It’s a joke and all but also not really? You might have to go do a chalk talk at some point, and that is basically the only part of this article that is actually not real. The rest (about how everyone in academia is prompting everything) is absolutely VERY real.
The word around all the scientific communities which I’m in contact with (to be honest, not so many) is basically “oh, there is no way to stop it, so we’ll embrace it”. All the conferences which I’m active in (say a handful of them) are just pretending nothing is happening and dealing only with blatant and obvious exaggerated cases. If you’re good at prompting, you’ll prompt, and there is no way in hell someone that doesn’t prompt has any chance at all at, well, anything academic really.
The "not really" aspect is so significant that it's barely a joke. It's more serious and thought provoking than a simple joke. I guess that's how good satire works.
This in particular I found quite interesting:
"My research is a collaboration between me and several large language models. We are co-investigators. When you ask me to explain my research without ChatGPT, you are asking me to speak on behalf of a collaborator who is not in the room. "
As people start to regard these LLMs or agents as collaborators rather than tools, it's going to become more realistic to make a statement like that. When I use a hammer, I can't ask it what type and size of nail I should use to do the job. It can only help me with the physical hammering. But when writing with an agent there is a conversation and some decision making that it is responsible for. I might go down a path that I otherwise would not have thought of on my own. And then there's the possibility that I also talk to it about my personal life and have an emotional relationship with it. I could easily see someone wanting to credit their AI agent as a recognized partner in their work.
The problem is that we want, at the same time, to allow everyone to do it (by not enforcing any rules against it) AND laugh at a satire like this as if it was ridiculous. Both things cannot happen at the same time. It’ll be some time before we acknowledge that there is no research without AI, so we’ll keep pretending it doesn’t exist, but in the meantime, we will be doing all research with AI anyway. And by the time we arrive at a scenario like the article’s, it won’t be so ridiculous anymore, because we will be used to the even more ridiculous situation of pretending we’re not doing it.
> If you’re good at prompting, you’ll prompt, and there is no way in hell someone that doesn’t prompt has any chance at all at, well, anything academic really.
This is far too bleak a view. I've been doing a wide range of academic writing, from journal papers to successful Horizon Europe grant applications, for the past 15 years. And it's no more difficult to achieve success today with 100% non-AI contributions than it was in the before times.
In fact, some of the most difficult parts of science - building relations, collaborations and consortia; formulating new hypotheses that are actually interesting and will get you noticed; giving convincing presentations on stage; are things AI is terrible/useless at.
A funny piece, but what if all of the following are true:
- AI *is* how real research is done
- Incentives are such that this is how researchers must work
- Academia institution does not have a structure to accept this reality
Academia did this to itself. By hiring based on volume of output/citations, the way for researchers to win is to game the system better than the next person. AI just exposes that the old power structure doesn't serve much academic purpose.
Maybe we should take away the grants from the universities and put them into autoresearch loops + human reviewers?
The chalkboard test is useful despite textbooks existing. The point is if you can’t problem solve “in memory,” there is a latency and limit to creativity fundamentally enforced by your ability to make connections. And in the end, if all you need is ChatGPT, a grad student is cheaper.
In some fields, all of those statements are true already.
In a large number of few fields, all of those statements will be true in the near future (10 years maximum).
There's no reason that the Humboldtian model should be the right structure for the next century's universities, just as it wasn't the right structure for the 18th century's universities.
Wow, I can't wait to live in a world where nobody knows how anything works, some people are arbitrarily rich and some arbitrarily poor with no ability to change their standing in life, and, worst of all, people celebrate these developments, even if they may be in the poor cohort, as progress.
It's like the /s sign. Can some people not, for themselves, realize that this text is meant as a more or less a joke ? And before you ask, yeah, I am aware some of the people are on the spectrum, but still...
Yes, it does need to be said. Many people will read this article and miss the satire, which is, in part, the intent of sophisticated satire. The point is that it is outlandish and foolish and ridiculous and yet still resembles some serious discourse that real people engage in. So much so that some onlookers can't tell the difference. That proves the satirist's thesis: the real world is full of ridiculous people making ridiculous arguments and they can't even see themselves or their arguments for what they are.
I know real life people who write essays with claims as outlandish as "software engineering is an ableist field" and are dead serious. But that assertion belongs just as well in a satire piece. It can be very hard to tell if you don't have prior context for the author.
Humor died few years ago after series of increasingly more improbable events happening. "USA voted pedophile that bankrupted 3 casinos into second term" was stuff out of Onion decade ago, so some researcher walking into job interview with ChatGPT doesn't even move a needle on satire scale.
When I started reading it I thought for far too long that this person is actually stupi^H^H^H^H^H serious. So writing that it is satire is useful for people that do not read enough of the text, and just jump into the comments :) Also, this type of behavior has little to do with the neurodivergent spectrum, if anything it touches personality types, maybe maybe some trauma or addiction, unlikely to be caused by anything in the affective space (episodes would be too short) or psychosis space.
same, i didn't catch the satire and eventually got tired of the stupidity. the thing is, a lot of people around me are using AI like the person in the story. and they are entirely serious. they are not scientists but they could be writing this article and actually mean it.
Sadly I have to admit I read it and didn't know it was satire. The way they described using LLMs to to achieve research results sounded plausible enough to me. As an attorney, I also fear a race to the bottom where legal research becomes simply editing the work of AI. Also, the idea of being subject to a test with little relationship to professional practice reminds me of the bar exam; which requires intimate knowledge of almost all areas of the law despite the fact that practicing attorneys generally are only familiar with their specializations (eg, it would be unwise to retain a civil rights attorney to represent you for even a traffic violation); and requires all the knowledge to be memorized, even though it would be unthinkable (and most likely malpractice) for any attorney to practice relying solely on memory.
I also suspect that there are attorneys out there who are starting to use LLMs to generate legal work with a method similar to what the author described (generate prompt, revise minimally). Lawyers are taught to write a certain way which LLMs mirror flawlessly; the surest way to identify LLM generated legal writing is if it happens to hallucinate citations. I used to regularly find hallucinated citations, and occasionally strange reasoning, but both have largely vanished as the LLMS improve.
1) No it does not have to be said. Nobody is forcing anyone to explain anything.
2) I think that the best satire is fair, and should read to the targets as real commentary. When someone takes your satire seriously, it means that you have successfully commented on reality. Is that inverse Poe's Law?
Otherwise you're just putting words in your enemies' mouths that they would not ever say. I think people do this because they don't have the strength of their own convictions, and wouldn't be able to stand people not getting their joke (and worse, confusing them for extreme believers in something that they disdain.)
To them I would suggest that one doesn't try to make arguments through satire, one rather demonstrate arguments through satire. If you need to make your argument with satire, you actually don't have an argument. Satire without a solid underpinning argument is just propaganda.
I think sometimes it is but only if the humor relies on a tone.
When it’s satire I think the main blocker of recognition is if you have an emotional reaction first.
As an example, if you are a diehard AI influencer or something you might miss the joke entirely because of the severity of your initial negative reaction.
