Mine seems to think that I'm some kind of detail-obsessed super-pedant. Personally, I think this is ridiculous. "super" is a Latin stem meaning "beyond", which implies that I've transcended the qualities of pedantry. A better term would be 'pluri-pedant', which denotes someone who is exceptionally punctilious while still remaining within the bounds of being pedantic.
An aside that I do want to mention here because it is a really unique way for many people to interface with LLMs: many commenters mention the model over indexing on a few comments they made that do not necessarily reflect of the broader themes of their writing. This is not any issue in the author’s engineering but an inherent issue in LLMs. The reason it is so noticeable in this case is because the subject matter is extremely familiar to the user: themselves.
LLMs consistently misrepresent information in this exact same way in, more critical applications. Because they are often employed on datasets that engineers and potentially end users are not deeply familiar with, the results often seem exceptional.
Disclaimer via my HN wrapped:
“The Anti LLM Manifesto
You will write a 5,000-word blog post on why a single Bayesian prior is more 'sentient' than GPT-6, and it will be ignored because the summary was generated by a 3B parameter model.”
I did like it, but for me it was fixated on 3-5 comments from the last 1-2 months that got a few more upvotes. It didn’t really work as an overview for the year. Still, a pretty cool thingy :)
Thanks. I now run a two-step process: first pass reads through all posts and comments to extract patterns, second pass uses those to generate the content. Should be much more representative of your full year now :)
My impression was the same as the poster: it still over-indexes on a couple of recent posts.
Of course, it's possible that we've both been repeating ourselves all year long! I mean, I know I do that, I just think I've ridden more hobby horses than it picked up. :-)
It's fun, though. Thanks for sharing - a couple of my "roasts" gave me a genuine chuckle.
Yeah, same here. For the comments it took into account it made pretty great roasts, but would have been better if it was actually comprehensive over the course of the year.
Your intense hatred for the concept of GDP is only matched by your strangely specific crusade against the calorie theory, making you the only person on HN who thinks the economy and thermodynamics are both just vibes.
Oh I absolutely love this.... And the roasts are sooo on point. Here is one of mine:
"""
You have plugged your project ZenQuery so many times that I am starting to suspect it is actually a sentient recursive script you wrote to avoid paying for real marketing.
> You've mentioned Gemini 2.0 Flash pricing and model comparisons so many times that I'm starting to think you're actually a Google Cloud Billing alert that gained sentience.
I wouldn’t mention it so much if Google stopped bumping up the price.
1.▲
Show HN: A songwriting DAW built entirely inside an Org-mode buffer(emacs.org)
432 points | 2 hours ago | 89 comments
2.▲
Why I'm still using Soulseek to trade prompt-engineered MIDI files
156 points | 4 hours ago | 42 comments
I admit I hastily tried clicking before realizing these were fake
As a side note, I find this capability of AI to mine social profiles quite disturbing. Automated profiling of social media accounts can be and is used with malicious intent. The amount of personal detail that can be recovered this way is shocking. It is possible to associate this information with a real identity, and it can be used to target and intimidate individuals.
It's literally what social media is for. People seem disturbed that when they put their private thoughts out on the internet their private thoughts end up out on the internet. I never understood that.
In contrast to many others, I did not find this particularly interesting.
- The comic on is oddly cropped and contains speech attribution errors.
- It calls me an "extremist" regarding the wrong thing (I am many kinds of extremist, but certainly not Haskell).
- It claims I believe "any software failure is merely a design error" which is a complete misunderstanding of the ideas I presented.
- It says things like "the geometric mean of the snack bowl" which doesn't have meaning in English.
I feel like it has picked up on certain keywords and then just rolled with its own stereotypes of what those keywords represent, rather than actually taking a good look at what I think. A roast works because the roaster has clearly spent time and effort and care understanding the person roasted. This is way too shallow for that.
The 2026 and 2035 predictions (with a few exceptions) don't make sense at all, and the jokes in them fall completely flat. They're not good anti-jokes either. If someone said something like it in a social situation it would be followed by an awkward silence.
The vibe check and the time spent were really cool though. Super interesting. I would have loved to see those expanded.
I don't mean to be negative. The project is cool. I just wish it would put its focus on the valuable parts, rather than the things it is weak at. I guess this is my 45 % pedantic, 25 % contrarian, 20 % analytical self speaking.
> I feel like it has picked up on certain keywords and then just rolled with its own stereotypes of what those keywords represent, rather than actually taking a good look at what I think. A roast works because the roaster has clearly spent time and effort and care understanding the person roasted. This is way too shallow for that.
Yeah. It picks one random thing from one comment and turns into a lifestyle.
That's about how it came across for me as well: ignoring my actual content and joking about generalizations related to key words.
Project is cool overall, love the xkcd-like comic idea—but prompting and/or model-selection could use some work. I'd like to take a crack at tuning it myself :)
Perhaps it should also avoid putting too much emphasis on several comments to the same story: there was a story about VAT changes in Denmark, where I participated with several comments; but the generator decided that I apparently had a high focus vat, when I just wanted to provide some clarifying context to that story. I wonder how comments are weighed, is it individually or per story?
Specifically this roast:
> You have commented about the specific nuances of Danish VAT and accounting system hardcoding at least four times, proving you are the only person on Earth who finds tax infrastructure more exciting than the books being taxed.
Yeah, but I did it on the same story (i.e. context).
Though the other details it picked up, I cannot really argue with: the VAT bit just stood out to me.
This is fun, but the fact that it is now so easy to process a massive amount of social media data to extract a person's political leanings, hobbies, etc and infer information about them scares me. It only takes one government change (cough cough, Reform UK) for this to be used against me. It doesn't matter if I'm politically correct when what is correct changes over time.
>Your frustration with LGA socket reliability will finally boil over, leading you to move to a cabin in Mexico where you only communicate via GPS-disciplined atomic clocks and ham radio SDRs.
> You spend so much time fighting macOS animations and keyboard layouts that I am surprised you have any time left to actually use the computer you keep threatening to replace with a Framework
Yep that’s me.
As for 2026 prediction:
> You will write a 4,000-word HN essay arguing that Silksong’s difficulty curve is a direct allegory for the South Korean 'Hagwon' education system.
