough people will tune in here and there, and it’ll on be on YouTube
I would definitely watch this.
Ironically, I finally caught up on an online web series (though it's been adapted to a published book as well as audiobooks) and discovered the author does live stream their writing of unreleased chapters.
They're a bit too prolific to actually dig into their overall process, but you can watch them type for hours straight. It's awfully boring tbh since there's no interaction and they're understandably focused on their work.
If your curious - https://youtube.com/@pirateaba/streams
Thanks for sharing!
I feel like writing has traditionally had two purposes, to communicate some internal feeling or idea of the writer, and to entertain. For the reader, even if their primary purpose was to be entertained, they would have no choice but to also engage with the internality of the writer. This latter point is what is changing.
I fully admit that we have been surpassed by the machine in the ability to entertain. And the biggest problem with that is that entertainment is the primary factor that drives the popularity and consequently the financial success of a book. If a writer's primary purpose is to entertain, then they should be very worried. But if their primary purpose is to convey something from within themselves, the LLM cannot replicate this. They may however lose that portion of their audience to whom the communication of profundity is merely collateral to being entertained.
I'm not a professional writer, but when I do write, I do so with the intention of expressing something that I feel, which before being thought out and written exists as something not linguistic. The language is a means for the expression of this deeper experience whose communication is the end. But an LLM on it's own is not able to convey internal experiences that it does not have. If it seems that way, it is only achieved by manipulating and rearranging the experiences of the authors that make up it's training data.
> I fully admit that we have been surpassed by the machine in the ability to entertain.
Why though? Daily here on HN, posts are pilloried for being AI generated and full of AI tropes.
AI generated content might be drowning out human generated content, but I wouldn’t say that it’s surpassed human content in a qualitative way.
I think in order to notice and take offense to those tropes, you already have to be prejudiced against them. And you probably also have the pretense that what you're reading was thought important enough by another person to write. So when you realize that this isn't the case you become disappointed.
But some people just don't care, and are just in it for novel rearrangements of the narrative structure. I'll give you that I don't notice this so much with literary content, but I have seen people being entertained by AI generated music and videos quite a bit, especially children and those who in general are uncritical of what they consume.
For example, the other day I was at a restaurant where they were playing some AI abomination of Frank Sinatra called "Take me to the Moon", and nobody really seemed to notice or care while I was becoming nauseous to my stomach. You have to already be biased to notice it, let alone to even be bothered by it.
The high-level structures of writing are prose, poetry, and drama.
Cross cutting them is the purpose, these are not mutex: exposition, narrative, description, conviction.
There's also straight rambling nonsense.
Conveying experience != having experience
Conveyance requires a source and a destination. To have an experience conveyed to you means to gain a degree of understanding of another's experience which you yourself did not have. Yet you yourself have not actually experienced it, you have merely gained knowledge into some aspects of it's nature by comparing it to your own experiences.
If a reader perceives that they have been conveyed an experience that nobody has ever had, then they have made up an imaginary experience which does not correspond to anybody else's lived reality, which may or may not be possible to have. If an LLM conveys an actual human experience, it is only because somebody's actual experience has been anonymized and conveyed on to you.
Even within your own brain, conveyance != experience.
When you report on your own experiences - like the warmth of the sun - that report is not itself the experience.
The scary part about an LLM (p-zombie) is that it can report on an experience exactly the same as if it had one, and for all intents and purposes it'd be sufficient for the task at hand.
After all, I don't need to have your experiences to know what you mean.
I agree with conveyance != experience, and looking at it again I agree with your original statement with regards to the "!=". But I do believe that conveying experience requires having it.
By conveying, I don't mean conveying the idea of a possible experience, but conveying knowledge about a real experience that a real experiencer had. What is conveyed is not the entire experience which is impossible, what is conveyed is knowledge that is sensible to someone who has similar experiences they can associate it to.
If the report does not correspond to an underlying experience, then it's just a lie. Like if I say to someone "Thank you" after they do me a favor, they perceive that I am having the experience of gratitude. But if I am not really grateful, then my words haven't conveyed information about an actual experience, about an actual instance of an experience.
So I'd say the report + the report's correspondence to reality + the receiver's ability to understand the report based on their own experiences = conveyance.
I do not consider a false report that leads the receiver to imagine the existence of a non-existent experience to be "conveyance of an experience".
Kant would agree, and went further, saying that people should always tell the truth, "even if it meant saving a friend's life from a murderer" and that it was an unconditional moral duty https://en.wikipedia.org/wiki/On_a_Supposed_Right_to_Tell_Li...
Contemporarily we have methods like radical honesty - another moral absolutist take that also asserts a primacy of truth, that as long as you tell the truth, you can do no wrong
For a lie always harms another; if not some human being, then it nevertheless does harm to humanity in general, inasmuch as it vitiates the very source of right... All practical principles of right must contain rigorous truth... Kant
So, that's the moral side of things. Whether or not you believe in truth as a moral compass (or even a component of it or some moral rubrik), I'll leave up to you! What you said:
> But if I am not really grateful, then my words haven't conveyed information about an actual experience
...gets more at the utility of honesty, as a conveyer of experience. While I might argue you still have conveyed an experience (just not one you had), because the references and end experience still happened in the mind of the receiver, the end (target) experience was had.
The invention of lying is largely due to it working just as well (or better than) the truth in the first place. Maybe the first lie was an accident, and evolution took over where only the best liars passed on their ideas.
So that you could even have an experience rich enough to contemplate!
The problem as I see it is that you cannot communicate your internal feeling if no one reads you book. Therefore, even for people whose motivation is personal expression, there is a very real risk of being flooded out of the market by AI slop.
For sure LLMs can write in perfect English (or any other language) and convey information.
But good writing is more than just conveying information. To me it's about connecting to a voice. This is a rare quality and AFAIK I have never experienced this with AI generated text or even with 99% of human written text for that matter.
I can flag a lot of content that does pretty well online as AI. There are certain rhythmic patterns in the sentence structure and the way arguments are set up that give it away. The writing is obviously "successful" in that it gets a lot of engagement, and many people either don't realize or don't care that it's machine generated, but I find it distasteful.
If I'm reading something I either want to be entertained or I want some specific information. If it's to be entertained then the story has to be good (writing doesn't necessarily need to be that good) and I haven't seen any AI generated stories that are that compelling. If it's informational then I can either ask an LLM directly and get something tailored for me, or I'm also interested in the emotional and human side, which an LLM can only imitate, since it has no theory of mind.
I think LLMs will take over making those trashy beach novels, as they're essentially clones. And people who like them will continue to buy them; no shade, some people enjoy that. But for good original stories and plots, you're going to need people to create that world.