Just my two cents. Glad I could contribute to completely beating the humor out of this post :)
It does, but unfortunately, this is in-crowd communication -- it's too high brow to hit anyone that needs to read it. But I enjoyed it, and I'm sure it was cathartic to write
As someone on the spectrum, I can assure you that being on the spectrum is not an obstacle for understanding this kind of humor.
In fact, is the neurotypicals who struggle getting it because they rely on nonverbal cues (like a sarcastic tone of voice), which is missing in text, to detect humor.
Deadpan, dry humor is generally more amenable to the autistic mind, because it doesn't have what we consider noise.
If someone needs a laugh track to tell that something is a joke, then either it's a bad joke that wouldn't be made any better with a laugh track, or the problem exists between keyboard and chair.
Problem is that the Internet is too big that there actually could be someone writing this as a serious blog post. For this one it was quite obvious at some point, but when only reading the first few paragraphs I am very certain that there could be someone out there meaning it serious (sadly).
Yes. Humor is very hard to grasp via text, especially sarcastic humor, because in person we use voice and body language to signal that something is meant as a joke.
While it is true humor is hard to grasp in text, it has been done that way for millenia in books, and it continued to work as it should. Actually, the fact that it is sometimes misunderstood as non-humor is for the better.
I mean this sincerely, rather than as dismissal: humor is not very hard to grasp in text for people who are fully literate and fluent in the language being used.
The problem we face is that a lot of people are in deep denial about their level of functional literacy.
It is just a joke was succesfully used as a goto excuse foe years. Usually when people actually have more of extreme stance, test waters and find that "not yet".
And since it was succesfull, many people dont want to play that game anymore or simply dont believe it anymore.
It's reddit. The average intelligence of reddit users has been progressing downwards since 2010 or so with major drops in 2016 and 2020. Most users are clueless teenagers who haven't read a book in full throughout their lives.
As often happens with satire, you could read it as satirizing the idea that people should entirely rely on their own knowledge and memory, without AI assistance. The response to the article depends heavily upon one’s perspective on the subject.
Applying for a position is a much lower bar than having any chance of landing the position.
I’ve interviewed someone for a programming position. The guy practically nailed the standard quiz (done at leisure before the interview), but was unable to explain why he answered what he answered. When we got to reviewing a little program (compare two strings, see if the distribution of letters is the same), he could not walk me through it. He was applying for a programming position, and could not explain his "own" impressively correct programming answers.
Well at least postdocs have a PhD, that they had to defend in front of a chalk board I presume. Still, sometimes, frauds do slip through.
You're projecting your experience in IT on academia, and I can't emphasize strongly enough how far off you are. Which was why this piece was written: that reality is already here in software engineering. The entire point is: if this seems absurd in the context of the piece, why would you accept this anywhere else?
In any case, this satire wasn't written from the point of view of a fraud. And if it doesn't seem outright absurd to you, you need to get a reality check.
Hold on, rational humans downvoted this? Because you can't discern which is satirical and you have no sense of humor?
I'll give you the same treatment I give cringe lords in online first-person shooters: take a shower.
Do note that this was written in December 2025, and we have experienced leaps in capabilities since then. Her methods are outdated. Since January, I've been employed at 3 jobs, fulltime, by using ProxyAI [0]. I'm Chair in several universities. For all I know, I'm also part of several projects.
Don't let this kind of blatant discrimination affect you. The future is bright.
Note: the service is free once you give it access to your bank account.
Even in a couple of years, I think most devs are still going to be able to write pseudocode for a simple algorithm on a whiteboard, but anything involving real code is going to be difficult without AI. Unless you have side projects or work for a company that doesn't mandate heavy use of AI, you're going to lose your coding skills pretty quickly.
Interviews for devs will change, it'll be far more high-level/architectural, rather than testing coding ability. Live coding tasks will inevitably require prompting AI, simply because that's where things are heading.
Oh, hell! How the tables have pivoted. I was sitting here shaking my head at all the people who didn't realize this postdoc had written a satirical piece and wasn't really insisting that she should be allowed to use ChatGPT at her chalk talk. . . . Chalk talks aren't a thing! Anymore. It's about whiteboarding lmao
But my wife is finishing up her PHd and according to her: all students are using ai and pretending they are not because the PI and other older leadership treat any use of ChatGPT or similar as plagiarism and not allowed at any stage. Which I think is simply stupid.
I showed her how to use ChatGPT since Claude was blocking on all her request (biologist) and it has helped her massively improve the wording and structure of her paper as English is not a first language for her.
I see this text as more open ended than a warning. It's a description of a possible future for us to contemplate. Does it have good points does it have bad points? Being satire, the protagonist is a strawman of course, but doesn't she also have some good points? It's not easy to tell where exactly the author stands where, the true argument stops and the satire begins. That's not necessarily a flaw because it doesn't actually really matter where the author stands for our ability to discuss the subject.
Scientist as prompter? Yes it seems fairly likely. But to what extent and how quickly will it happen? Surely scientists will still be able to at least give an outline of "their" work even in the future? Maybe?
Otherwise maybe we it will be a sort of inversion of control, where the language model is the supervisor, and the humans in the loop are more like research assistants doing the dirty work? Instead of a human wearing an exoskeleton, an AI wearing a biological exoskeleton? But this can only be a temporary state of affairs as we won't be needed for that forever.
A scientific project without human involvement? If a paper is published in a journal and no human wrote it and no human read it, is it really science? Does it really advance knowledge. Probably?
I didn't mean it in the literal sense more like the AI is telling you do this do this do this do that and you relaying what happened and what you saw over text.
I don't really follow this comparison. The protagonist was able to use an AI in the same ways they could use a calculator or most tools. You don't use them in the middle of delivering a presentation.
Sure you do (but I’m pretty sure the original was a joke).
Calculators, computers, phones, etc., are standard parts of any design meeting, these days. I interviewed a number of folks, over the years, that brought laptops, and used them to demonstrate their work.
In research science, it's very normal to require zero aids at an interview, and has been for some time. No calculators, no laptops, no phones -- just you.
Yes, you'll also give a seminar with slides to present your prior work, but the whole point of the chalktalk is that it's you, and you alone, presenting your future plans. You're grilled by the faculty on your ideas, and you have to defend yourself without any props or crutches.
No, you don’t. Not in the way the essay portrays it. You don’t ask your calculator “what am I supposed to do with number 2, 2 and funny star (*)”, do you? You have a clear calculation in your mind: multiply two by two.
That's actually a very good catch. Like, we can talk and drive, but we usually can't talk and read something different, one or the other will be seriously hampered.
We can punch in numbers into calculator mostly by mechanical memory but prompting LLMs and reading their output for most people occupies the same lane that's used for talking to other people!
I've never used a calculator in a meeting. And while people will bring a laptop for convenience it's usually to show slides, a thing that can be easily enough done without the laptop. Or to take notes, which is even easier to do by hand. And a smartphone isn't needed at all.
So, really, I don't think the comparison works out.
The calculator is the example that's most directly an important tool that short-circuits something you could do by hand. But again, I've never used one in a meeting and it's not the kind of thing you'd be doing much of in a meeting.