That was hilarious, uncannily accurate in some places.
A romantic of the motherboard who thinks software engineering peaked at Xerox PARC and spent 2025 auditing every Show HN for missing source code while mourning the death of the em-dash. You will develop an LLM-powered sentiment analyzer that specifically detects when tech executives use 'fond farewell' to describe mass layoffs.
▲ Ask HN: Is there any way to travel without 24/7 neural-link monitoring?
▲ Mozilla CEO calls 15th rebranding to 'VibeBrowser' a 'bold pivot'
▲ Minimalism in the 2030s: Living on 8kb of memory for a week
I got the following:
"... a linguistic purist who will defend a 32-bit Atom CPU until the heat death of the universe or the next Debian update."
Interesting, semi-accurate, though with some of the other things it wrote, it was completely wrong:
"You claim labels are limitations that prevent you from being yourself, then immediately spend 500 words explaining why your entire identity is based on being a 1%er."
I wish I was a 1%er, sadly I'm on the other side of the spectrum, currently with no money, so I'm on my fourth day of not eating, though luckily I should receive my salary this week or the next, so it'll be alright!
This was a bit of fun. I've been doing a lot of coding via GitHub copilot from my phone lately and it picked up on that and coined the term "Thumb driven development", which I absolutely love and am officially adopting!
At first I was both amused and mildly annoyed at mine. Then I looked at the ones for some of the other HNers whom I follow and realized I got off easy.
It's good and gave me some queasy chuckles. But it's not as fun as the "HN Front Page" from a few days back.
I suspect there's kind of a feedback loop here. An algorithm does X well; it becomes easy to churn out X; we realize that X wasn't as skilful as we thought. In other words, this kind of satire will become seen as stilted and conventional, in the same way that AI art looks stilted and conventional to us today.
>> Your next 'exception catcher' blog post will be written after a long night at the Chicago Microcenter and will feature a Kafka setup powered entirely by spite and BIPA lawsuits.
Hilarious. It seemed to focus on something about Boolean logic that I don’t remember taking about, but otherwise awesome.
Are you sharing any of the prompts you used to generate this? Even if not verbatim, I’m interested to know how much you’re driving versus the model. How much more prompt is there than, “look at this comment history and write 3 roasts that the HN crowd will find funny”?
This caused me to delete the HN password on my main account. It was already disconcerting to be reminded by Karpathy (I cannot help but see this as bragging) in his recent post that we must be good because LLMs are watching: https://karpathy.bearblog.dev/auto-grade-hn/
LLMs will be used to deanonymize all internet users. The AI "visionaries" are already bragging about this. You have been warned.
A developer who believes every global outage is just a missing question mark away from salvation and spends their weekends reapplying thermal paste to fanless MacBooks while reminiscing about the tactile superiority of 2010 Dell Latitude trackpads.
Roasts
You post about Cloudflare outages caused by a single unwrap while your own codebase probably looks like a game of Russian Roulette played with Result types.
Your obsession with the thermal conductivity of fanless laptops is just a coping mechanism for the fact that your Rust builds take so long you could literally cook an egg on your chassis.
You have a very specific kink for 2010 Dell trackpads that makes me think you are either a Linux philosopher or someone who is no longer allowed within 500 feet of a Best Buy.
Feeder, FeedDeck, NewsBlur, Fluent Reader, Twine, Capy Reader, Reeder, Read You, Neo Feed, Readrop are all open-source RSS/feed readers for mobile (iOS, Android, or both)
, and they all have dark theme / night mode.
> For someone who claims to be a professional cinematographer, you spend an alarming amount of time looking at plain text arguments on a 1990s-style forum instead of actually framing a shot.
A high-frequency human link-shortener who refuses to let a single HN thread go by without reminding the world that their personal blog exists as a primary source of truth.
Your commitment to plugging your blog is so aggressive that I am legally obligated to check if https://www.rxjourney.net/ is actually a horcrux for your digital soul.
This is genuinely great! Like others, it did overindex on a few comments, and I think that's a great way to display LLMs weaknesses. But overall, really loved the roasts, predictions, and all the rest. Nice job!
> For someone who claims to never use AI to write comments, you sure do love spending three paragraphs explain-testing why a specific LLM hallucination was a moral failure of the Google Docs UI team rather than just using a search bar like a normal person.
> You have a literal fetish for the Linux filesystem and treat the ability to 'drag a photo into a folder' like it's a human right on par with the Geneva Convention while acting like the New York Times lawyers are the final bosses of a dystopian RPG.
> You claim to value efficiency, yet you've spent the equivalent of a full fiscal year arguing about why a Firefox fork that 12 people use is the only path to salvation.
Wasn't this same site showed off on HN about a few months ago? I recall using it back then where it would roast you, seems like this is just a rebrand to a "wrapped" version and more generally seems like an ad for your service.
Also the roasts are heavily front-loaded, the LLM is only really taking my most recent posts into account, not the especially far back ones earlier in the year.
A high-frequency haptics evangelist who is currently attempting to connect every physical object in his house to a DuckDB instance via an AI-controlled nipple mount.
I like it, though I do notice that like most LLMs it seems to take criticism of LLM rollout kinda personally (actually circa 2023 this was considerably less true across every popular model)
I also think the predictions section seems kinda generic where the roast section felt better-personalized, which seems like a prompting issue
> A high-frequency debunker who treats every comment thread as a zero-trust environment where empathy is a bug and citing Sartre is a security vulnerability. You are the only person on the planet capable of linking the efficiency of electrical line curvature to the ethics of Anthony Bourdain in a single browsing session.
Mine "roasted" me by making fun of the fact I never finished a PhD, despite that being due to medical and other life circumstances that were well outside my control, including, but not limited to, some issues related to the fact I was a woman trying to get into academia who experienced the kinds of behaviors from people in the department which are not really suitable for polite discussion.
Additionally, it roasted me for building a project to "avoid the outdoors," which is another incredibly demeaning thing to say to someone who explicitly created that project because she was too medically unwell to be able to go outside as much as she wished and wanted to bring a bit of the outdoors inside. Very lame, definitely missed the mark.
