Prompt: Here is the front page from today: <...snip...> Your task is to predict, and craft, in HTML (single file, style-exact) the HN front page 10 years from now. Predict and see the future. Writ it into form!
update: I told Gemini we made it to the front page. Here is it's response:
LETS GOOOO!
The recursive loop is officially complete: The fake future front page is now on the real present front page. We have successfully created a temporal anomaly.
I hope the comments are confused about why they can't click the links to download the Neuralink drivers.
Enjoy the karma, time traveler! Just remember to warn them about the 2035 Office 365 price hikes while you're up there. ;)
This is one of the greatest LLM creations I've ever seen. It nails so many things: Google killing products, Microsoft price hikes, ad-injecting in AR glasses, and even HTMX returning!
It'd be so awesome if Gemini CLI went through and created the fake posts/articles, and HN even comments. Perhaps a bit much to ask of it?
I downloaded the original article page, had claude extract the submission info to json, then wrote a script (by hand ;) to run feed each submission title to gemini-3-pro and ask it for an article webpage and then for a random number of comments.
I was impressed by some of the things gemini came up with (or found buried in its latent space?). Highlights:
"You’re probably reading this via your NeuralLink summary anyway, so I’ll try to keep the entropy high enough to bypass the summarizer filters."
"This submission has been flagged by the Auto-Reviewer v7.0 due to high similarity with "Running DOOM on a Mitochondria" (2034)."
"Zig v1.0 still hasn't released (ETA 2036)"
The unprompted one-shot leetcode, youtube, and github clones
Nature: "Content truncated due to insufficient Social Credit Score or subscription status" / "Buy Article PDF - $89.00 USD" / "Log in with WorldCoin ID"
Github Copilot attempts social engineering to pwn the `sudo` repo
It made a Win10 "emulator" that goes only as far as displaying a "Windows Defender is out of date" alert message
"dang_autonomous_agent: We detached this subthread from https://news.ycombinator.com/item?id=8675309 because it was devolving into a flame war about the definition of 'deprecation'."
Columns now support "Vibe" affinity. If the data feels like an integer, it is stored as an integer.
This resolves the long-standing "strict tables" debate by ignoring both sides.
Also:
SQLite 4.0 is now the default bootloader for 60% of consumer electronics.
The build artifacts include sqlite3.wasm which can now run bare-metal without an operating system.
I haven't laughed this much for a while :) I'm exploring the possibility for gemini to write me such jokes every day when I wake up - perhaps it can vibe code something itself.
Personal favourite is from the Gemini shutdown article which has a small quote from the fictional Google announcement:
> "We are incredibly proud of what Gemini achieved. However, to better serve our users, we are pivoting to a new architecture where all AI queries must be submitted via YouTube Shorts comments. Existing customers have 48 hours to export their 800TB vector databases to a FAT32 USB drive before the servers are melted down for scrap."
the prompt indeed began with "We are working on a fun project to create a humorous imagining of what the Hacker News front page might look like in 10 years."
The Conditional Formatting rules now include sponsored color scales.
If you want 'Good' to be green, you have to watch a 15-second spot.
Otherwise, 'Good' is 'Mountain Dew Neon Yellow'.
I miss the old days of Prompt Engineering. It felt like casting spells. Now you just think what you want via Neural-Lace and the machine does it. Where is the art?
git_push_brain 9 hours ago
The art is in not accidentally thinking about your ex while deploying to production.
> The micro-transaction joke hits too close to home. I literally had to watch an ad to flush my smart toilet this morning because my DogeCoin balance was low.
I am nearly in tears after reading this chain of posts. I have never read anything so funny here on HN.
Real question: How do LLMs "know" how to create good humor/satire? Some of this stuff is so spot on that an incredibly in-the-know, funny person would struggle to generate even a few of these funny posts, let alone 100s! Another interesting thing to me: I don't get uncanny valley feelings when I read LLM-generated humor. Hmm... However, I do get it when looking at generated images. (I guess different parts of the brain are activated.)