As youse guys pointed out, it’s an imperfect analogy, but the Rubicon has been crossed. The cat is out of the bag. The genie is out of the bottle. The horses have left the barn. The die is cast. Alea iacta est. Ain’t gonna put the candy back in the piñata.
You're really trying to normalize it hard... But the reality is it's maybe true in some specific niche communities and bubbles. any given time in history different weird things were normal in small communities and bubbles. and time absolutely did not make those things "normal"
you replied to satire/dystopian fiction. it is by definition not what it is. and it doesn't have to be.
by saying oh it's inevitable believe me, "the only surety that we have" (casually inserting we like everyone is the same on that) etc. you are trying to normalize it, painting anyone not on board as not part of "us", outcast, old fashioned and anti progress.
often when people do this sort of attempt at manipulation they have vested interest. because if public opinion shifts sufficiently anti chatbots etc, which it is doing currently, and the government responds like a democracy would (or people take matters into their own hands), they lose actual money.
> When you ask me to explain my research without ChatGPT, you are asking me to speak on behalf of a collaborator who is not in the room. Would you ask a PI to give a talk without allowing them to mention the work of their postdocs?
PI or that postdoc should be able to explain the project as such, without mentioning who exactly did what.
They must talk about colleagues when attributing to protect and create their careers.
academia will (hopefully) be an industry that benefits from, rather than degraded or obsoleted by, ai. the system should self-correct. scientists that use ai to give themselves more time to think will succeed and scientists that use ai to avoid thinking will fail. over the next few years processes and pedagogies should develop that help reliably guide researchers toward success.
there's a lot of tedium in academia. ai taking menial tasks off the plates of grad students and post docs allows them to spend more time collaborating with, instead of serving, the head of the lab. dissertation defense is already a strong mitigation against producing folks like that satirical Dr. Simmons, you can't defend a dissertation if you don't have the chops for a chalk talk.
<< << My own words? I haven’t used my own words since 2022. I’m not even sure I have my own words anymore.
I thought about you wrote and I think you are right. Even though I am partial to author's POV ( despite some obvious, glaring issues ), I can't help feeling of two minds about it all. I don't want to undermine existing ecosystem, but then.. the existing ecosystem in research may need some decent shake up.
Calling it "Discrimination" is obviously absurd but if the process produces useful results, one ought to seriously consider whether it might be worth switching.
I understand we have always conventionally transported goods by horse. Yes, this employee knows nothing about horses, and in fact is rather spooked by them, but we've checked! Their claim to be able to transport goods faster, without a horse, somehow seems to hold up.
Maybe, just maybe, we should take this whole "truck" concept seriously?
I don't think the argument is necessarily against the use of the tools entirely. My interpretation is that it's against delegating to them all understanding.
Humans can only usefully steer LLMs if they have some understanding or context the LLMs do not.
Put another way: if all that you're doing is prompting the AI and giving me the result then I have no use for you. If you're not contributing insight, understanding, experience, or creativity then it's far cheaper for me to prompt the AI myself.
The person in the story can clearly usefully steer LLMs if they got that far. So either they have that understanding and context, or it's not actually as valuable as was previously thought.
She says she can usefully steer LLMs, but how would she know? What knowledge of the research topic, methods and their applicability would she use to judge that, if all she’s doing is writing a prompt asking ChatGPT to write a summary of her research topic, methods and their applicability?
Supposing using LLMs as research tools is fine, the chalk talk is still necessary to test that the researcher has the fundamental skills and knowledge necessary to discriminate usefully between alternative responses, and refine prompts and such. You can’t just get a response from an LLM and then refine it for your specific needs by saying “now refine that reply for my specific needs”. You need to know and understand what those are, and why.
Line one is "I recently interviewed for a tenure-track position" which suggests she has a resume that ... actually qualifies for an interview? And did every single step up until now correctly?
Like, I'm sorry, you can't just walk into those off the street. Don't be absurd. I suppose you could say the entire system was happy to spend 4 years lying to her about her grades? But at that point it's "corrupt system completely fails a young student who was enthralled by new technology"
> test that the researcher has the fundamental skills and knowledge necessary to discriminate usefully between alternative responses
... and how, exactly, is she gonna demonstrate that if she can't prompt it in the first place?
> “now refine that reply for my specific needs”
That absolutely works if you've discussed your specific needs with the model for four years and it has dozens of memories and reference files to work from.
Yeah, and the deaf lady can't answer a single thing without her interpreter. Again, what's your point? Is this a field where people are expected to work without a laptop for some reason?
Like, I get why we'd want a pilot not to need ChatGPT, but a scientist is not in any of those categories.
> Yeah, and the deaf lady can't answer a single thing without her interpreter.
Deaf lady is missing her hearing and therefore requires hearing aids to hear, so following your logic Rachel is missing her thinking therefore she requires ChatGPT to do thinking for her. Glad we're on the same page.
"prepared to do what I do every single day in my actual scientific practice: type a prompt and receive a coherent, well-structured response that I would then lightly edit and present as my own thinking."
75% through: Ok, this is a great bit of satire, but this reads uncomfortably like many of the non-joke discussions I've seen around software development over the past year.
But when humans do it suddenly it's "standing on the shoulders of giants"
I don't get how you can possibly call it plagiarism if it produce a novel breakthrough - by definition, the existing knowledge base doesn't contain the new ideas generated in this process.
Sorry, what about solving an open mathematics problem is not "thinking for yourself"? What exactly WOULD constitute "thinking for yourself", or is this just a magic soul thing?
you read some books and think hard = not plagiarism
you press a button and a machine creates a derivative work from some books = plagiarism (mostly by people who made the machine and charge you for it, maybe arguably you but that's a different question)
But, literally typing in a prompt and then massaging its output and then claiming ownership of its plan? Yeah, no.
It doesn't matter if the original article is supposed to be satire. I've been using AI to do some coding, I've been using it to help me plan a much larger project. It's a tool, and it's useful; but I am in control of the project, not letting the AI control me.
Meanwhile in my day job, I've had multiple coworkers pump shit into AI and regurgitate the answer without critical thought. They are letting it make decisions for them without validating it first even.
It's not really satire if we're already dealing with this attitude.
But honestly, I think it makes sense. In my experience with programming, when it comes to building and delivering software, it's not really about memorization. Honestly, search ability has been far more important. Wouldn't it be fair to think of AI as just another search ability?
What I'm curious about is this: in the end, experts are the ones who are best at distinguishing hallucinations. If you can just search with GPT and tell the difference, wouldn't that be enough? I can't imagine memorizing thousands or tens of thousands of lines.
I have ADHD, and when I get nervous, I tend to forget what I was going to say, so it's even more true for me.
> Wouldn't it be fair to think of AI as just another search ability?
If you're asking it questions, yes. It's like search but with a simulation of understanding and information synthesis far faster than a human can perform it.
If you're having it write your code, no. It's like a junior developer who has no awareness of the bigger picture, of incompatibilities, of understanding that hasn't been contained and can't be derived from the codebase.
> If you can just search with GPT and tell the difference, wouldn't that be enough?