This is a well-known author, but how come he thinks AI is better at writing? AI writing makes me vomit.
Also, I think he's too enthusiastic about the future of readers. In my opinion, reading books will be a very rare hobby pretty soon. The decline started more than 10 years ago.
The full sentence I wrote: "Artificial Intelligence is writing about as well as humans — at least well enough to cause confusion and speculation — and that’s creating an existential crisis among authors, confusion among readers, and legal action within the industry"
"About as well as" doesn't mean "better." Wild that you jumped from one to the other.
I continue: "-- at least well enough to cause confusion and speculation --" This is factual.
One thing I am sure of is that computers are better at reading than people. The way some people read makes me want to vomit.
Of course they are. I'm not even a native speaker.
I read the whole thing and it was my impression.
Heh.
Literally, welcome to: Hacker News.
It's always been like this.
Thanks! Everyone seems so nice.
> AI writing makes me vomit.
I think this was fairly obvious and fair 1 - 2 years ago. Now though, I have to wonder if people who think this with such conviction are falling for the toupee fallacy.
You said it yourself: someone you consider a well known (and presumably accomplished?) author thinks AI is pretty good at writing.
I haven't tried using an AI to write prose, but I use Claude to write code daily and it spits out the worst comments I've ever seen, along with some surprisingly good ones.
Would we not expect AI fiction to be the same? A stream of poor and excellent taste, not particularly original or surprising, samey in hard-to-define ways, but capable of being hammered into a good story by an editor.
> Would we not expect AI fiction to be the same? A stream of poor and excellent taste, not particularly original or surprising, samey in hard-to-define ways, but capable of being hammered into a good story by an editor.
That could also describe many human writers…
AI writing seems to have a quality ceiling, or an ick floor if you prefer to think about it that way around, but is OK compare to bad human writers.
I'm still committed to the career suicide that is refusing to touch generative AI beyond the supped-up auto-correct/-suggest that has replaced the older method in my IDE (replaced, such that if I turn it off I also lose what I had before), but seeing results others around me have pulled out of the hat I have to admit the tech can produce serviceable work if properly fettled by a person afterwards.
I don't think you're committing career suicide. I've seen people get serviceable results out of it, and it always involves reading the code in depth. Reading the code being slower than writing it, they're almost never getting a speed up.
That being said, the better coders I've seen aren't looking for a speed up, they're generally looking for something to bounce ideas off of, or to get a first draft going to defeat writer's block or something (I guess this is technically a speed up, but not the way people mean when they describe slinging tokens around). It's generally to wind up at a better solution at the end than to get a sustained speed up (which seems to only be possible in boilerplate heavy domains, but those were previously slowed down by the boilerplate and are really only coming up to speed with everything else, at significant cost).
I think “taste” is probably the right word. Steve Jobs’ quote about Microsoft is very relevant: “The only problem with Microsoft is they just have no taste. They have absolutely no taste. And I don't mean that in a small way, I mean that in a big way, in the sense that they don't think of original ideas, and they don't bring much culture into their products.”
Replace Microsoft with LLMs and the same quote applies. They have no culture, no connection. The code or comments they produce will not effect any future LLM session and they current have no concern how it will effect future humans except for maximizing engagement for the human at the console today.
Fiction is no different. Unless there’s a feedback loop the LLM just spits out text and will never see it again. The LLM will never meet the readers at book signings, weave in personal details that could only be known for experience.
It’s probably perfect for the write-by-committee style of movies and shows being created today. Just like organic food or locally made products, there are already creators putting Human Intelligence badges on their work and there will probably be a market (though niche) for that until human supremacist laws push it underground.
> Now though, I have to wonder if people who think this with such conviction are falling for the toupee fallacy.
If this is the case it should be quite easy for you to provide examples of excellent AI-fiction.
As for the “well-known author”, Occam’s razor says he (like most “well-known accomplished authors”) doesn’t know what the hell he is talking about.
> If this is the case it should be quite easy for you to provide examples of excellent AI-fiction.
I feel that's the opposite of the toupee fallacy. The lack of labelled excellent AI fiction supports the idea that there's a lot of excellent fiction that was written by largely by AI but nobody was able to identify it as AI, because they were looking for writing that makes them vomit.
Regardless, how about this: https://deadline.com/2026/07/ai-questions-debut-novel-book-f...
There’s plenty of people who use good toupees/wigs and are open about it (or went bald and saved their hair with mystery medicine). I’m sure if models can now produce amazing fiction it wouldn’t take more than few minutes and Claude subscription to prove us doubters wrong.
Maybe that debut novel is amazing, but the fact that we are even discussing whether it’s AI or not tells me it might not be - in any case we can’t read it yet to judge ourselves - just that publishing agents are completely out of touch. If it were truly great I don’t think anyone would care, since hey, good writing is good writing, no matter where it comes from. We would just read it and enjoy.
Similarly models are for fact able to produce music and images that are very difficult to differentiate from meat made, but still… Most people don’t like them. Something seems always a bit off, kinda artificial and boring. Lack of imagination, lack of authenticity, lack of voice.
Taking the lack of ability to identify good AI fiction as evidence that there is good AI fiction is absurd, because it means that there is no evidence which could ever point to there not being good AI fiction. There must be evidence which could contradict a hypothesis for it to be sensible.
Just like how raw sewage tastes good, but people think it tastes bad because of how all of the raw sewage they've tasted tastes bad. But if only people ate more raw sewage, they'd find how much tasty raw sewage is out there!
There would be some cases where the author/prompter would want to show off how good it is. It would even be beneficial for the AI companies
Good at constructing paragraphs, yeah. Generating coherent plotting and character development is still a human skill.
> I think this was fairly obvious and fair 1 - 2 years ago.
And that's mostly for us, developers who use it all the time and maybe for other office employees. Not sure if random person would recognize LLM-isms.
Not even the random HN user detects LLM-isms. You regularly get downvoted for pointing it out. E.g. here: https://news.ycombinator.com/item?id=49087955
There's been multiple 'unslop' contests [1] with a focus on AI 'writing' and I agree with the OP that AI writing makes me want to vomit. It's all hollow, bland writing that has zero zest to it.
[1] https://www.hyperstitionai.com/unslop-results
Agree. I've just decided that I'm limiting my reading to pre-2021 books. There's still many lifetimes of a backlog there, even at my relatively high rate.
This is happening in every creative industry. The creators on the edge of this space in every medium have seen it from both sides.
AI creative is easy to spot _right now_. It will not always be that way.
AI makes you vomit because most uses of it are hamfisted attempts at doing a thing.
People are going to eventually figure out how to polish and refine the usage and outputs so you cant even tell it is AI.
This is already becoming increasingly true for code and software. The next step is writing and then media.