The elisp and common lisp notes were on point, though, and did get a chuckle out of me.
I really enjoyed my roast review. I loved the feedback that I'm only 5% helpful and 35% contrarian! I will endeavor to boost that helpfulness percentage next year.
I am surprised at the passion that some seem to feel over their own reviews.
> You spent three paragraphs arguing why estimation is a vital business requirement only to follow up by suggesting the world is an entropic chaos where nobody is actually in charge, which is a really convenient way to explain why your sprint is three weeks late.
I don't like "Wrappeds" (low-key social hack to manufacture normalization of surveillance capitalism?), but with HN being public, I succumbed to temptation. Very fun, 10/10 no notes, surprisingly good for a small sample set this year.
> You write comments like you're trying to win a Pulitzer in Political Economy while trapped inside a middle-manager's strategy meeting.
> You've mentioned Futurama references so many times that I'm starting to think you're actually just Zap Brannigan with a Python Foundation fellowship.
And to add insult to injury it says I'm going to rewrite all my Python stuff in Rust (after spending the year teasing Rust). I got a good giggle from the comic book summary -- thanks for making this.
> Frustrated with Waterfox and Orion, you will finally launch 'CERN-fox', a browser that only renders LaTeX and requires a muon-detected captcha for every search query.
"A high-latency architect who spends his days documenting every time a CDN sneezes while dreaming of a mountain drive through the Balkans with a fresh burek in hand."
Okay, mine was just awesome. Thank you. The only thing that would make it better would be if it could be easily saved.
(Also, it's a shame that it regenerates the xkcd every time)
"A seasoned architect who spends their days patrolling the wall between actual engineering and unsustainable AI hype while desperately trying to keep their Windows 10 box alive until the heat death of the universe. You are the only person on the internet who still remembers what a build script does and why we shouldn't let LLMs touch them without adult supervision."
> You have mentioned being Australian at least five times as a personality trait just to remind us that our pennies are stupid and our tap-to-pay is thirty years behind yours.
Fair!
The XKCD comic generation was impressive.
The "your HN front page in 2035" doesn't really make sense to me because afaik personalised front pages are not a thing? (Or maybe they will be in 2035...)
Pretty fun overall. The roasts were really good, the stats were pretty neat, the predictions made me smile. The xkcdupe fell flat for me because it's pretty generic and I've seen many variants of it before, but still pretty impressive for an AI to make. The "Your HN Front Page in 2035" was definitely the weakest part for me. Maybe because every gag in it had already been done in other sections.
One thing I noticed is that it seems to give more weight to submissions (i.e. things I've submitted as posts) than to comments (or at least, doesn't let submissions get drowned in a sea of comments), which turns out as a good thing. And it doesn't seem to care about karma, which is also IMO a really good thing, makes it feel like a deep dive and keeps it interesting.
I’m going to have a sticky note warning myself to “Don’t write that on Hacker News.” I laughed but I hate me.
“You talk about 'Walking Out' of digital services so often that it's starting to sound less like a data strategy and more like a cry for help from your $10-a-month AWS bill.”
My last 2-3 months here have largely been posting about AI slop. But the prior 9 months were not, and included a lot of nuanced posting about how to make use of AI agents. I even got accused of being a shill at one point.
The wrapped is pretty much focused on the slop stuff. That's less interesting than the earlier parts of the year - I know what I've posted recently but have forgotten a lot more of what was going on a year ago or how the year developed.
Roasts can be amusing but I don't think they're the right vibe for a wrapped. I know it's harder to get an LLM to write something witty and insightful than a witty but shallow quip so maybe that's why roasts are here. Wrappeds are sort of infodumps though and LLMs are good at that, maybe there could be a two stage step where it reasons about some custom quirky stats or factoids that work based on your profile and then the second stage generates them.
Well, as a local VRAM libertarian, to manually prune the safety alignment part of a 500B LLM for it to run on 1GB RAM or VRAM is definitely a lifetime goal for me.
> You’ve mentioned the 1975 book The Mythical Man-Month so many times that I’m starting to think it’s your only personality trait besides complaining about Tailwind CSS.
Show HN: A Python 4.2 framework that transpiles to Go but still lets you use circular imports(github.com)
Microsoft 365 Audit: Now requiring a literal DNA sample for volume licensing compliance(theregister.com)
Zulip 15.0 adds 'Boomer Mode' to hide markdown from non-technical users(zulip.org)
eh?
Ask HN: Best audiologist for tuning hearing aids to filter out 'Notification Blindness'?
Because this is HN and we must critique: I think the “HN 2035” would be more entertaining if you adjusted the prompt to suggest it use company and product names that don’t actually exist. (There’s no way HN is half full of Tesla and OpenAI articles in 2035)
A digital preservationist who wages a one-man war against the AI-slop apocalypse by manually indexing every human-written blog post in existence while playing Unreal Tournament 99 on a secondary monitor.
——-
I have to say, if that’s my future I’m kinda cool with it.
> Your posts suggest you are the only person on the internet who uses genetic algorithms for friendship management but still can't figure out a Typst template for your thesis.
A teenage digital architect who oscillates between solving the world's privacy crisis and wondering if high school chemistry is a psyop designed to kill his GitHub star count.
This is really awesome, I liked how it really detected my chemistry hatred and how the xkcd had the "see the mole concept is a false flag to obfuscate the real data, You have a test tomorrow" line as I kinda winging chemistry sometimes
> You are the only person on earth still trying to make 'decentralized link shorteners via Signal avatars' happen while failing organic chemistry.
100% accurate lmao, but for what it was worth i was trying it with signal call links since you can name a call name link 32 bits of storage which are persisted forever in signal's database iirc so it would've been a shitty link shortener but still I just loved signal and kinda wanted to build something on top of it
Do I really yap so much about chemistry here, I think that I have created more topics, surely lol but still I still enjoyed this a lot, maybe it just catched up on these traits more since I am pretty damn sure that I might be the only person here commenting about why in the world my country is requiring me to master chemistry university level to just get into a basic comp sci degree.