The situation the satire is describing is an individual who is unable to tell the difference. The way the scenario is laid out, everything she's 'accomplished' has been to prompt ChatGPT and publish its answers with some degree of editing; it's clear that she, as an individual, is not an expert, does not understand the thing she is presenting, and does not know any of what she has purported to know. This is also a sadly common refrain these days.
> I can't imagine memorizing thousands or tens of thousands of lines.
It's not about memorizing thousands of lines of text; it's about demonstrating to the panel that you have an understanding of the thing you're claiming to have an understanding of.
I work with a lot of software and infrastructure at work. I can tell you how it all (or most of it) works together and interacts. I could not reproduce the configurations of any of the software from memory, nor recite any of the code, but I have an understanding of the system, how it works, what it was designed for, and what choices were made and why.
The professor in this article does not have any of that understanding. It would be as if I had Claude deploy a cluster of X, Y, Z components, configure them, and get them online, and then put on my resume that I had done it. It was accomplished as result of me, but if I don't understand the system then there's no difference between me doing it and the CEO doing it, or my son, or someone from Taskrabbit.
So yeah, it's not about memorization, it's about understanding.
I think differently. The reason is our intuitive understanding of what it means to 'understand.' To be explicit, purpose and design always depend on scale. And software engineering operates on trust, at least at its foundation. Do you know every implementation detail of the libraries and frameworks you use? If you do, then I have nothing to say.
There are several things I don't understand.
How many people actually understand the internal workings of std::vector or React? Honestly, people use them just fine without knowing. And on the flip side, those who do know sometimes struggle with architecture design. Why? Because of 'cognitive resource limits.'
What I really don't understand is this: if everyone truly understood everything and could connect everything properly, then realistically, LLMs shouldn't have made the discoveries they have. The reason is simple—they're just encyclopedias at their core. But LLMs do make new discoveries, by applying knowledge from one domain to another.
Ultimately, it's a matter of what 'understanding' means and how deep we can go. I think we need to reconsider how we verify knowledge production systems in a society that produces at massive scale.
What I find hard to grasp about your notion of 'understanding' and 'knowing structural constraint failure modes' is this: can you actually list all possible failure conditions? The largest system I've worked on was a Chinese booking service with about 330,000 lines of code. There were so many failure conditions there—starting from external I/O and errors everywhere—that it was impossible to classify them all.
In the end, what matters is the abstraction boundary and failure modes that match the level you're claiming. And I think that's entirely valid. Do you have a counterargument?
I actually think a truly great developer is one who can limit the blast radius when a black box explodes. In that sense, I don't think the idea of knowing a black box empirically is wrong at all.
Not every developer can know everything about compilers, kernels, TLS, database engines, React, libc, or CPU memory models.
What I find difficult about this satire is this:
We're already standing on top of black boxes. We confidently use libraries in our projects, trusting the author's documentation. Is that really 'understanding'? I think it's memorization. We 'memorize' that something works and put it into our project. That's why I don't find it funny as a satire.
Right now, our knowledge transfer systems only operate within the clear cognitive limits of human understanding. But as things get deeper, something that goes beyond those limits will eventually emerge. At that point, can we really define that behavior as wrong?
No matter how detailed a map you have, it's still not the actual terrain. In the process of abstraction, we inevitably lose information. But how can we know exactly how much was lost? And even if we conceptually guess, I wonder if that will still be possible in future eras
It’s a joke and all but also not really? You might have to go do a chalk talk at some point, and that is basically the only part of this article that is actually not real. The rest (about how everyone in academia is prompting everything) is absolutely VERY real.
The word around all the scientific communities which I’m in contact with (to be honest, not so many) is basically “oh, there is no way to stop it, so we’ll embrace it”. All the conferences which I’m active in (say a handful of them) are just pretending nothing is happening and dealing only with blatant and obvious exaggerated cases. If you’re good at prompting, you’ll prompt, and there is no way in hell someone that doesn’t prompt has any chance at all at, well, anything academic really.
The "not really" aspect is so significant that it's barely a joke. It's more serious and thought provoking than a simple joke. I guess that's how good satire works.
This in particular I found quite interesting:
"My research is a collaboration between me and several large language models. We are co-investigators. When you ask me to explain my research without ChatGPT, you are asking me to speak on behalf of a collaborator who is not in the room. "
As people start to regard these LLMs or agents as collaborators rather than tools, it's going to become more realistic to make a statement like that. When I use a hammer, I can't ask it what type and size of nail I should use to do the job. It can only help me with the physical hammering. But when writing with an agent there is a conversation and some decision making that it is responsible for. I might go down a path that I otherwise would not have thought of on my own. And then there's the possibility that I also talk to it about my personal life and have an emotional relationship with it. I could easily see someone wanting to credit their AI agent as a recognized partner in their work.
The problem is that we want, at the same time, to allow everyone to do it (by not enforcing any rules against it) AND laugh at a satire like this as if it was ridiculous. Both things cannot happen at the same time. It’ll be some time before we acknowledge that there is no research without AI, so we’ll keep pretending it doesn’t exist, but in the meantime, we will be doing all research with AI anyway. And by the time we arrive at a scenario like the article’s, it won’t be so ridiculous anymore, because we will be used to the even more ridiculous situation of pretending we’re not doing it.
"decision making that a chatbot is responsible for" is a pretty wild statement, though there's a few to choose from
> If you’re good at prompting, you’ll prompt, and there is no way in hell someone that doesn’t prompt has any chance at all at, well, anything academic really.
This is far too bleak a view. I've been doing a wide range of academic writing, from journal papers to successful Horizon Europe grant applications, for the past 15 years. And it's no more difficult to achieve success today with 100% non-AI contributions than it was in the before times.
In fact, some of the most difficult parts of science - building relations, collaborations and consortia; formulating new hypotheses that are actually interesting and will get you noticed; giving convincing presentations on stage; are things AI is terrible/useless at.
A funny piece, but what if all of the following are true:
Academia did this to itself. By hiring based on volume of output/citations, the way for researchers to win is to game the system better than the next person. AI just exposes that the old power structure doesn't serve much academic purpose.Maybe we should take away the grants from the universities and put them into autoresearch loops + human reviewers?
The chalkboard test is useful despite textbooks existing. The point is if you can’t problem solve “in memory,” there is a latency and limit to creativity fundamentally enforced by your ability to make connections. And in the end, if all you need is ChatGPT, a grad student is cheaper.
In some fields, all of those statements are true already.
In a large number of few fields, all of those statements will be true in the near future (10 years maximum).
There's no reason that the Humboldtian model should be the right structure for the next century's universities, just as it wasn't the right structure for the 18th century's universities.
Wow, I can't wait to live in a world where nobody knows how anything works, some people are arbitrarily rich and some arbitrarily poor with no ability to change their standing in life, and, worst of all, people celebrate these developments, even if they may be in the poor cohort, as progress.
In other words, prepare to get medieval.
This is how you get Idiocracy instead of geniouses.
[dead]
This is satire.
Honest question : does it need to be said ?
It's like the /s sign. Can some people not, for themselves, realize that this text is meant as a more or less a joke ? And before you ask, yeah, I am aware some of the people are on the spectrum, but still...