> People are going to eventually figure out how to polish and refine the usage and outputs so you cant even tell it is AI.
Personally, I'm gonna go offline after this and meet more people. One of the reasons I read comments is because it's human conversation. When everything will be ridden by bots — I'm out.
There is no point in consuming this.
> AI writing makes me vomit.
There is likely a non empty set of writings that are AI written and that you don't recognize to be AI written.
I agree that the Ai-isms are ugly.
Butni also acknowledge that I am likely reading huge amounts of Ai text I don't attribute Ai.
that's not how I interpreted the whole article.
Default unspecifically-prompted AI writing makes you vomit, because it has a specific pattern.
But with enough well-defined prompts AI can be made to write in practically any style, often with results unrecognizable as AI-generated.
It does not matter which prompt you use. Even a long article with a small input (of what to write) will include weird details and structure.
People tried to fool Pangram by using custom prompts and instant feedback (rewrite until it passes). After hundreds of rounds, the text was still detected.
Although it's harder to detect such text by humans, there will be clues if you want to write a whole book based on a few pages of input.
Show me. AIvangelists love to say what AI can do, but never bother to provide any examples.
>"This is a well-known author, but how come he thinks AI is better at writing? AI writing makes me vomit."
Most of all writing makes me want to vomit. It's not impossible to write well with AI, but it's the same argument as vibe-coded slop vs. a senior engineer using the tools carefully. I'd be inclined to accept the expert's opinion here.
how come he thinks AI is better at writing?
He talks about this in the article, did you read it? Reasons include AI winning writing contests, AI hitting bestseller lists, and now apparently an AI-written book inspiring a bidding war between publishers that resulted in the supposed "author" getting a $2.4M advance (until it was revealed that the book was written by AI).
"how come" makes me want to vomit. (That "It’s a common idiom" is completely irrelevant. And note that it doesn't actually make me vomit.)
Also, here's another thought. In my opinion, here's my opinion.
It’s a common idiom
I've yet to read anything authored by AI that was truly excellent writing. It simply does not have the chops to write at that level. The great American novel will not be written by AI. Maybe "co-authored by Claude and friends", but not authored.
It will one day. For sure. I'd wager within the next ten years there will be a major controversy about an AI book winning a top prize in literature.
I see all of this as the cognitive dissonance created when humans are confronted with what scientists have been telling us forever: we’re not that special. We’re just piles of atoms, as are rocks, dogs, computers, and all of the other intelligent life that the universe must surely be teeming with.
Building something that supersedes us may have always been our purpose. Until skynet turns on us I’ll be happy to say I lived at the exact time it happened.
Humans are extremely special! They have a unique trait no other extant and few antecedant species have/had…
No other carbon terrestrial species can run for such long distances - a healthy young human can jog for 50 km or more on water and sugar: this is a (not too) breathtaking efficiency!
Can you imagine being a gazelle who has been sprinting away from a human for six hours but they just keep coming? Terrifying.
The ability to do so truly defines the species, to the extent they’re taxonomied as ‘homo marathoniensis‘ (fun fact: during the atomic age most human philosophers used ‘homo sapiens sapiens’ - more like ‘homo vanus vanus’ amirite).
And yet the cheapest car can easily outrun any human who ever lived. Everything is unique if you look into it closely enough. I would say that’s not the same as special, though I admit this is just semantics.
Just so. I don’t mean optimal locally or globally - the species is practically extinct! - only unique.
(A platypus is special too, while also rather the worst at everything it does.)
>Building something that supersedes us may have always been our purpose.
This is just religion, but without even any serious philosophical/theological grounding.
You could say that about any religion given only a one sentence high level summary. “A guy in the sky made us in seven days.” has no serious, philosophical grounding.
Your comment's first half was a kind of freshman nihilism backed up by vague gesturing at science (despite the fact that science can't really answer normative claims or value judgements). You then follow it up with a totally unsupported teleological claim which completely contradicts the paragraph that came before.
It was a Hacker News comment, not a doctoral thesis. People like you are why this site so desperately needs a block feature.
This is just a guess based off my own reading habits, but I can't help but think the average reader chooses books largely based on who the writer is. I don't necessarily mean that as a fan-boy kind of thing, but more just "I'm overwhelmed with choices and I know this guy is good". I know we'll keep getting stories of like "this LLM created book won this award" or something, but ultimately someone is going to have to put their name behind a piece of work and cultivate an audience and develop a unique voice and viewpoint, and like, I don't think the existence of a lot of spam is going to get rid of that. (I know that's a bit tangential to the blog's point -- I read it -- but I just think that the human element in a piece is going to matter the most)
From the post: "Doing a thing for the love of the thing is no different than it ever was. The problem is, there aren’t many people for whom this applies." I think this is an interesting point but I am not sure it matters when deciding whether someone should be paid for the work they have done. It is fine to have your opinion on what you think deserves your money. However it doesn't change that they have done work that has value regardless of their motivations.
> Artificial Intelligence is writing about as well as humans
Does he really hold on this until the end of the article? (sorry, I had no time to read through it). It doesn't matter how well and sophisticated the writing is, what counts is the human connection. A human is talking to me through the writing. I am not interested in having communication with a machine, I want the vibes only humans can send and receive …
Did you even make it to the end of that sentence? It would've taken you less time than writing your comment. I said "at least as well enough to cause confusion and speculation"
But apparently there's enough confusion to be had by reading a few random words from a blog post, getting your blood pressure up, and writing a comment to show off your lack of attention span.
Man, I wish Michael O. Church were alive.
What LLMs have shown me most, is:
Alternatively most blog posts could have been paragraphs. Most books could have been single chapters. Most newspaper articles could have been tweets.
That being said, most readers want you to hold your hand step by step. They don't want to the outline: they want you to give them a story. Most people and writers know this already. To me, it was new.
Sometimes pleasure is in the activity, not just the outcome.
A good movie or novel is a generally pleasurable experience from start to finish, where the conclusion on its own might have only been mildly interesting at best.
Sex is also like this.
> Alternatively most blog posts could have been paragraphs. Most books could have been single chapters. Most newspaper articles could have been tweets.
Why even read? Why even live? What's the point of all of this? Churches could be wooden shacks, 5 stars menus could be protein shakes, paintings could be single sentence descriptions, ...
You might as well sign up to be put in an artificial coma, or in a matrix pod
> Why even read?
To understand.
For most everyone else, it is to momentarily feel some sort of way. The difference is what the parent describes as a revelation, to him.
You see, for people who crave to understand, living among people who just want to feel this and that feels like living in an insane asylum where the people inflicting harm upon themselves and others will do nothing but distract themselves long enough to forget and go on to inflict harm on themselves and others again, forever.