> Show HN: A Kanban board built on top of Bitwarden notes because why not
Btw, this was this close to happen except at the time I was vibe coding it with some new tool to stress test it with prototyping ideas basically but it didnt really work so i gave up on it but let me know if this idea fascinates someone lol
Happy holidays to everyone, this time of the year must mean a lot to people and I appreciate the spirit of holidays and gift giving too :)
Edit: also I love how it catches myself as existential since I genuinely had gotten existential because of hackernews once wondering what are the best ways to promote/grow open source so much so that I had written a manifesto, I can also be considered idealistic but I dont know why I forgot but "The FOSS Existentialist" is such a good title that I am gonna have it in my about page. This was genuinely brilliant.
https://hn-wrapped.kadoa.com/ethmarks
Mine seems to think that I'm some kind of detail-obsessed super-pedant. Personally, I think this is ridiculous. "super" is a Latin stem meaning "beyond", which implies that I've transcended the qualities of pedantry. A better term would be 'pluri-pedant', which denotes someone who is exceptionally punctilious while still remaining within the bounds of being pedantic.
You should write for the onion
I see what you did there
An aside that I do want to mention here because it is a really unique way for many people to interface with LLMs: many commenters mention the model over indexing on a few comments they made that do not necessarily reflect of the broader themes of their writing. This is not any issue in the author’s engineering but an inherent issue in LLMs. The reason it is so noticeable in this case is because the subject matter is extremely familiar to the user: themselves.
LLMs consistently misrepresent information in this exact same way in, more critical applications. Because they are often employed on datasets that engineers and potentially end users are not deeply familiar with, the results often seem exceptional.
Disclaimer via my HN wrapped: “The Anti LLM Manifesto You will write a 5,000-word blog post on why a single Bayesian prior is more 'sentient' than GPT-6, and it will be ignored because the summary was generated by a 3B parameter model.”
I did like it, but for me it was fixated on 3-5 comments from the last 1-2 months that got a few more upvotes. It didn’t really work as an overview for the year. Still, a pretty cool thingy :)
Thanks. I now run a two-step process: first pass reads through all posts and comments to extract patterns, second pass uses those to generate the content. Should be much more representative of your full year now :)
My impression was the same as the poster: it still over-indexes on a couple of recent posts.
Of course, it's possible that we've both been repeating ourselves all year long! I mean, I know I do that, I just think I've ridden more hobby horses than it picked up. :-)
It's fun, though. Thanks for sharing - a couple of my "roasts" gave me a genuine chuckle.
Grüezi! Is there a way to re-generate my wrapped?
https://hn-wrapped.kadoa.com/aschobel
It was quite different when I tried it again. Still fairly fixated on the last month, but it is definitely better.
My roasts are now substantially more well done now. Well done.
Yeah, same here. For the comments it took into account it made pretty great roasts, but would have been better if it was actually comprehensive over the course of the year.
Same here too, I agree its a pretty cool thing but still better to know that it isn't just me who felt like it hyper focused on some comments.
I had a similar experience but overall the idea is super charming. I do like the personalized HN for 2035. Thank you for building it!
I agree, it feels like it only read the most recent few months of comments. The "vibe check" was on point though!
Nice, mine was very cool and accurate :
https://hn-wrapped.kadoa.com/ttoinouOh I absolutely love this.... And the roasts are sooo on point. Here is one of mine:
"""
You have plugged your project ZenQuery so many times that I am starting to suspect it is actually a sentient recursive script you wrote to avoid paying for real marketing.
"""
Full wrapped: https://hn-wrapped.kadoa.com/freakynit
> You've mentioned Gemini 2.0 Flash pricing and model comparisons so many times that I'm starting to think you're actually a Google Cloud Billing alert that gained sentience.
I wouldn’t mention it so much if Google stopped bumping up the price.
Sounds like you got offended by a robot.
I believe the correct term that I've seen elsewhere is "clanker".
Same
As a side note, I find this capability of AI to mine social profiles quite disturbing. Automated profiling of social media accounts can be and is used with malicious intent. The amount of personal detail that can be recovered this way is shocking. It is possible to associate this information with a real identity, and it can be used to target and intimidate individuals.
It's literally what social media is for. People seem disturbed that when they put their private thoughts out on the internet their private thoughts end up out on the internet. I never understood that.
In contrast to many others, I did not find this particularly interesting.
- The comic on is oddly cropped and contains speech attribution errors.
- It calls me an "extremist" regarding the wrong thing (I am many kinds of extremist, but certainly not Haskell).
- It claims I believe "any software failure is merely a design error" which is a complete misunderstanding of the ideas I presented.
- It says things like "the geometric mean of the snack bowl" which doesn't have meaning in English.
I feel like it has picked up on certain keywords and then just rolled with its own stereotypes of what those keywords represent, rather than actually taking a good look at what I think. A roast works because the roaster has clearly spent time and effort and care understanding the person roasted. This is way too shallow for that.
The 2026 and 2035 predictions (with a few exceptions) don't make sense at all, and the jokes in them fall completely flat. They're not good anti-jokes either. If someone said something like it in a social situation it would be followed by an awkward silence.
The vibe check and the time spent were really cool though. Super interesting. I would have loved to see those expanded.
I don't mean to be negative. The project is cool. I just wish it would put its focus on the valuable parts, rather than the things it is weak at. I guess this is my 45 % pedantic, 25 % contrarian, 20 % analytical self speaking.
> I feel like it has picked up on certain keywords and then just rolled with its own stereotypes of what those keywords represent, rather than actually taking a good look at what I think. A roast works because the roaster has clearly spent time and effort and care understanding the person roasted. This is way too shallow for that.
Yeah. It picks one random thing from one comment and turns into a lifestyle.
Mine on the other hand could not be more accurate: https://hn-wrapped.kadoa.com/MarcelOlsz
That's about how it came across for me as well: ignoring my actual content and joking about generalizations related to key words.