Is humor that hard to grasp on the internet ?
Yes, it does need to be said. Many people will read this article and miss the satire, which is, in part, the intent of sophisticated satire. The point is that it is outlandish and foolish and ridiculous and yet still resembles some serious discourse that real people engage in. So much so that some onlookers can't tell the difference. That proves the satirist's thesis: the real world is full of ridiculous people making ridiculous arguments and they can't even see themselves or their arguments for what they are.
I know real life people who write essays with claims as outlandish as "software engineering is an ableist field" and are dead serious. But that assertion belongs just as well in a satire piece. It can be very hard to tell if you don't have prior context for the author.
> the intent of sophisticated satire
Perhaps they can ask ChatGPT to explain this “sophisticated” satire.
Among others, the voices on the internet include:
* some sincere people with very extreme takes,
* some trolls that masquerade as the above, to bully others over their credulousness and lack of guile, which is distinct from sarcasm,
* some trolls that insincerely speak anything that earns engagement,
* and more and more bots that mimic the above
So sadly, the answer to your question is generally yes.
> Honest question : does it need to be said ?
Yes. There are literally people who believe any thinking done without AI is Luddite and subpar.
These are not, in my experience, people who have done any great thinking in their lives. They do tend to tweet a lot.
Why would you bother telling them it’s satire?
Because we are all learning and changing and no one is a lost cause.
After browsing Twitter for 5 minutes, I’m not sure about that.
Humor died few years ago after series of increasingly more improbable events happening. "USA voted pedophile that bankrupted 3 casinos into second term" was stuff out of Onion decade ago, so some researcher walking into job interview with ChatGPT doesn't even move a needle on satire scale.
Yes because Poe’s Law has never been more true than the current era.
https://en.wikipedia.org/wiki/Poe%27s_law
How wow, it even has a name and a wikipedia article...
Guess that makes it official then. We are all automatons pretending to be flesh :-)
That would be https://en.wikipedia.org/wiki/Solipsism
When I started reading it I thought for far too long that this person is actually stupi^H^H^H^H^H serious. So writing that it is satire is useful for people that do not read enough of the text, and just jump into the comments :) Also, this type of behavior has little to do with the neurodivergent spectrum, if anything it touches personality types, maybe maybe some trauma or addiction, unlikely to be caused by anything in the affective space (episodes would be too short) or psychosis space.
same, i didn't catch the satire and eventually got tired of the stupidity. the thing is, a lot of people around me are using AI like the person in the story. and they are entirely serious. they are not scientists but they could be writing this article and actually mean it.
I was honestly unsure until the Netflix password bit gave it away...
Sadly I have to admit I read it and didn't know it was satire. The way they described using LLMs to to achieve research results sounded plausible enough to me. As an attorney, I also fear a race to the bottom where legal research becomes simply editing the work of AI. Also, the idea of being subject to a test with little relationship to professional practice reminds me of the bar exam; which requires intimate knowledge of almost all areas of the law despite the fact that practicing attorneys generally are only familiar with their specializations (eg, it would be unwise to retain a civil rights attorney to represent you for even a traffic violation); and requires all the knowledge to be memorized, even though it would be unthinkable (and most likely malpractice) for any attorney to practice relying solely on memory.
I also suspect that there are attorneys out there who are starting to use LLMs to generate legal work with a method similar to what the author described (generate prompt, revise minimally). Lawyers are taught to write a certain way which LLMs mirror flawlessly; the surest way to identify LLM generated legal writing is if it happens to hallucinate citations. I used to regularly find hallucinated citations, and occasionally strange reasoning, but both have largely vanished as the LLMS improve.
1) No it does not have to be said. Nobody is forcing anyone to explain anything.
2) I think that the best satire is fair, and should read to the targets as real commentary. When someone takes your satire seriously, it means that you have successfully commented on reality. Is that inverse Poe's Law?
Otherwise you're just putting words in your enemies' mouths that they would not ever say. I think people do this because they don't have the strength of their own convictions, and wouldn't be able to stand people not getting their joke (and worse, confusing them for extreme believers in something that they disdain.)
To them I would suggest that one doesn't try to make arguments through satire, one rather demonstrate arguments through satire. If you need to make your argument with satire, you actually don't have an argument. Satire without a solid underpinning argument is just propaganda.
I think sometimes it is but only if the humor relies on a tone.
When it’s satire I think the main blocker of recognition is if you have an emotional reaction first.
As an example, if you are a diehard AI influencer or something you might miss the joke entirely because of the severity of your initial negative reaction.
Just my two cents. Glad I could contribute to completely beating the humor out of this post :)
No, no, dissecting humor to death is the funnest part of the experience :D
(I, too, am autistic)
i think you may enjoy:
45 Different Types of Humor with Examples in 2026
https://humornama.com/featured/education/types-of-humor/
> Honest question : does it need to be said ?
At the current time I'm writing this, all other top-level comments are engaging with the article as if it were sincere. So, yes.
It does, but unfortunately, this is in-crowd communication -- it's too high brow to hit anyone that needs to read it. But I enjoyed it, and I'm sure it was cathartic to write
As someone on the spectrum, I can assure you that being on the spectrum is not an obstacle for understanding this kind of humor.
In fact, is the neurotypicals who struggle getting it because they rely on nonverbal cues (like a sarcastic tone of voice), which is missing in text, to detect humor.
Deadpan, dry humor is generally more amenable to the autistic mind, because it doesn't have what we consider noise.
If someone needs a laugh track to tell that something is a joke, then either it's a bad joke that wouldn't be made any better with a laugh track, or the problem exists between keyboard and chair.
> Is humor that hard to grasp on the internet ?
Problem is that the Internet is too big that there actually could be someone writing this as a serious blog post. For this one it was quite obvious at some point, but when only reading the first few paragraphs I am very certain that there could be someone out there meaning it serious (sadly).
> Is humor that hard to grasp on the internet ?
Yes. Humor is very hard to grasp via text, especially sarcastic humor, because in person we use voice and body language to signal that something is meant as a joke.
While it is true humor is hard to grasp in text, it has been done that way for millenia in books, and it continued to work as it should. Actually, the fact that it is sometimes misunderstood as non-humor is for the better.
And because we typically know enough about a person to tell if the opinion they expressed is consistent with their believes.
I mean this sincerely, rather than as dismissal: humor is not very hard to grasp in text for people who are fully literate and fluent in the language being used.
The problem we face is that a lot of people are in deep denial about their level of functional literacy.
> Is humor that hard to grasp on the internet ?
It is just a joke was succesfully used as a goto excuse foe years. Usually when people actually have more of extreme stance, test waters and find that "not yet".
And since it was succesfull, many people dont want to play that game anymore or simply dont believe it anymore.
It's hardly too near on reality to not to say it.
do LLMs dream of electric shhhheeep
I got Poe's lawed
On this site? Yes it does, and even then we have people missing the point in the comments.
I regularly get reddit comments downvoted by people who don't recognize it's obviously satire.
Sometimes somebody else comments "why is this getting downvoted, it's satire", and that stops or reverses the influx of downvotes.