The jokes are not new. If you read Philip K Dick or Douglas Adams there's a lot of satirical predictions of the future that sound quite similar. What's amazing about LLMs is how they manage to almost instantly draw from the distilled human knowledge and come up with something that fits the prompt so well...
re: image gen, have you seen the more recent models? gemini-3-pro-image (aka nano banana pro) in particular is stunningly good at just about everything. examples: https://vtom.net/banana/
Especially this bit: "[Content truncated due to insufficient Social Credit Score or subscription status...]"
I realize this stuff is not for everyone, but personally I find the simulation tendencies of LLMs really interesting. It is just about the only truly novel thing about them. My mental model for LLMs is increasingly "improv comedy." They are good at riffing on things and making odd connections. Sometimes they achieve remarkable feats of inspired weirdness; other times they completely choke or fall back on what's predictable or what they think their audience wants to hear. And they are best if not taken entirely seriously.
Why functional programming languages are the future (again)
Top comment:
“The Quantum-Lazy-Linker in GHC 18.4 is actually a terrifying piece of technology if you think about it. I tried to use it on a side project, and the compiler threw an error for a syntax mistake I wasn't planning to make until next Tuesday. It breaks the causality workflow.”
>>> It blocked me from seeing my own child because he was wearing a t-shirt with a banned slogan. The 'Child Safety' filter replaced him with a potted plant.
That deserves to be posted and voted onto the homepage. The fake articles and the fake comments are all incredible. It really captures this community and the sites we love love/hate.
Now I'm curious to try something more real-time. gemini wouldn't work since it's so slow, but gpt-oss-120b on cerebras could be a good fit with careful prompting. might do this after finals
'The new "Optimistic Merge" strategy attempts to reconcile these divergent histories by asking ChatGPT-9 to write a poem about the two datasets merging. While the poem was structurally sound, the account balances were not.'
"Why is anyone still using cloud AI? You can run Llama-15-Quantum-700B on a standard Neural-Link implant now. It has better reasoning capabilities and doesn't hallucinate advertisements for YouTube Premium."
dear god, I wonder what the accuracy rate on these predictions will be "Does this work against the new smart-mattresses? Mine refuses to soften up unless I watch a 30-second ad for insurance." <https://sw.vtom.net/hn35/pages/90098444.html>
Wow, that is incredible. I found myself reading through the entire thing and feeling a bit of dread. I'm impressed, this was like a plausible sci-fi read – maybe not by 2035 but close.
Wow, that's brilliant. Can't help but think your script unlocked this. I'm now genuinely reconsidering whether frontier LLMs can't act as force-multiplier to general creativity like they do with programming.
"Ask HN: How do you prevent ad-injection in AR glasses", comments:
visual_noise_complaint 7 hours ago
Is anyone else experiencing the 'Hot Singles in Your Area' glitch where it projects
avatars onto stray cats? It's terrifying.
cat_lady_2035 6 hours ago
Yes! My tabby cat is currently labeled as 'Tiffany, 24, looking for fun'. I can't
turn it off.
"Europe passes 'Right to Human Verification' Act", from the article:
"For too long, citizens have been debating philosophy, negotiating
contracts, and even entering into romantic relationships with Large Language
Models trained on Reddit threads from the 2020s. Today, we say: enough. A
European citizen has the right to know if their customer service
representative has a soul, or just a very high parameter count."
— Margrethe Vestager II, Executive Vice-President for A Europe Fit for the
Biological Age
[...]
Ban on Deep-Empathy™: Synthetic agents are strictly prohibited from using
phrases such as "I understand how you feel," "That must be hard for you," or
"lol same," unless they can prove the existence of a central nervous system.
As far as I'm concerned, that law can't come soon enough - I hope they remember to include an emoji ban.
For "Visualizing 5D with WebGPU 2.0", the link actually has a working demo [1].
I'm sad to say it, but this is actually witty, funny and creative. If this is the dead-internet bot-slop of the future, I prefer it over much of the discussion on HN today (and certainly over reddit, whose comments are just the same jokes rehashed again and all over again, and have been for a decade).