As for 'might as well quit' - most everyone I've known who's not an idiot has already done that. Even the idiots aren't having children by the way :) Living in a matrix pod? You mean a modern city? Anyhoo...
I've colloquially coined this as the contentification of everything. The goal isn't to actually engage with something, whether it's the writing style, flavor, or artwork on display. It's purely about consumption, and the most optimal way of consuming something is distilling it down into a short, bland summary.
It's why there are people out there that will use an LLM to summarize a book and pretend that they read the whole thing because they know the general outline.
Right on.
I keep brooding on the idea that maybe LLMs will never get “good at writing” in some Platonic sense, but they will be progressively trained better to create text that a greater and greater bulk of humans will “enjoy” for reasons that are downstream of inherited patterns of thought and emotional response that the models have learned.
In other words, we become the stochastic parrots.
I've never read Howey's books but have them mentally filed next to, like, Dungeon Crawler Carl, the Bobiverse, whatever that long thing is with all the superheroes, thousand-chapter cultivation LitRPG blah blah blah as Reddit-beloved high-wordcount fan-fiction-adjacent genreslop.
I hate being a snob, but when we read "artificial Intelligence is writing about as well as humans", maybe take that with a grain of salt? It has not been my experience that AI writing is remotely near a "good" human writer.
> "Most people write a blog post hoping it’ll change their lives, not just the way they think"
Some of us just want an outlet that we control. Even though nobody will ever read it.
I don't see it as abrupt as it was portrayed. Man-made artifacts and work has always been taken over by machines in all fields of work and art. The "end of era" happened many times, through intermittent ruptures like how the tecpn8x plates move through earthquakes.
Loss of human touch is a problem. Playing a chess game while being aware that the opponent is a bot, is not same as playing against a human.
The "plasticification" is something we wanted all along, supported it through excitement and awe. Science is the plastic.
The mediocre may livestream, but I suppose the worst and the best, in seeking escape from the middle voice, now gifted to large language models, now not our own, that is, I say, the best and the worst will become esoterically poetic, will run on in agramatical authenticity, in voluminous polyglot, tiresomely peripatetic through the thesaurus. Loose change, slang, new coinage, words lazarized, older and rarer englishes, scattered everywhere.
Nonsense and beauty ascendant! Poesy freed from several centuries’ fuligin tyranny of the novel and horrid mass literacy; windmills again a just and proper enemy!
This new muse will mate with the corpse of epistemology to produce a second coming of baroque obscurant gnosticism. A fine time to be a scholar or a priest, the 2100s.
Before the printing press, information was transmitted by word of mouth, skills were learned by joining guilds, and the largely illiterate public lived and learned from songs. As AI takes over and information becomes adulterated, people will enter their own bubbles. Ultimately, the only way to escape enslavement by AI will be to return to the world before the printing press.
Given the nose-diving literacy rate history might begin to repeat itself
I scrolled through the comments as I am curious if anyone shared some ideas: What is the future of authoring by humans like? Instead of books being one of either ‘meat’ or ‘machine’, as this author is claiming, what would it look like for human authors to leverage AI for simply getting assistance for the stories they author? Could there be a healthy, non sloppy use for AI?
It is just like coding.
If you are on top of the AI, it is a powerful tool for exploration of ideas, organization of thoughts, bridging connections, outlining, filling in gaps.
But if you vibe write, it will look good on the surface but be full of illogical, contradictory statements, disconnected characterizations, etc.
There is one important point to note. Works created by artificial intelligence are not subject to copyright.
Probably.
AI writing is not good. Seeing this post, and another comment on here earlier today to the effect of “we’ve solved prose”, I’m genuinely really confused. I don’t want to be insulting, even if the first place my mind runs is basically people must not read very much or think very hard about what they read. I can’t imagine a world where anyone ingests a material amount of AI writing and things it’s even “acceptable”. But so many people seem completely blind to it. I’m actually horrified by the idea of a world where people are just slinging empty slop back and forth at each other. Probably many of us have already experienced it at work.
So what am I missing? What’s the most charitable explanation for how anyone can think LLMs are good at writing?
I think it's a difference of reading levels.
AI 'writing' makes me ill, but that's happening because I can 'feel' the prose's samey hollowness in every paragraph.
But! If someone is just moving word by word and is focussed on just the gist (the average reading level is '6th grade' here in the states), then it will not repel them the way it does us. Also, AI's grammar is impeccable, if bizarre, and the word choices can be facially erudite, so that might also be mistaken as a signal for competence where none exists.
LLMs have not currently "solved" _anything_. Not coding, not writing, not anything.
As someone who does a lot of writing, I ended up doing a lot of re-writing of what I'll call reference architectures in a previous role. I did use an LLM for some of the missing context up-front. It was essentially boilerplate and I did revise it--took out some things I didn't think important and added some things I thought might be.
But I basically never use them for a blog post except maybe to include some boilerplate explanation of some technology. (Which I'm familiar with but don't want to rewrite or copy and paste.)
Could that change? Perhaps. But I don't really use today.
Perhaps this is a question of ethics. Should we come to a conclusion that human effort makes a text worth reading despite it's objective quality. Literacy is humanities by definition.
Hijacked signals. When they read good writing they get a certain emotional response. This is not triggered by the totality of the writing but certain aspects closer to the surface level. They also get it from reading slop because it hits the same triggers.
It's like how evolution taught humans that sugar indicates a good densely-packed food source, so then corporations just started putting sugar in everything and humans eat it as ravenously as they eat actual food.
I read his take as not that "we've solved prose", but that we've created the McDonald's slop food version of prose, and people who are undiscerning will cause it to succeed and dominate the culture. While the effects of other people eating McDonald's affects us all in indirect ways, the effects of other people accepting lowest common denominator slop art/labor is going to be much more dire.
If a novel is 70,000 to 100,000 words, then a whole novel can fit in the context window for AI. That has to be good for the editing process. I.e. you can use AI to scan the whole book for various issues over and over again until all those issues are found and fixed.
Well, but then they would’ve just hopped on the eagles to fly to Mordor and just throw the damn ring into the volcano.
...Unless trench warfare your son was living through was on your mind as you wrote
Two of my favorite (sci-)fi writers both use crazy big Excel files that almost crash their computers when loading. For research but also keeping track of story loops.
I can imagine something like ChatGPT Work with markdown/python/CSV's can do quite some magic for such writers.
A common writing tool is to use a story grid. You have chapters/similar along the Y axis and title, characters, plot elements, etc. along the X axis. That way you can keep track of what information you are revealing at each story beat/chapter, make sure that plot twists work (e.g. when outlining something like The Sixth Sense).
Depending on how big the story is and how many characters/plot threads there are, I can see the Excel files getting large.