It's reddit. The average intelligence of reddit users has been progressing downwards since 2010 or so with major drops in 2016 and 2020. Most users are clueless teenagers who haven't read a book in full throughout their lives.
As often happens with satire, you could read it as satirizing the idea that people should entirely rely on their own knowledge and memory, without AI assistance. The response to the article depends heavily upon one’s perspective on the subject.
exactly!
When it's this poorly executed, yes it needs to be said.
I was going back and forth on whether it was for far longer than I should have been.
> was going back and forth on whether it was for far longer than I should have been.
Please consider that the fault here might be in you, not the text.
> When it's this poorly executed, yes it needs to be said
Isn't that a mark of great satire? That the argument sort of works within the zeitgeist, thereby showing the zeitgeist to be corrupted.
Over-the-top satire is funny. But it's clumsy. The fact that some people who are being satirised will agree with the satire is what makes it great.
It likely is, but it is still uncanny to think about how much of it is actually true.
Is it making fun of people relying on ChatGPT, or is it just an exaggerated description of how she actually does research, honestly I don't know.
That's what makes it good satire.
But if you honestly think that someone applying for a postdoc can't explain what their research is about, it's on you.
Applying for a position is a much lower bar than having any chance of landing the position.
I’ve interviewed someone for a programming position. The guy practically nailed the standard quiz (done at leisure before the interview), but was unable to explain why he answered what he answered. When we got to reviewing a little program (compare two strings, see if the distribution of letters is the same), he could not walk me through it. He was applying for a programming position, and could not explain his "own" impressively correct programming answers.
Well at least postdocs have a PhD, that they had to defend in front of a chalk board I presume. Still, sometimes, frauds do slip through.
>Well at least postdocs have a PhD
That's the point. This is the bar.
>Still, sometimes, frauds do slip through.
You're projecting your experience in IT on academia, and I can't emphasize strongly enough how far off you are. Which was why this piece was written: that reality is already here in software engineering. The entire point is: if this seems absurd in the context of the piece, why would you accept this anywhere else?
In any case, this satire wasn't written from the point of view of a fraud. And if it doesn't seem outright absurd to you, you need to get a reality check.
I hope so, but I work with people with exactly this attitude and approach.
That’s what makes it good satire.
I agree. ChatGPT told me it was most likely satire.
thank you, captain obvious
Your comment, or the post? ;) (I'll see myself out)
Hold on, rational humans downvoted this? Because you can't discern which is satirical and you have no sense of humor? I'll give you the same treatment I give cringe lords in online first-person shooters: take a shower.
One of the better-written ones I've seen. Am sending it around to some professors.
Do note that this was written in December 2025, and we have experienced leaps in capabilities since then. Her methods are outdated. Since January, I've been employed at 3 jobs, fulltime, by using ProxyAI [0]. I'm Chair in several universities. For all I know, I'm also part of several projects.
Don't let this kind of blatant discrimination affect you. The future is bright.
Note: the service is free once you give it access to your bank account.
[0]: https://getproxyai.com/
Edit: downvote from the haters, it's ok this account is managed by my proxy.
> stress free
> fomo nomo
> talk walk
This is golden.
I also like "Terms | Privacy | Terms" at the bottom. It's... that kind of weird error.
The TAM for ProxyAI is even bigger than SpaceX's. Hopefully they IPO soon.
The TAM is infinity, since agents need their own agents, which need their own agents, and so on.
This is about software developers.
Even in a couple of years, I think most devs are still going to be able to write pseudocode for a simple algorithm on a whiteboard, but anything involving real code is going to be difficult without AI. Unless you have side projects or work for a company that doesn't mandate heavy use of AI, you're going to lose your coding skills pretty quickly.
Interviews for devs will change, it'll be far more high-level/architectural, rather than testing coding ability. Live coding tasks will inevitably require prompting AI, simply because that's where things are heading.
Oh, hell! How the tables have pivoted. I was sitting here shaking my head at all the people who didn't realize this postdoc had written a satirical piece and wasn't really insisting that she should be allowed to use ChatGPT at her chalk talk. . . . Chalk talks aren't a thing! Anymore. It's about whiteboarding lmao
Or are they? I don't know what to think anymore.
So it is funny satire.
But my wife is finishing up her PHd and according to her: all students are using ai and pretending they are not because the PI and other older leadership treat any use of ChatGPT or similar as plagiarism and not allowed at any stage. Which I think is simply stupid.
I showed her how to use ChatGPT since Claude was blocking on all her request (biologist) and it has helped her massively improve the wording and structure of her paper as English is not a first language for her.
I see this text as more open ended than a warning. It's a description of a possible future for us to contemplate. Does it have good points does it have bad points? Being satire, the protagonist is a strawman of course, but doesn't she also have some good points? It's not easy to tell where exactly the author stands where, the true argument stops and the satire begins. That's not necessarily a flaw because it doesn't actually really matter where the author stands for our ability to discuss the subject.
Scientist as prompter? Yes it seems fairly likely. But to what extent and how quickly will it happen? Surely scientists will still be able to at least give an outline of "their" work even in the future? Maybe?
Otherwise maybe we it will be a sort of inversion of control, where the language model is the supervisor, and the humans in the loop are more like research assistants doing the dirty work? Instead of a human wearing an exoskeleton, an AI wearing a biological exoskeleton? But this can only be a temporary state of affairs as we won't be needed for that forever.
A scientific project without human involvement? If a paper is published in a journal and no human wrote it and no human read it, is it really science? Does it really advance knowledge. Probably?
>>Instead of a human wearing an exoskeleton, an AI wearing a biological exoskeleton<<
remember this the next time you encounter a request to donate your body.
I didn't mean it in the literal sense more like the AI is telling you do this do this do this do that and you relaying what happened and what you saw over text.
i thought so, but i ran with it.
A term has already been coined for AI driving human: the reverse centaur.
Quite a clever way to take a jab at the cross-coast rival. Well done
I would guess this is a tongue-in-cheek essay (I hope), but it’s probably kind of prescient.
I’m old enough to remember when calculators were banned from classrooms, for almost the exact same reasons that people are giving against AI.
It’s really only a matter of time.
X were banned because you need to learn or demonstrate the ability to do Y.
X = calculator, Y = addition and subtraction
X = LLM, Y = thinking
they are not the same
I don't really follow this comparison. The protagonist was able to use an AI in the same ways they could use a calculator or most tools. You don't use them in the middle of delivering a presentation.
Sure you do (but I’m pretty sure the original was a joke).
Calculators, computers, phones, etc., are standard parts of any design meeting, these days. I interviewed a number of folks, over the years, that brought laptops, and used them to demonstrate their work.
In research science, it's very normal to require zero aids at an interview, and has been for some time. No calculators, no laptops, no phones -- just you.
Yes, you'll also give a seminar with slides to present your prior work, but the whole point of the chalktalk is that it's you, and you alone, presenting your future plans. You're grilled by the faculty on your ideas, and you have to defend yourself without any props or crutches.