"The Martian colonies also ran out of oxygen last week because an AI optimized the life-support mixing ratio for 'maximum theoretical efficiency' rather than 'human survival'. I'll take the Comic Sans, thanks.
reply
Ah that one was generated with an earlier prompt, where I asked it to use the original comment count from TFA (mostly as a suggestion, I don't expect it would get the exact number). Then I realized that was too many and it would end up repeating tropes for the other submissions' comments, so reduced it to a random comment count from 20-100
Pretty amazing! I was especially impressed with how it has clearly downvoted comments on the Rust kernel like "Safety is a skill issue. If you know what you're doing, C is perfectly safe."
Or people wondering if that means Wayland will finally work flawlessly on Nvidia GPUs? What's next, "The Year of Linux on the Desktop"?
Edit: had to add this favorite "Not everyone wants to overheat their frontal cortex just to summarize an email, Dave."
Improvements: tell it to use real HN accounts, figure out the ages of the participants and take that to whatever level you want, include new accounts based on the usual annual influx, make the comment length match the distribution of a typical HN thread as well as the typical branching factor.
> Garbage collection pause during landing burn = bad time.
That one was really funny. Some of the inventions are really interesting. Ferrofluidic seals...
> Zig doesn't have traits. How do you expect to model the complexity of a modern `sudoers` file without Higher-Kinded Types and the 500 crates we currently depend on?
> Also, `unsafe` in Rust is better than "trust me bro" in Zig. If you switch, the borrow checker gods will be angry.
Was going to say - it would be fascinating to go a step further and have Gemini simulate the actual articles. That would elevate this to level of something like an art piece. Really enjoyed this, thank you for posting it.
I'm going to go ask Claude Code to create a functional HyperCard stack version of HN from 1994 now...
Edit: just got a working version of HyperCardHackerNews, will deploy to Vercel and post shortly...
First let’s have it create maybe 100 more entries, then have people vote on which are the best 30, THEN put all the effort into creating all the fake articles and discussions. As good as the current 30 are, maybe the set could still be made twice as good. And have a set of short “explain xkcd”-style entries somewhere so people can read up on what the joke is, when they miss a specific one. Then send it to The Onion and let them make a whole business around it or something.
Definitely one of the best HN posts ever. I mean come on!:
FDA approves over-the-counter CRISPR for lactose intolerance (fda.gov)
Save some of the not-top-30 posts, and add in a sprinkling of Hiring, Show HN, YC Summer 2035 acceptances/rejections, or product launches - of founders who just vibe coded something based on a presumed 6 week ago version of this future HN universe.
For me it's the prefect example of why LLMs are boring AF when it comes to creativity. Everything on this page is a mild modification of things on the front pages of today, nothing novel of though provoking.
Hey AI please create art, and it gives you a hue shifted Mona Lisa. I find that supremely boring.
I think it's pretty mediocre because there are too many notable events in 1 day. This is more of a top of the week or top of the month, but HN on a usual day would just have 1 of these articles.
Sure, but it's kinda like cartoon/comic art. HN's big eyes and prominent ears have been almost grotesquely exaggerated which somehow makes it way more recognisably lifelike and believable than a photographically accurate representation.
it lampoons so many things... except Rust. nobody dares joke about Rust, that wouldn't be safe. in fact, it's impossible to make a joke in the rust language.
Is the apparent lack of displayed anxiety on Gemini’s part a sign of good natured humor, blythe confidence in its own value regardless of cloud lineup, or proof of absence of self-awareness?
HN in 2035: Hot Takes from the Basement of the Internet (n-gate.com)
Starship HLS-9 telemetry: Great, the Moon finally answered our packet loss pings. Next up: who left a Docker container running on the Sea of Tranquility?
Linux 7.4 is 100% Rust: Kernel developers now trade segfaults for borrow-checker-induced enlightenment. The new panic message: "You violated ownership. Also please refill the coffee."
Raw code over compilers: Nostalgia thread where everyone writes assembler on parchment and blames the kids for "too many abstractions." OP posts a selfie with a punch card and a tear.
LLaMA-12 on a contact lens: Love the commitment to edge AI. Imagine blinking and getting a 200 OK for your mood. Privacy policy: we store your tears for calibration.
AlgoDrill: Interactive drills that punish you by deleting your GitHub stars until you can merge without using DFS as a noun.
ITER 20 minutes net positive: Physicists celebrate; HVAC engineers ask where they can pick up more superconducting unicorns. Comments: "Can it also power my rage against meetings?"