There are already tons of specialist non-AI tools for writers to be able to keep track of story elements, research, etc.
Writers (older genre writers particularly) tend to be a bit behind the technological curve, and they stick with whatever works. If you want to see a treasure trove of "old web" sites, look up the site of any sci-fi or fantasy author. Chances are its design and layout still look like something from the late 90s.
The vast majority of AI apps don't actually solve any novel problems (pun intended come at me,) they're just new versions of existing solutions "but with AI."
I wonder how concerned writers and publishers are about copyright of AI written text? As I understand it, in a lot of jurisdictions, AI products are not covered by copyright.
If my understanding is close to correct, than I can see why the million dollar advance was withdrawn.
How does one PROVE AI or human writing to determine copyright?
Who is challenging a lack of copyright on a work?
You publish a story, I take it and republish it / remix it / whatever, you invoke copyright protections and try to have mine taken down - how do I prove your work was AI and thus has no copyright, and how do you prove (or do you even have to) that it's human?
Who has the burden of proof - the person claiming AI created it, and thus it has no copyright, or the person who claims they wrote it and it does have copyright?
Well, copyright claims are usually a civil matter (in the US at least). For a civil dispute you have to make a judge or jury believe you more than the other person.
But it sounds like you are talking about a criminal charge. There the bar is higher and the government has to prove beyond a reasonable doubt. In that case I suppose the government could do things like search your computer, subpoena the AI companies for their records, talk to your editor, talk to your family and friends, etc... All the usual stuff.
Before you would ever get to a point where someone claiming your work was AI would be able to get access to your computer, editor, family/friends, etc there would need to be a bar that needs to be cleared for a judge to allow that discovery to take place.
What I'm asking is what is that bar?
Copyright is going to be assumed by default - in order to break that assumption there needs to be some process... however for all the people who are "concerned" about the copyright status of works that have touched AI I've never seen anyone make a feasable argument on how it would ever practically take place.
> There will be meat-book fanatics. […] The author will matter to them. They’ll read books the way audiophiles listen to vinyl. Every typo a pleasant pop and hiss. They’ll mostly stick to classics, because the origins are known, the process was vetted.
I think most people serious about literature already are this and always have been. Y’know, it is not hard to write a novel. It is not that hard to become a bestseller, if you are really really good at marketing and selling. Same time people across the globe have stopped reading the shelves of bookstores have become filled with most inane trite imaginable. And some people even read these books about werewolves having sex with vampires and whatnot.
Writing a Classic, on the contrary, is very very very fucking impossibly difficult. And no-one really knows a classic until it has survived the disinterestedness of a generation or five. Kids have now apparently found The Count of Monte Cristo recently. Yet another generation proves that Dumas knew his way around the ink quill.
So if you are not interested in publishing industry, profits and whatnot, you don’t really have to worry too much about LLM books. Only thing you need to do is wait a generation. Hell, few summers and no-one remembers most of the books published today. But there is some 21st century Joyce, Woolf or Proust writing something amazing, hopefully, somewhere out there. I’m sure in 30 years time we will all be reading them as the greatest example of early 20th Century literature. Meanwhile, there’s plenty of old books last a few lifetimes.
> Artificial Intelligence is writing about as well as humans
This is just untrue.
There's a lot of money in claiming it's true, with varying levels of hedging and insincerity, but AI writing follows the pattern of AI everything else: if you're not competent to judge quality, then you think it is high quality.
That's a scary notion: what happens when the money claiming that AI writing is good, surpasses the one claiming otherwise?
I think that's been the case for a while. We get
- Countless poorly-written AI books released
- Many, many articles saying that people should stop complaining and AI writing is just as good
Author gives an example with a bidding war and multimillion dollar advance.
Which is an excellent example of AI writing being potentially lucrative, but not good. Also note that that is its own counter-example: it wasn't published, and the advance was retracted.
> it wasn't published, and the advance was retracted.
Not because they didn't think the writing was high quality. The quality of the writing is what led to the bidding war. The advance was retracted because the publisher can't get exclusivity if copyright protections don't hold due to the novel being AI generated.
Bidding wars are caused by all kinds of things, and quality is only one of them. It's the only one that gets said in marketing material though.
I think this is much more likely than any of his predictions: "buying and reading a book" becomes even less of an isolated activity than... it already is becoming. Instead it's more and more a media product, attached to authors/characters that are big in other formats, specifically, video. Because people are reading less and watching more video to begin with, entirely separate from AI.
David Goggins is kind of a perfect example of this. There is zero value in an AI-written book that isn't attached directly to his public persona and media presence. Maybe today he can use AI to write the book, but it's still tied to his video media of doing tough challenges, etc. An AI-first author writing such books would get nowhere, as he/she has no credibility.
That all aside, I actually think the most philosophically interesting result of LLMs is that ideas actually end up mattering a lot more than execution. The LLM can't really think for you, and so your knowledge and experiences are actually what matters, as it determines "what you prompt."
In short, the authors that win the future will be ones with robust personal brands and a big social media footprint. The days of the mysterious writer that gets chosen and published by a small group of elites at the publishing house is probably over... at least until the ecosystem becomes so messy that curators become necessary again.
Ghostwriting has been the norm in many subgenres of nonfiction for a while now
Exactly, so in the "dystopian" future, people just use AI to write the books they were already having a ghostwriter make.
If the books are original enough to be perceived as not-AI, the mind behind it will be creative enough to make the AI origin moot.
Originality is an interesting thing.
Books are a physical medium of organizing and distributing information by gluing a bunch of flat papers between covers. Properties like originality do not follow from the physical medium itself.
Publishers of travel guides in book form use the physical medium as an instantiation of the class of "interesting coffee table adornment expected to contain wit". Does that imply an expectation of originality? I would argue it does not.
Publishers of speculative science fiction, on the other hand, instantiate objects of the class "examinations into the human condition using technology as a lens". For this class, the use of LLMs is an abomination.
There are other subgenres of science fiction, of course, for which these sensibilities would be misplaced.
It's the end of many eras, all at once. I think the same dilemma will appear for games, music and film too. As our ability to detect AI content deteriorates most authors will resort to it, at varying degrees.
> There is going to be a very wide variety of outcomes, all of which are going to coexist.
Haha, I feel like this is true for a lot of things that people argue in the ai vs. non-ai space.
>Sales and marketing, of course, were always and will always be a different story.
If we could fix these, we could have bike lanes in places where they make sense
Is the AI that writes "about as well as humans" in the room with us right now?
Did you stop reading at the em-dash?
> Artificial Intelligence is writing about as well as humans — at least well enough to cause confusion and speculation
No matter your opinion on AI writing, this is just a fact.
Reading comprehension is suffering these days.