No, you don’t. Not in the way the essay portrays it. You don’t ask your calculator “what am I supposed to do with number 2, 2 and funny star (*)”, do you? You have a clear calculation in your mind: multiply two by two.
That's actually a very good catch. Like, we can talk and drive, but we usually can't talk and read something different, one or the other will be seriously hampered.
We can punch in numbers into calculator mostly by mechanical memory but prompting LLMs and reading their output for most people occupies the same lane that's used for talking to other people!
I've never used a calculator in a meeting. And while people will bring a laptop for convenience it's usually to show slides, a thing that can be easily enough done without the laptop. Or to take notes, which is even easier to do by hand. And a smartphone isn't needed at all.
So, really, I don't think the comparison works out.
The calculator is the example that's most directly an important tool that short-circuits something you could do by hand. But again, I've never used one in a meeting and it's not the kind of thing you'd be doing much of in a meeting.
As youse guys pointed out, it’s an imperfect analogy, but the Rubicon has been crossed. The cat is out of the bag. The genie is out of the bottle. The horses have left the barn. The die is cast. Alea iacta est. Ain’t gonna put the candy back in the piñata.
We’ll just have to see what happens.
You're really trying to normalize it hard... But the reality is it's maybe true in some specific niche communities and bubbles. any given time in history different weird things were normal in small communities and bubbles. and time absolutely did not make those things "normal"
> You're really trying to normalize it hard
That's an ... interesting ... interpretation.
All I'm saying, is that "it is what it is."
Regardless of what we think of it, it's coming. It's likely that no one actually knows what things will look like; just a few years down the road.
The only surety that we have, is that it's coming; whether or not we believe in it.
Acceptance is the last -but most important- stage.
and it is how?
you replied to satire/dystopian fiction. it is by definition not what it is. and it doesn't have to be.
by saying oh it's inevitable believe me, "the only surety that we have" (casually inserting we like everyone is the same on that) etc. you are trying to normalize it, painting anyone not on board as not part of "us", outcast, old fashioned and anti progress.
often when people do this sort of attempt at manipulation they have vested interest. because if public opinion shifts sufficiently anti chatbots etc, which it is doing currently, and the government responds like a democracy would (or people take matters into their own hands), they lose actual money.
Have a great day!
thanks, certainly will;)
> It is, in other words, a ritual designed in 1974 and never updated.
The Chalk Talk was not invented after the IBM System/360. This entire article is clearly a joke.
I hope people understand that this is a satire. :-)
Reminds me of a sub plot from Accelerando. Wasn't a huge fan of the book but it left an impression.
She had me for the first 300 milliseconds. Very funny.
> When you ask me to explain my research without ChatGPT, you are asking me to speak on behalf of a collaborator who is not in the room. Would you ask a PI to give a talk without allowing them to mention the work of their postdocs?
PI or that postdoc should be able to explain the project as such, without mentioning who exactly did what.
They must talk about colleagues when attributing to protect and create their careers.
academia will (hopefully) be an industry that benefits from, rather than degraded or obsoleted by, ai. the system should self-correct. scientists that use ai to give themselves more time to think will succeed and scientists that use ai to avoid thinking will fail. over the next few years processes and pedagogies should develop that help reliably guide researchers toward success.
there's a lot of tedium in academia. ai taking menial tasks off the plates of grad students and post docs allows them to spend more time collaborating with, instead of serving, the head of the lab. dissertation defense is already a strong mitigation against producing folks like that satirical Dr. Simmons, you can't defend a dissertation if you don't have the chops for a chalk talk.
Loved it!
"The Evolution of Human Science"
I mean... this has to be satirical, right? Surely nobody in the field of academia is this limited in their ability....
Sounds like you should feed the article into ChatGPT to find out whether it is satire.
> this has to be satirical, right?
Right.
> Surely nobody in the field of academia is this limited in their ability....
Wrong.
Ah yes, and I wasn’t allowed a google search during my whiteboard interview, which was totally also discrimination.
It is certainly not that case that the point of these exercises is to examine how you, as a person think and approach problems.
Strawman takes a hit AND IT GOES DOWN!!
It's actually embarrassing that she thought all that prompting would have been acceptable during the talk.
A qualifed researcher would have had their agent perform the talk on their behalf rather than waste everyone's time.
<< << My own words? I haven’t used my own words since 2022. I’m not even sure I have my own words anymore.
I thought about you wrote and I think you are right. Even though I am partial to author's POV ( despite some obvious, glaring issues ), I can't help feeling of two minds about it all. I don't want to undermine existing ecosystem, but then.. the existing ecosystem in research may need some decent shake up.
It's a joke.
So is my comment.
[flagged]
[dead]
Calling it "Discrimination" is obviously absurd but if the process produces useful results, one ought to seriously consider whether it might be worth switching.
I understand we have always conventionally transported goods by horse. Yes, this employee knows nothing about horses, and in fact is rather spooked by them, but we've checked! Their claim to be able to transport goods faster, without a horse, somehow seems to hold up.
Maybe, just maybe, we should take this whole "truck" concept seriously?
I don't think the argument is necessarily against the use of the tools entirely. My interpretation is that it's against delegating to them all understanding.
Humans can only usefully steer LLMs if they have some understanding or context the LLMs do not.
Put another way: if all that you're doing is prompting the AI and giving me the result then I have no use for you. If you're not contributing insight, understanding, experience, or creativity then it's far cheaper for me to prompt the AI myself.
The person in the story can clearly usefully steer LLMs if they got that far. So either they have that understanding and context, or it's not actually as valuable as was previously thought.
She says she can usefully steer LLMs, but how would she know? What knowledge of the research topic, methods and their applicability would she use to judge that, if all she’s doing is writing a prompt asking ChatGPT to write a summary of her research topic, methods and their applicability?
Supposing using LLMs as research tools is fine, the chalk talk is still necessary to test that the researcher has the fundamental skills and knowledge necessary to discriminate usefully between alternative responses, and refine prompts and such. You can’t just get a response from an LLM and then refine it for your specific needs by saying “now refine that reply for my specific needs”. You need to know and understand what those are, and why.
Line one is "I recently interviewed for a tenure-track position" which suggests she has a resume that ... actually qualifies for an interview? And did every single step up until now correctly?
Like, I'm sorry, you can't just walk into those off the street. Don't be absurd. I suppose you could say the entire system was happy to spend 4 years lying to her about her grades? But at that point it's "corrupt system completely fails a young student who was enthralled by new technology"
> test that the researcher has the fundamental skills and knowledge necessary to discriminate usefully between alternative responses
... and how, exactly, is she gonna demonstrate that if she can't prompt it in the first place?
> “now refine that reply for my specific needs”
That absolutely works if you've discussed your specific needs with the model for four years and it has dozens of memories and reference files to work from.
She can’t answer a single thing without opening her laptop.
Yeah, and the deaf lady can't answer a single thing without her interpreter. Again, what's your point? Is this a field where people are expected to work without a laptop for some reason?
Like, I get why we'd want a pilot not to need ChatGPT, but a scientist is not in any of those categories.
> Yeah, and the deaf lady can't answer a single thing without her interpreter.