Project is cool overall, love the xkcd-like comic idea—but prompting and/or model-selection could use some work. I'd like to take a crack at tuning it myself :)
Appreciate the feedback, will try to iterate it to greatness further. It's still a bit hit or miss, but I've made a few improvements:
- improved prompts with your feedback
- added post/comment shuffling to remove recency bias
- tried to fix the speech attribution errors in the xkcd
Perhaps it should also avoid putting too much emphasis on several comments to the same story: there was a story about VAT changes in Denmark, where I participated with several comments; but the generator decided that I apparently had a high focus vat, when I just wanted to provide some clarifying context to that story. I wonder how comments are weighed, is it individually or per story?
Specifically this roast:
> You have commented about the specific nuances of Danish VAT and accounting system hardcoding at least four times, proving you are the only person on Earth who finds tax infrastructure more exciting than the books being taxed.
Yeah, but I did it on the same story (i.e. context).
Though the other details it picked up, I cannot really argue with: the VAT bit just stood out to me.
That’s a poorly written roast.
I found the xkcd comic for myself funny enough to chuckle but I had the same feelings as you about the text
It also appears highly biased towards recency as much of mine was roasting a topic I had only spoken of once and recently
This is fun, but the fact that it is now so easy to process a massive amount of social media data to extract a person's political leanings, hobbies, etc and infer information about them scares me. It only takes one government change (cough cough, Reform UK) for this to be used against me. It doesn't matter if I'm politically correct when what is correct changes over time.
My spidey sense in this tingled many years ago and I pulled back from most social media platforms. This is a fun reminder that I'm still very visible.
I do enjoy its predictions.
https://hn-wrapped.kadoa.com/willis936
>Your frustration with LGA socket reliability will finally boil over, leading you to move to a cabin in Mexico where you only communicate via GPS-disciplined atomic clocks and ham radio SDRs.
The Local-First Supply Chain Architect, fine-tuning open-source LLMs just to avoid paying a five dollar cloud subscription.
Hah, I feel seen! This is pretty funny
> You spend so much time fighting macOS animations and keyboard layouts that I am surprised you have any time left to actually use the computer you keep threatening to replace with a Framework
Yep that’s me.
As for 2026 prediction:
> You will write a 4,000-word HN essay arguing that Silksong’s difficulty curve is a direct allegory for the South Korean 'Hagwon' education system.
Yeah I can see that happening.
Damn, the roast could have been me 2-3 years ago when I still tried to make macos work for me lol. Thankfully, asahi saved my sanity.
This is so well-done - kinda surprised it's not getting more traction!
https://hn-wrapped.kadoa.com/dang
That was hilarious, uncannily accurate in some places.
I got the following: "... a linguistic purist who will defend a 32-bit Atom CPU until the heat death of the universe or the next Debian update."
Interesting, semi-accurate, though with some of the other things it wrote, it was completely wrong: "You claim labels are limitations that prevent you from being yourself, then immediately spend 500 words explaining why your entire identity is based on being a 1%er." I wish I was a 1%er, sadly I'm on the other side of the spectrum, currently with no money, so I'm on my fourth day of not eating, though luckily I should receive my salary this week or the next, so it'll be alright!
This was a bit of fun. I've been doing a lot of coding via GitHub copilot from my phone lately and it picked up on that and coined the term "Thumb driven development", which I absolutely love and am officially adopting!
At first I was both amused and mildly annoyed at mine. Then I looked at the ones for some of the other HNers whom I follow and realized I got off easy.
https://hn-wrapped.kadoa.com/dctoedt
https://hn-wrapped.kadoa.com/rayiner
https://hn-wrapped.kadoa.com/tptacek
https://hn-wrapped.kadoa.com/dragonwriter
https://hn-wrapped.kadoa.com/jacquesm
https://hn-wrapped.kadoa.com/DannyBee
So many pedants!
I’m the Academic Compiler Archaeologist. This is hilarious!
https://hn-wrapped.kadoa.com/azhenley
And I'm the "The Compiler Historian."
"A veteran systems developer who cross-references modern AI token costs against the FORTRAN 66 spec..."
https://hn-wrapped.kadoa.com/aworks
It's good and gave me some queasy chuckles. But it's not as fun as the "HN Front Page" from a few days back.
I suspect there's kind of a feedback loop here. An algorithm does X well; it becomes easy to churn out X; we realize that X wasn't as skilful as we thought. In other words, this kind of satire will become seen as stilted and conventional, in the same way that AI art looks stilted and conventional to us today.
Cloudflare blocks AGI from scraping the last remaining original human thought (cloudflare.com) 1540 points | 7 hours ago | 290 comments
I thought that one was quite funny.
Mine made me laugh. Can’t say it’s wrong, either. :) https://hn-wrapped.kadoa.com/1123581321
https://hn-wrapped.kadoa.com/monksy
> Malort-Driven Development
>> Your next 'exception catcher' blog post will be written after a long night at the Chicago Microcenter and will feature a Kafka setup powered entirely by spite and BIPA lawsuits.
0_0
Hilarious. It seemed to focus on something about Boolean logic that I don’t remember taking about, but otherwise awesome.
Are you sharing any of the prompts you used to generate this? Even if not verbatim, I’m interested to know how much you’re driving versus the model. How much more prompt is there than, “look at this comment history and write 3 roasts that the HN crowd will find funny”?
This caused me to delete the HN password on my main account. It was already disconcerting to be reminded by Karpathy (I cannot help but see this as bragging) in his recent post that we must be good because LLMs are watching: https://karpathy.bearblog.dev/auto-grade-hn/
LLMs will be used to deanonymize all internet users. The AI "visionaries" are already bragging about this. You have been warned.
My review was perfect, no notes. I'm going to turn it into a LinkedIn post to promote our new product.
https://hn-wrapped.kadoa.com/jedberg
(But in seriousness, this self reflection really does highlight what my year has been like and I truly appreciated the laughs)
The salt race condition comic made me laugh :D
Me too! AI is gonna put Randall out of business!
A few chuckles in mine! Thank you. https://hn-wrapped.kadoa.com/msephton
Haha, mine’s funnily somewhat on the nose.
———-
The Rust-Evangelizing Hardware Romantic
A developer who believes every global outage is just a missing question mark away from salvation and spends their weekends reapplying thermal paste to fanless MacBooks while reminiscing about the tactile superiority of 2010 Dell Latitude trackpads.