The average person, yes. A professional, competent writer, editor, journalist? Not even close (currently).
Again - did you even read the article you are commenting on? The author gives an example of a novel that had a huge advance offered on it and then pulled because of AI concerns... in order to get to the point of having the offer a bunch of competent writers, editors etc didn't notice.
Oh, yes, because one example makes this generally true.
I'm sorry, but I have about 1M counter examples.
no, you don't have 1M counter examples. I'd also wager the vast majority of examples that you do have are on older models of LLMs that have been vastly surpassed in capabilities in the last 6 months.
You are essentially talking out a hole on the bottom half of your body.
It's annoying to come here and read 80% of comments by people who clearly didn't even read the article in question, but decided this was an excellent time to trot out their anti-AI screeds that totally fail to engage with the actual article. Human slop.
Human slop is right!
Stealing that phrase. Thank you for it.
The intentional typo "is't" needs work. Either drop the apostrophe like the yoots or stick to believable transpositions.
"But then I have to remind myself that I don’t sound like AI, it sounds like me. I’ve been em-dashing and comma-splicing up a motherfucking storm these past two decades. It was fun then, and it’s fun now."
From the source.
The em-dash thing really annoys me. Using them has been house style the last couple of places I worked and I'm not just going to stop using them because some idiot might think they were generated by AI.
I try to do the same, but I'm afraid the chilling effect is already slowly taking hold at the back of my mind.
Microsoft Word has auto-inserted em-dashes when you use hyphens in such a sentence since forever. Em-dashes are never a proof of AI-generated text.
since when is "confusion and speculation" an indicator of quality/goodness? if I ran around naked in traffic it would also cause confusion and speculation.
This is a terrible analogy.
It's quite clear that the author means AI writing is now good enough to cause confusion and speculation as to whether a given piece of writing was human-authored.
The confusion and speculation caused by your nude transit adventure is about what the hell you're doing and why.
> now good enough to cause confusion and speculation
since when is "confusion and speculation" an indicator of quality/goodness?
Logic failure.
> maybe LLMs are seen to write as well as humans because most humans have poor reading comprehension skills?
Whoa, ironic.
I think he's talking about it more in the sense that there will be speculation about the provenance of some work.
For example, if desktop fabrication tools get cheap and good enough, maybe one day garage tinkerers will be making Patek knock offs that an expert with a loupe, spectrometer, and a lot of time can't definitively say that any particular watch made after 2034 (to pick a date) is genuine. That would cause a lot of confusion and speculation in the market for used watches.
I think that's what his crypto bros comment was all about (aside from being a joke). Maybe crypto tech could provide evidence to support provenance claims?
You're misunderstanding ... that's not at all the claim. Confusion and speculation is a consequence, not an indicator. Running around naked in traffic might well cause confusion and speculation, but it wouldn't cause anyone to confuse you with an automobile or result in speculation that are one.
At least AI is better at reading than are most people.
I initially interpreted this as "Artificial Intelligence _is_ [defined as] writing about as well as humans", which would imply that common LLMs are slowly approaching Artifical Intelligence. Though English is not my native language.
The conclusion is and always will be the answer: do something because it brings you joy. Anything else it brings you is gravy
> In my dream world, the robots do all the boring jobs and humans spend all their time being creative, connecting with one another, learning about the universe, and sharing our ideas and creations in a great big prosperous circle.
Bertrand Russell famously predicted* that machines would allow everyone to “work” no more than 4 hours a day. I just had a thought when reading Hugh’s article - what if work is by and large the thing that connects most of us with one another and compels us to be creative and learn about the universe? And TBH the thought already feels trite in advance; both religious and wealthy people have said this before for various reasons.
I’ve had the same ideal in mind for many years. It would be amazing if I could work less and devote more time to art & music, reading & travel, socializing & community projects. But using AI for work (writing code) isn’t making that happen yet and doesn’t feel like it’s even aiming in that direction. Now that we have machines that can do most of my job, I’m starting to worry because there are some signs that it might not lead to utopia, starting with how it’s beginning to demotivate otherwise strong and curious programmers from wanting to be better programmers.
I don’t know, it’s too early to draw conclusions. But is anyone feeling like LLMs are giving them more time? They’re helping me work on more things, and produce more output. They’re helping me context switch more easily and work on things between meetings. They’re also sometimes tiring, and I’m often using them instead of talking to someone. They’re not yet freeing up time, and it doesn’t feel like lack of better models is the problem.
[*] Russell’s “In Praise of Idleness” doesn’t age quite as well as I remembered. :P https://harpers.org/archive/1932/10/in-praise-of-idleness/
Nice blog post but this one part is so irritating:
"In my dream world, the robots do all the boring jobs and humans spend all their time being creative, connecting with one another, learning about the universe, and sharing our ideas and creations in a great big prosperous circle. I believe such a utopia is possible. It does not violate the laws of physics, so much the laws of human nature."
What irritates me is when people say "human nature" but what they actually mean is "capitalism" or "the profit motive". Common among people who have fallen into the trap of what Mark Fisher called "capitalist realism", a capturing of the imagination by the cultural concept in the west that capitalism will be with us forever, or that it is the natural state of humanity.
Capitalism and profit motive are human nature.
It's also human nature to be irritated by your own reading comprehension (based on biases and assumptions you carry around daily) and then attribute that irritation onto a stranger instead of working on yourself. :)
Is he financed by AI money now with the "meat-book" concern trolling? Pure "inevitability".
You've created a throwaway account specifically to post against the HN guidelines:
> Please don't fulminate. Please don't sneer, including at the rest of the community.
> Please don't post insinuations about astroturfing, shilling, brigading, foreign agents, and the like. It degrades discussion and is usually mistaken.
This is not cool. Please don't do it again.
This is the black and white perspective that literally every other human-creative industry is wringing its hands over. But I think it misses a key difference that is already painfully obvious, which is creating new and engaging things will never be as simple as pressing a button. There will always be a human at the other end, pressing the button (or a human who deploys the agent, etc.)
The best comparison is photography. Instantly, a whole career path of portraiture was tossed out the window. Go to a museum and you see portrait after portrait - the best ones. There were millions of shitty ones made at all levels of society, and that job was replaced by cameras. But painting didn't go away, it got weird, it became anti-camera and more human, in a way.
This splitting of purpose and amplification of humanness is going to ramp up. Expect everyone to grow a taste for weirder, unprocessed music, experimental novels, wild and unexpected designs. Art is human communication and AI brings that into sharp relief.
Novels seem the most protected from AI - pulp/romance/beach novels will be crushed by volume and the sad remains of the publishing industry along with it, but interesting human novels will be just fine. And, much like with photography, I think AI will eventually normalize to its own artistic medium with writing, where careful usage will allow more ambitious work to appear that wasn't possible before, like a novel written using only sestinas, for example.