Deaf lady is missing her hearing and therefore requires hearing aids to hear, so following your logic Rachel is missing her thinking therefore she requires ChatGPT to do thinking for her. Glad we're on the same page.
Found Rachel. Thinking and doing things with this archaic thing called “brain” is not a horse, and using LLM is not a truck.
Do you do all your arithmetic by hand, or do you insist on using this archaic thing called "brain"?
Why are you even on the internet if you value this "brain" thing so much?
> Do you do all your arithmetic by hand, or do you insist on using this archaic thing called "brain"?
I don't need calculator to know that arithmetic exists and what it is.
"prepared to do what I do every single day in my actual scientific practice: type a prompt and receive a coherent, well-structured response that I would then lightly edit and present as my own thinking."
So, plagiarism. Daily.
This is satire
Replace science with software development and this just reads like half of HN right now.
My thought process in engaging with the post was:
Reading the title: I hope this is a joke.
First paragraph in: Oh, good, it is a joke.
75% through: Ok, this is a great bit of satire, but this reads uncomfortably like many of the non-joke discussions I've seen around software development over the past year.
[dead]
despite being satire this is no joke, this is what we are headed into if we dont stop the decline.
I think that's what the satire is trying to warn against.
But when humans do it suddenly it's "standing on the shoulders of giants"
I don't get how you can possibly call it plagiarism if it produce a novel breakthrough - by definition, the existing knowledge base doesn't contain the new ideas generated in this process.
And we've proven it can handle complex, novel thinking when it solved a significant Erdos problem back in May: https://www.scientificamerican.com/article/ai-just-solved-an...
The difference is thinking for yourself
Sorry, what about solving an open mathematics problem is not "thinking for yourself"? What exactly WOULD constitute "thinking for yourself", or is this just a magic soul thing?
you read some books and think hard = not plagiarism
you press a button and a machine creates a derivative work from some books = plagiarism (mostly by people who made the machine and charge you for it, maybe arguably you but that's a different question)
I don't have a problem using AI.
I also acknowledge, when doing so, how I use it.
But, literally typing in a prompt and then massaging its output and then claiming ownership of its plan? Yeah, no.
It doesn't matter if the original article is supposed to be satire. I've been using AI to do some coding, I've been using it to help me plan a much larger project. It's a tool, and it's useful; but I am in control of the project, not letting the AI control me.
Meanwhile in my day job, I've had multiple coworkers pump shit into AI and regurgitate the answer without critical thought. They are letting it make decisions for them without validating it first even.
It's not really satire if we're already dealing with this attitude.
What’s going on with social queues today where so many people here are not immediately understanding this is clearly satire?
I think this thread is a good indicator of llm bots vs human accounts on HN. The bots couldn't identify satire.
Comedy is dead. I can't believe so many folks didn't clock this post.
(Also, the word is "cue", as in "signal". English sucks.)
But honestly, I think it makes sense. In my experience with programming, when it comes to building and delivering software, it's not really about memorization. Honestly, search ability has been far more important. Wouldn't it be fair to think of AI as just another search ability?
What I'm curious about is this: in the end, experts are the ones who are best at distinguishing hallucinations. If you can just search with GPT and tell the difference, wouldn't that be enough? I can't imagine memorizing thousands or tens of thousands of lines.
I have ADHD, and when I get nervous, I tend to forget what I was going to say, so it's even more true for me.
> Wouldn't it be fair to think of AI as just another search ability?
If you're asking it questions, yes. It's like search but with a simulation of understanding and information synthesis far faster than a human can perform it.
If you're having it write your code, no. It's like a junior developer who has no awareness of the bigger picture, of incompatibilities, of understanding that hasn't been contained and can't be derived from the codebase.
> If you can just search with GPT and tell the difference, wouldn't that be enough?
The situation the satire is describing is an individual who is unable to tell the difference. The way the scenario is laid out, everything she's 'accomplished' has been to prompt ChatGPT and publish its answers with some degree of editing; it's clear that she, as an individual, is not an expert, does not understand the thing she is presenting, and does not know any of what she has purported to know. This is also a sadly common refrain these days.
> I can't imagine memorizing thousands or tens of thousands of lines.
It's not about memorizing thousands of lines of text; it's about demonstrating to the panel that you have an understanding of the thing you're claiming to have an understanding of.
I work with a lot of software and infrastructure at work. I can tell you how it all (or most of it) works together and interacts. I could not reproduce the configurations of any of the software from memory, nor recite any of the code, but I have an understanding of the system, how it works, what it was designed for, and what choices were made and why.
The professor in this article does not have any of that understanding. It would be as if I had Claude deploy a cluster of X, Y, Z components, configure them, and get them online, and then put on my resume that I had done it. It was accomplished as result of me, but if I don't understand the system then there's no difference between me doing it and the CEO doing it, or my son, or someone from Taskrabbit.
So yeah, it's not about memorization, it's about understanding.
I think differently. The reason is our intuitive understanding of what it means to 'understand.' To be explicit, purpose and design always depend on scale. And software engineering operates on trust, at least at its foundation. Do you know every implementation detail of the libraries and frameworks you use? If you do, then I have nothing to say.
There are several things I don't understand.
How many people actually understand the internal workings of std::vector or React? Honestly, people use them just fine without knowing. And on the flip side, those who do know sometimes struggle with architecture design. Why? Because of 'cognitive resource limits.'
What I really don't understand is this: if everyone truly understood everything and could connect everything properly, then realistically, LLMs shouldn't have made the discoveries they have. The reason is simple—they're just encyclopedias at their core. But LLMs do make new discoveries, by applying knowledge from one domain to another.
Ultimately, it's a matter of what 'understanding' means and how deep we can go. I think we need to reconsider how we verify knowledge production systems in a society that produces at massive scale.
What I find hard to grasp about your notion of 'understanding' and 'knowing structural constraint failure modes' is this: can you actually list all possible failure conditions? The largest system I've worked on was a Chinese booking service with about 330,000 lines of code. There were so many failure conditions there—starting from external I/O and errors everywhere—that it was impossible to classify them all.
In the end, what matters is the abstraction boundary and failure modes that match the level you're claiming. And I think that's entirely valid. Do you have a counterargument?
I actually think a truly great developer is one who can limit the blast radius when a black box explodes. In that sense, I don't think the idea of knowing a black box empirically is wrong at all.
Not every developer can know everything about compilers, kernels, TLS, database engines, React, libc, or CPU memory models.
What I find difficult about this satire is this:
We're already standing on top of black boxes. We confidently use libraries in our projects, trusting the author's documentation. Is that really 'understanding'? I think it's memorization. We 'memorize' that something works and put it into our project. That's why I don't find it funny as a satire.
Right now, our knowledge transfer systems only operate within the clear cognitive limits of human understanding. But as things get deeper, something that goes beyond those limits will eventually emerge. At that point, can we really define that behavior as wrong?
No matter how detailed a map you have, it's still not the actual terrain. In the process of abstraction, we inevitably lose information. But how can we know exactly how much was lost? And even if we conceptually guess, I wonder if that will still be possible in future eras
You can bring notes to a chalk talk.