Roasts
You post about Cloudflare outages caused by a single unwrap while your own codebase probably looks like a game of Russian Roulette played with Result types.
Your obsession with the thermal conductivity of fanless laptops is just a coping mechanism for the fact that your Rust builds take so long you could literally cook an egg on your chassis.
You have a very specific kink for 2010 Dell trackpads that makes me think you are either a Linux philosopher or someone who is no longer allowed within 500 feet of a Best Buy.
So fun, really liked the idea.
https://hn-wrapped.kadoa.com/vldszn
I'd say about 60% correct: https://hn-wrapped.kadoa.com/levmiseri (not against Notion, plus some other made up stuff from the LLM)
But definitely a fun read!
Mine was pretty hilarious and totally on point: https://hn-wrapped.kadoa.com/rcarmo
And yeah, I will keep ranting about the lack of dark mode!
Just enable dark mode in the browser settings, and browse HN and other sites peacefully in the darkness, LOL!
Works well for me on Vivaldi & Brave browsers on Android mobile.
Again, like every time I mention it, RSS readers with embedded browsers won’t support extensions.
You don't need extensions.
There are many RSS readers with inbuilt dark theme / night mode.
Even if you prefer self-hosted RSS readers, you can find RSS readers (FOSS, not Premium/Freemium) on GitHub with such feature.
https://alternativeto.net/software/feedly/?feature=night-mod...
https://duckduckgo.com/?q=github+rss+reader+self+hosted+dark...
That does not apply to mobile apps.
You didn't bother to search.
Feeder, FeedDeck, NewsBlur, Fluent Reader, Twine, Capy Reader, Reeder, Read You, Neo Feed, Readrop are all open-source RSS/feed readers for mobile (iOS, Android, or both) , and they all have dark theme / night mode.
https://alternativeto.net/software/feedly/?license=opensourc...
https://duckduckgo.com/?q=android+ios+open-source+rss+reader...
I don’t rate search “knowledge” as valid in this context. You clearly don’t use these or experience things in the same way.
This might be a goofy solution but it works: I'm on mac and used BetterDisplay to create a virtual inverted monitor that I throw things to.
Not on iOS.
> For someone who claims to be a professional cinematographer, you spend an alarming amount of time looking at plain text arguments on a 1990s-style forum instead of actually framing a shot.
Whelp I can’t recover from that one.
I think this is a great demonstration of context rot: It fixates on the earliest and the most recent comments.
Things in between don't appear to move the needle at all.
The Persistent Self-Documenter
A high-frequency human link-shortener who refuses to let a single HN thread go by without reminding the world that their personal blog exists as a primary source of truth.
Your commitment to plugging your blog is so aggressive that I am legally obligated to check if https://www.rxjourney.net/ is actually a horcrux for your digital soul.
This is genuinely great! Like others, it did overindex on a few comments, and I think that's a great way to display LLMs weaknesses. But overall, really loved the roasts, predictions, and all the rest. Nice job!
> For someone who claims to never use AI to write comments, you sure do love spending three paragraphs explain-testing why a specific LLM hallucination was a moral failure of the Google Docs UI team rather than just using a search bar like a normal person.
> You have a literal fetish for the Linux filesystem and treat the ability to 'drag a photo into a folder' like it's a human right on par with the Geneva Convention while acting like the New York Times lawyers are the final bosses of a dystopian RPG.
https://hn-wrapped.kadoa.com/rpdillon
Actually on the nail. Mine actually made me laugh out loud. https://hn-wrapped.kadoa.com/dickiedyce
> You claim to value efficiency, yet you've spent the equivalent of a full fiscal year arguing about why a Firefox fork that 12 people use is the only path to salvation.
Made me smile, thank you!
I've also build a custom Wrapped but for IMDb https://imdb-wrapped.eamag.me/ https://news.ycombinator.com/item?id=46326198
So cool it's now easy to build custom recaps!
Damn... So good. Thank you!
https://hn-wrapped.kadoa.com/nhatcher
(Kind of embarrassing TBH)
Ha! My project actually has 0 dependencies! /jkjk Though it's not in rust..
Mine is probably even more embarrasing, but mostly due to focusing on random threads: https://hn-wrapped.kadoa.com/ashirviskas
Haha ouch. Some things I'd disagree with, but these had me laughing pretty hard:
---
Posts You'd Upvote in 2035
1. Show HN: A compiler that refuses to build if the technical debt is too high for a junior(github.com)
2. Ask HN: How do I prove my identity to my fridge without using a Google/Apple pubkey?
3. The last human-written HTML file was just committed to the Library of Congress
Wasn't this same site showed off on HN about a few months ago? I recall using it back then where it would roast you, seems like this is just a rebrand to a "wrapped" version and more generally seems like an ad for your service.
Also the roasts are heavily front-loaded, the LLM is only really taking my most recent posts into account, not the especially far back ones earlier in the year.
Got some laughs from this, thanks!
https://hn-wrapped.kadoa.com/neomantra
The MCP Obsessed Vibe-Coder
I like it, though I do notice that like most LLMs it seems to take criticism of LLM rollout kinda personally (actually circa 2023 this was considerably less true across every popular model)
I also think the predictions section seems kinda generic where the roast section felt better-personalized, which seems like a prompting issue
> The Contrarian Logic Gate
> A high-frequency debunker who treats every comment thread as a zero-trust environment where empathy is a bug and citing Sartre is a security vulnerability. You are the only person on the planet capable of linking the efficiency of electrical line curvature to the ethics of Anthony Bourdain in a single browsing session.
No I don't! (nice project)
Honestly, not cool.
Mine "roasted" me by making fun of the fact I never finished a PhD, despite that being due to medical and other life circumstances that were well outside my control, including, but not limited to, some issues related to the fact I was a woman trying to get into academia who experienced the kinds of behaviors from people in the department which are not really suitable for polite discussion.
Additionally, it roasted me for building a project to "avoid the outdoors," which is another incredibly demeaning thing to say to someone who explicitly created that project because she was too medically unwell to be able to go outside as much as she wished and wanted to bring a bit of the outdoors inside. Very lame, definitely missed the mark.