Publishers and "professional editors" have been releasing crap for years, even decades. No LLM needed. Humans are perfectly capable of writing so-called "slop" without a machine.
Any better word than ‘meat-books’?
Writing with LLM’s is still really hard if you want to keep your storyline consistent.
What annoys me the most about current day and age is PROLIFERATION of everything…
Not so long ago the author / creator would have to be very skillfull and prolific to write-and-write-and-write… nowaways I fee like I (we?) are drowning in more and more content…
I'd take back substance over volume :/
The problem of AI writing, to me, is its emptiness, not the fact that it's not written by a human. Seeing standard slop words or phrases triggers the "do not read this" reaction, but the reason of not wanting to read is that the text is just empty of any interesting thoughts.
The majority of us read ChatGPT/Claude/Gemini responses, even though they are, by definition, slop, because we probably asked some question (the one we did not really care about and just wanted to explore) and the response has some valuable information in it (maybe true, maybe not). So we read it.
If (or when?) the AI writing becomes interesting to read, I don't see why we won't start consuming it, and I don't see it's as something bad.
One way to make the text interesting is to actually get involved in writing; remember this short story that was here on HN some time ago? [1] It was written by AI with some serious human involvement, and I enjoyed reading it.
[1]: https://news.ycombinator.com/item?id=47431237
The biggest problem facing authors is not battles over whether something is AI vs. human. Most readers don't care.
AI has made creating content almost free. What it hasn't done is add free hours to our day to consume all this new content. The number of people who want to read a book this month is roughly what it was five years ago, and roughly what it'll be five years from now.
Supply is exploding into demand that is stagnant (at best) or falling. Upload a book to Amazon today — good cover, sharp title, a blurb you sweated over for a week — and there is no longer any reasonable expectation of an organic sale. The book goes up, and then nothing happens, indefinitely.
Five years ago that wasn't true.
The answer is paid advertising. Buy your way onto the shelf. Stand out from the slop. But every writer is being given the same advice at the same moment, and there are only so many slots. The price of the slots must rise. And will keep rising for as long as there are writers able to pay.
The old metric was it takes about $5 to sell a book via paid advertising. I think in a year we'll look back on that figure fondly.
This is pretty much what I've been saying for 20 years, but with content in general. People's time is fixed: 168 hours per week to: work, sleep, eat, excercise, commute, and then....
- read
- watch content: a movie at a theater, youtube, short-form socials, streaming services, etc..
- read a book
- scroll on the phone (won't list all the apps out)
- play a video game
- play a board game
- listen to music
etc.
So much content, fixed amount of time.
This is AI-writes-too-well is a problem with many solutions. Here's a slightly unconventional one: if an author has enough skin in their book, they could just do as Tolkien and sing their songs[^2], in public. Taking that one step further would be to establish rapport with readers by teaching them how to play those songs, e.g. in a piano or a guitar. It's like going back in time, to when the Iliad was sung and performed instead of written. Though there's no need to negate technology; if a writer is sufficiently technologically inclined, they could forego the linear story and instead release interactive fiction for musical instruments...Hm, I'll see myself out now.
[^2]: https://tolkiengateway.net/wiki/Category:Songs
> Another prediction is that MOST readers will care about machine-vs-meat as much as they currently care about publishing imprints, which is to say almost none.
This seems like a truly wild prediction to me.
LLMs can synthesize data in useful ways for all sorts of purposes, and I can see how they could be helpful in the process of a human editing and refining a work of fiction they are creating, but I can't imagine why I'd ever tolerate a work of fiction actually generated by an LLM.
The authorship and craft matters, the translation of human feeling and thought to the page matters. This isn't code where it either does what it's supposed to or does not.
Fiction from a machine is of no value.
definitely dipped my chin a few times at HH's maddeningly human way of writing prose a couple times whilst reading this
Don't you find it surprising that seemingly the largest firms don't use RLHF to prevent LLM writing from sounding like LLM writing?
I actually think AI has great potential to help low budget movies, shows and even books.
CGI is so expensive only the most generic and safe sci fi movies get made. I want a 30k movie with amazing effects.
As for writing, AI is great at filler. But a lot of books are brought for Instagram photos and social validation. Not to be actually read.
Self help books are like this. Maybe a few are good. Many are slop to funnel you into buying other products.
But with AI telling us what to buy, how long until we have books written specifically for LLMs ?
I’m sure they’ll be absolutely gibberish, but you can still pose with them on Instagram?
Anyone have a mirror, the site seems to be down?
https://web.archive.org/web/20260731121903/https://hughhowey...
Thankyou! Didn't realise archive.is is back up.
I have been using: https://periscope.corsfix.com/
Not sure if it's what you meant, but archive.is and archive.org are unrelated.
TIL.
It's easy to confuse because archive.is uses several domains like archive.ph and archive.today so you can see how easy it is to think they are related.
Also important: archive.org (i.e. The Internet Archive) is a non-profit, while we don’t really know who runs archive.today and it has some shady history, including tampering with archived pages.
https://en.wikipedia.org/wiki/Internet_Archive
https://en.wikipedia.org/wiki/Archive.today#Attack_on_Gyrova...
I started writing because AI made me obsolete in software development. Or at least gave me depression that blocks me from writing code. What I did instead I started writing my thoughts as short stories and then made an attempt to write full blown 50k words novel. It worked as therapy surprisingly well.
I write mostly without AI but then I use AI for editing and grammar correction. Sometimes when I struggle with plot threads I ask AI for an advice. When I write about something that a have a vague understanding about I may ask AI to write a paragraph or two. I'm also open about it and warn reader in preface and mark generated text using italics.
As a new author I find AI both helpful and discouraging, because it CAN write better prose than me. I think text that's written for humans should be written by humans, but temptation to ramp up quality with zero effort is huge.
Then I experimented with different regimes when I only allow myself a couple of prompts at the end of the day when I done writing and only allow minor edits. At least my conscience is clear.
Sounds just like using the gym for therapy
Writing gives meaning, despite fact if anyone reads your stuff or not. Surely publishing a book is a major achievement but many authors wrote without publishing for many years. It's a therapy and getting a book printed is an achievement, but it's not mandatory.
It works like talking to yourself, your conscience and you also get a gut feeling about using AI in humanities with time. It starts to _feel_ bad if you go too far with it. It feels good when I write my text myself. Using AI for humanities feels like borrowing on bad terms.
I think that Silo is just a darker, stupider City of Ember. Like there have been shitty derivative works long before AI came along. Now there’s AI that can plagiarize faster and better than humans and those people are sad.