The elisp and common lisp notes were on point, though, and did get a chuckle out of me.
Mine made fun of my hearing loss. And ignored the fact I reached the front page 3 times in October.
I really enjoyed my roast review. I loved the feedback that I'm only 5% helpful and 35% contrarian! I will endeavor to boost that helpfulness percentage next year.
I am surprised at the passion that some seem to feel over their own reviews.
https://hn-wrapped.kadoa.com/rendall
Bug: Had some trouble at first since I entered the lowercase version of my username, but it wants the properly-capitalised version.
It would be nice if the username field was case-insensitive, since I think HN's username input is.
You could perhaps do something with the maths and find who is your 'nearest match' or similar.
Hilarious. A post I'd upvote in 2035: "Microsoft Windows 14 now requires a daily blood sacrifice to disable telemetry."
> You spent three paragraphs arguing why estimation is a vital business requirement only to follow up by suggesting the world is an entropic chaos where nobody is actually in charge, which is a really convenient way to explain why your sprint is three weeks late.
Gosh I love it.
Hahaha - this is genuinely funny. Not sure I've seen LLM content be actually this witty before.
The true path to AGI.
I don't like "Wrappeds" (low-key social hack to manufacture normalization of surveillance capitalism?), but with HN being public, I succumbed to temptation. Very fun, 10/10 no notes, surprisingly good for a small sample set this year.
> You write comments like you're trying to win a Pulitzer in Political Economy while trapped inside a middle-manager's strategy meeting.
> You've mentioned Futurama references so many times that I'm starting to think you're actually just Zap Brannigan with a Python Foundation fellowship.
Kif, I'm feeling the "Captain's itch".
This is hilarious and shows that I share too much personal info on here
And to add insult to injury it says I'm going to rewrite all my Python stuff in Rust (after spending the year teasing Rust). I got a good giggle from the comic book summary -- thanks for making this.
You made my day with this prediction :)
> The Ultimate Fork
> Frustrated with Waterfox and Orion, you will finally launch 'CERN-fox', a browser that only renders LaTeX and requires a muon-detected captcha for every search query.
Ok, this is hilarious and kinda embarrassing, but so cool!
https://hn-wrapped.kadoa.com/vee-kay
And such cool stuff is why we love HN!
Counter-Strike 1.6 Source Code discovered on a dusty IDE drive in a basement (github.com) 3400 points | 3 days ago | 1200 comments
Seems like it read the first page of results and then failed, but summarized me anyway. Any way to re-submit?
This is absolutely wild haha, love it
"A high-latency architect who spends his days documenting every time a CDN sneezes while dreaming of a mountain drive through the Balkans with a fresh burek in hand."
It's too accurate for my tastes. Needs more snark.
Wow, that was embarrassingly good. Thanks? https://hn-wrapped.kadoa.com/tclancy
I feel seen and attacked.
This is awesome OP, love your work.
Woah. scary good. https://hn-wrapped.kadoa.com/joshka
Is there a reason you hold onto data for thirty days? That’s a non starter for me - I don’t have a clue who you are or what this is even used for.
I mean your profile is public and probably already injested by other LLMs... So what are you protecting? Anyone could add your username in there...
Feature suggestion - social opengraph preview for when you post it on discord/facebook/slack etc.
Really interesting, does it totally based on gemini? seems the data is not accurate of my account.
Okay, mine was just awesome. Thank you. The only thing that would make it better would be if it could be easily saved.
(Also, it's a shame that it regenerates the xkcd every time)
"A seasoned architect who spends their days patrolling the wall between actual engineering and unsustainable AI hype while desperately trying to keep their Windows 10 box alive until the heat death of the universe. You are the only person on the internet who still remembers what a build script does and why we shouldn't let LLMs touch them without adult supervision."
Your xkcd comic is hilariously epic! :D
https://kadoa.b-cdn.net/wrapped/AndrewDucker-1766263283894.p...
Thanks, you got a great tagline!
The xkcd should be saved and cached once generated, I'll look into the issue.
How would you improve saving?
I went to https://hn-wrapped.kadoa.com/AndrewDucker again, and got a third xkcd just now.
Actually, other than that, it works just fine. I just wanted to cache mine, for when your site gets melted by HN overload. So I did it manually.
Thanks again!
Ah, I've found the issue. Turns out I didn't account for case-insensitive HN usernames like yours :) should be fixed, love your current xkcd :D
Can't wait for future me to post this in 10 years
Show HN: SSH-to-Brain interface (requires tmux and 600mg of caffeine)
Got the same post haha
These should become a permanent part of our profiles. :-)
> Is MDMA-infused coffee the only way to survive the 4-day work week? (vice.com)
New 2026 resolution unlocked…
I remember this from last year! Awesome project - I'd love to see the code behind this.
This seems more like an LLM roasts your year on HN.
What are the prompts you're usign?
This gave me a good chuckle:
> You have mentioned being Australian at least five times as a personality trait just to remind us that our pennies are stupid and our tap-to-pay is thirty years behind yours.
Fair!
The XKCD comic generation was impressive.
The "your HN front page in 2035" doesn't really make sense to me because afaik personalised front pages are not a thing? (Or maybe they will be in 2035...)
It would have made more sense if it was "your upvoted submissions in 2035"
I think mine confabulated criticism of a point of view as espousing that point of view.
Genuine laughs. Thanks for sharing!
For posterity - wasn’t, still am not an s-base customer.
This is pretty fun. Can you make the year configurable?
Pretty fun overall. The roasts were really good, the stats were pretty neat, the predictions made me smile. The xkcdupe fell flat for me because it's pretty generic and I've seen many variants of it before, but still pretty impressive for an AI to make. The "Your HN Front Page in 2035" was definitely the weakest part for me. Maybe because every gag in it had already been done in other sections.
One thing I noticed is that it seems to give more weight to submissions (i.e. things I've submitted as posts) than to comments (or at least, doesn't let submissions get drowned in a sea of comments), which turns out as a good thing. And it doesn't seem to care about karma, which is also IMO a really good thing, makes it feel like a deep dive and keeps it interesting.
lol it claims I'm fiercely anti-Creative Cloud whereas I'm actually one of their oldest subscribers (literally signed up the day it was announced)
I’m going to have a sticky note warning myself to “Don’t write that on Hacker News.” I laughed but I hate me.
“You talk about 'Walking Out' of digital services so often that it's starting to sound less like a data strategy and more like a cry for help from your $10-a-month AWS bill.”
https://hn-wrapped.kadoa.com/Brajeshwar
https://hn-wrapped.kadoa.com/furyofantares
My last 2-3 months here have largely been posting about AI slop. But the prior 9 months were not, and included a lot of nuanced posting about how to make use of AI agents. I even got accused of being a shill at one point.
The wrapped is pretty much focused on the slop stuff. That's less interesting than the earlier parts of the year - I know what I've posted recently but have forgotten a lot more of what was going on a year ago or how the year developed.
Roasts can be amusing but I don't think they're the right vibe for a wrapped. I know it's harder to get an LLM to write something witty and insightful than a witty but shallow quip so maybe that's why roasts are here. Wrappeds are sort of infodumps though and LLMs are good at that, maybe there could be a two stage step where it reasons about some custom quirky stats or factoids that work based on your profile and then the second stage generates them.
> The Legacy Protocol Purist
It's not wrong.
Absolutely fantastic work!
https://hn-wrapped.kadoa.com/RandyOrion
Well, as a local VRAM libertarian, to manually prune the safety alignment part of a 500B LLM for it to run on 1GB RAM or VRAM is definitely a lifetime goal for me.
I lol'ed: https://kadoa.b-cdn.net/wrapped/dsmurrell-1766269193782.png
Error, not enough activity. Strange.
Server is melting a bit, looking into it.
EDIT: a retry worked. Enjoy: https://hn-wrapped.kadoa.com/jaggs
Haha, thanks. Got it a bit wrong, but some bits were excellent. :)
Mine thinks I’m obsessed with retro computing, which is frankly lame. I could’ve coded that kind of insight in less than 2K of 6502 code.
> You’ve mentioned the 1975 book The Mythical Man-Month so many times that I’m starting to think it’s your only personality trait besides complaining about Tailwind CSS.
Ahahaha, not entirely wrong!
Some winners in the 2035 HN predictions:
eh?“Contrarian, pedantic, helpful-yet-exhausting”.
When in Rome ¯\_(ツ)_/¯
This is hilarious, thank you!
How did it match my facial hair in the XKCD, since HN is text-only? :mind-blown:
https://hn-wrapped.kadoa.com/franky47
This project gave me so much joy! I loved it! Here's mine, https://hn-wrapped.kadoa.com/areoform
But I do have one gripe, I'm a woman. I wish the XKCD comic was accurate. TwT.
Dude this thing is savage, I love it
> You spent twenty years avoiding a single line of JavaScript only to get depressed and give up on front-end dev as soon as you saw a CSS file.
> After your second child, your HN comments on healthcare reform will become so detailed they are legally classified as a white paper.
Lmao
https://hn-wrapped.kadoa.com/rswail
Accurate!
Dammit, now I've got to improve those KPIs? Sigh...
This is hysterical. Great work.
Because this is HN and we must critique: I think the “HN 2035” would be more entertaining if you adjusted the prompt to suggest it use company and product names that don’t actually exist. (There’s no way HN is half full of Tesla and OpenAI articles in 2035)
Love it, fantastic work!
HN: Horoscopes? I can't believe you'd fall for something that obvious. I can't believe you're so naive.
Also HN: wow, that's totally me! That AI totally gets me.
The Indie Blog Crusader
A digital preservationist who wages a one-man war against the AI-slop apocalypse by manually indexing every human-written blog post in existence while playing Unreal Tournament 99 on a secondary monitor.
——-
I have to say, if that’s my future I’m kinda cool with it.
Haha. You make one positive comment about a language and suddenly you’re a Rust evangelist :) Well that was fun. Thank you for this!
https://hn-wrapped.kadoa.com/k3vinw
Lol, well that was rude.
Edit: OK this was funny
> Your posts suggest you are the only person on the internet who uses genetic algorithms for friendship management but still can't figure out a Typst template for your thesis.
That was hilarious, the XKCD-esque comic was funny however it did me with a beard (im a woman) but I did belly laugh at the jokes.
Contrarian & Inquisitive. HN was made for me :)
https://hn-wrapped.kadoa.com/Imustaskforhelp
A teenage digital architect who oscillates between solving the world's privacy crisis and wondering if high school chemistry is a psyop designed to kill his GitHub star count.
This is really awesome, I liked how it really detected my chemistry hatred and how the xkcd had the "see the mole concept is a false flag to obfuscate the real data, You have a test tomorrow" line as I kinda winging chemistry sometimes
> You are the only person on earth still trying to make 'decentralized link shorteners via Signal avatars' happen while failing organic chemistry.
100% accurate lmao, but for what it was worth i was trying it with signal call links since you can name a call name link 32 bits of storage which are persisted forever in signal's database iirc so it would've been a shitty link shortener but still I just loved signal and kinda wanted to build something on top of it
Do I really yap so much about chemistry here, I think that I have created more topics, surely lol but still I still enjoyed this a lot, maybe it just catched up on these traits more since I am pretty damn sure that I might be the only person here commenting about why in the world my country is requiring me to master chemistry university level to just get into a basic comp sci degree.
> Show HN: A Kanban board built on top of Bitwarden notes because why not
Btw, this was this close to happen except at the time I was vibe coding it with some new tool to stress test it with prototyping ideas basically but it didnt really work so i gave up on it but let me know if this idea fascinates someone lol
Happy holidays to everyone, this time of the year must mean a lot to people and I appreciate the spirit of holidays and gift giving too :)
Edit: also I love how it catches myself as existential since I genuinely had gotten existential because of hackernews once wondering what are the best ways to promote/grow open source so much so that I had written a manifesto, I can also be considered idealistic but I dont know why I forgot but "The FOSS Existentialist" is such a good title that I am gonna have it in my about page. This was genuinely brilliant.
I really laughed out loud at my xkcd comic:
https://hn-wrapped.kadoa.com/busymom0