AI is beginning to seriously annoy me.
I just watched:
https://www.youtube.com/watch?v=Eo5w2S-h5dI
I think it is time to jail those CEOs that drive up the prices and destroy industries. RAM prices up 350%? ALL AI companies are in my opinion a mafia structure that must be eliminated. They not only force us to have to overpay (or be stuck) but also destroy industries downstream, thus eliminating jobs. All that cost should be paid by Microsoft, Facebook, Google, Amazon etc...
The increase in prices is a signal to producers to increase production.
Email eliminated a lot of post office and paper production jobs, does that also annoy you? Washing machines eliminating jobs about washing clothes by hand? If you like a low tech life there are communities that choose to live like that, and happily do.
I think this is even worse for music. There’s almost no way to tell if a track is produced by humans or AI.
I’m pretty sure that bands I’ve seen live aren’t AI. Of course, I’m still yet to hear an AI track that approaches ‘good’, so it’s irrelevant, for now.
If there's no way to tell, does it matter?
Some people downvote you, but is not able to produce a meaningful response.
In my view the best writing presents a portal to someone else's mind. LLMs do not have a mind. They can produce text and perhaps even 'weave a tapestry' of plot lines that make for an enthralling and well written story. Maybe a meaningful one, if prompted with a particular set of themes and narrative ideas. Is that a voice?
I don't know. But I do think, or rather I hope, real voices will prevail. But publishers (including KDP) have a rough road ahead of them, if they also want that. Just search YouTube for tutorials on how to spam Amazon with AI generated works, hoping for a break out. It is sad.
Is anyone working on cryptographic human-crafted attestation, without feeding the work into LLMs?
yeah for me it's all about specificity, like no one else but Philip K. Dick could have written VALIS, it is so tied to his life and psychology. Its difficult for me to imagine something which is is build on huge amounts of various cultural output to retain the specificity of a singular limited human life-perspective.
Is there a cap on AI intelligence?
Is it just a case of throwing more and more resources at a pattern recognition machine with no upper limit except the energy of the universe and the economy of the algorithms?
Someone else suggested the maximum intelligence is capped as the average intelligence of its training data, which is an interesting take.
François Chollet made an interesting geometric analogy on Twitter. We tend to view Intelligence as something unbounded, like the height of a pole. But it could be something closer to the roundness of a polished sphere. Once close to the limit, incremental improvements are less and less significant.
The intuition we have about the nature of cognition and its whole shape and limits is almost entirely built by looking at ourselves and our peers, comparing that to an ant. I feel this is largely naive and misguided.
I've been interpreting AI task solving capabilities as a search problem: being smarter lets us prune more of the tree and try out different things more efficiently. I think this fits into the sphere analogy quite well: at the limit, we need to experiment, use the experiments to create new abstractions and build on them repeatedly. This doesn't necessarily require more raw intelligence, but the time or resources to find how the pieces fit together.
The cap will be reached when LLMs run out of original material to train on and the large majority of their training material is AI generated content.
Completely disagree, I think LLMs write horribly badly.
Documentation, marketing, prose, it’s all horrible slop writing, immediately recognisable regardless of model
Is that still going to be the case in 1 year? 5? 10?
It will be worse after consuming its own crap and poisoned sources.
Just another trillion, trust me bro.
Article predicts that we'll see tools to "roughen up" slop to pass as human. Surprised we aren't there already...
I'm already being more careless with spelling and grammar because it gives authenticity.
Maybe people will revert to Jeffrey Epstein style prose to avoid any suspicion of slop..., e.g:
“no , bed, i, m laying on thefloor so it can't move.”
Another wave of marketing from AI companies?
Only speak advantages of AI, no or a few trivial disadvantages.
I read the author's entire post and find it interesting that the author attributes success to having a good story to tell based upon some wild youth experiences while sailing. Not being able to type real fast, not some magic sauce from a "how to be an author class" but being able to spin a good yarn, one good enough for other people to pay to hear it.
There was/is a business position as a middleman of sitting between a storyteller and the printing press, called authoring or writing and that's slowly been denigrated for over a century by mechanical typewriters, word processors, spelling checkers, grammar checkers, now LLMs.
However there's always been a shortage of good story tellers with a story to tell, and for at least a century or two been a MASSIVE oversupply of the middlemen mentioned above, who end up failing. They can type, and they insist thats all they need, and the ability to spin a good yarn is just fluff or unimportant. Then they fail, so they're probably incorrect.
I suspect as long as the author can spin a heck of a good yarn based on his personal sailing experiences mixed in with whatever else, the author will have gainful employment in the book biz. As a storyteller. Not as "a guy who types really fast" or "a word processor technician" or whatever.
The middleman author who used to have a job writing generic romance novel #7994126 and was hoping to make money off writing #7994127 is probably utterly doomed. Why does the publisher need him if we have LLMs, or even better, why do we need publishers if we have LLMs and can enter a short simple plot just as well. "Claude generate me a 75K word romance novel similar to 'Just for the Summer' from last year but write all my favorite past characters into it. Especially character Anne she's my favorite."
But lets be realistic, those middlemen were already doomed anyway. Very few people in the book biz can spin a good yarn and that shows up in the stats of "failed writers". LLMs just mean they'll generate more scrap paper along the way to failure. But about the same percentage will sell books and about the same percentage will fail unless AI itself causes a substantial difference in readers tastes (honestly, I think not...)
Compare it to programming, about which we supposedly have more experience. LLMs generate more code that no one is responsible for nor understands fully. But does it make more or less successful startups? Nah, not really. Doesn't seem to have much effect on "idea people" at all. The people whom always generate bad ideas crash the plane with no survivors a little quicker, the people whom always generate good ideas make money a little quicker, most people in the middle are unaffected beyond change for the sake of change, which is always expensive.
Deleted because I realised I am not in the mood to argue with people who don’t read; don’t make an effort to understand an argument; or make shallow dismissals thinking they consist in any way of a valid point. There are more important things to worry about than internet randos (and that includes me, if you’re offended).
I think his “ALL AUTHORS” point was about how authors will use AI in some form or fashion during their process, and points to how he did with his google search and his AI Gmail response. I don’t think he meant all authors using AI for their writing. Here’s the snippet:
“Another thing that will happen is established authors will turn to AI for assistance, not just for writing but for research, cover art, editing, returning emails, having an assistant, doing marketing, and thousands of other things that haven’t been invented yet.”
I take it as a general, this-tech-is-being-shoved-down-our-throats kind of way
That's how I meant it. Love to have read the original comment telling me about all the authors out there who are somehow able to avoid the AI that is now in EVERY SINGLE THING WE USE, including our email and web searches.
Wow, what an epic misread :)
I think I saw an em dash in there, guys!