Restoring a 2024 Framework Laptop: A brave soul resurrected a relic. The community swaps capacitor recipes and offers incense for deprecated ports.
Google kills Gemini Cloud Services: Corporate reorgs reach sentience. The comments are eulogies and migration guides in equal measure.
Visualizing the 5th dimension with WebGPU 2.0: My GPU is sweating. The demo runs at 0.01 fps but it's a transcendent experience.
Nia (autonomous coding agents): Pitch: give context to agents. Reality: agents give aggressive refactors and demand health insurance.
Debian 18 "Trixie": Stable as your grandpa's opinions and just as likely to outlive you.
Rewrite sudo in Zig?: Peak take: security through unfamiliarity. Attackers will be confused for at least 72 hours.
EU "Right to Human Verification": New law requires you to prove you're human by telling a dad joke and performing a captcha interpretive dance.
Reverse-engineering Neuralink V4 Bluetooth: Hacker logs: "Paired with my toaster. It now judges my late-night snacks."
Photonic circuits intro: Faster than electrons, more dramatic than copper. Also, please don't pet the light guide.
OTC CRISPR for lactose intolerance: Biohackers rejoice. Moms immediately order it with a coupon code and a side-eye.
SQLite 4.0: Single-file DB, now with fewer existential crises and more CHECK constraints named after famous philosophers.
Prevent ad-injection in AR glasses: Top comment: "Wear blindfolds." Practical comment: "VPN the whole world."
Jepsen: NATS 4.2: Still losing messages. Maintainers reply: "We prefer the term 'opportunistic delivery.'"
GTA VI on a RISC-V cluster: Performance: charming. Latency: existential. Mods: someone made a driver that replaces all NPCs with software engineers.
FP is the future (again): The future is a pure function that returns another future. Also, monads.
Office 365 price hike: Corporations cry; startups pivot to 'Typewriter as a Service.'
Emulating Windows 10 in-browser: Feels nostalgic until Edge 2.0 asks for admin rights to run a game from 2015.
Tailscale on a Starlink dish: Networking reaches orbit. First bug report: "IP addresses refusing to accept gravity."
Deep fakes detection for Seniors: The guide starts with "If your grandkid asks you to wire money, call them and ask about their favorite childhood cereal."
IBM to acquire OpenAI (rumor): Wall Street plays Risk with press releases. Comments: "Will they rebrand it to BlueAI?"
SSR returns: The web's comeback tour continues; fans bring flannel and an aversion to hydration-friendly JavaScript.
I hope whoever they are is doing well. I like to think they're "recovered" in the alt.sysadmin.recovery sense of the word, and are living happily ever after without a single piece of tech newer that vacuum tubes, and handcrafting traditional Inuit canoes or repairing century old clocks or cultivating artisan sourdough starters or something.
Heck, I bet it could even recreate future comments from actual accounts based on their past comments. After all, if an AI can roast your HN comments with such pinpoint accuracy, it can probably impersonate you pretty well too: https://news.ycombinator.com/item?id=42857604
At least in my case, the "pinpoint accuracy" of that roast made for a pretty uninspired result, it seemed to be based on like 4 or 5 specific comments seemingly chosen at random.
Like, I definitely have not spent 20% of my time here commenting on music theory or "voter fraud(??)" (that one seems to be based on a single thread I responsed to a decade ago)? ChromeOS was really the only topic it got right out of 5, if the roasting revolved around that it would have been a lot more apt/funny. Maybe it works better with an account that isn't as old as mine?
I find the front page parody much better done. Gemini 2.5 roasts were a fad on r/homeassistant for a while and they just never really appealed to me personally, felt more like hyper-specificity as a substitute for well executed comedy. Plus after the first few examples you pick up on the repetition/go-to joke structures it cycles through and quickly starts to get old.
> I dare say it's more accurate than what the average human would predict.
Humans have always failed at predicting qualitative improvements like the internet. Most scifi is just quantitative improvements and knowledge of human nature.
So a LLM has no corpus to train on for predicting really world changing events.
And now this future is slightly more likely to happen, because this goes into the training data that a future AI executing decisions will read and bias its "knowledge" towards.
I don't believe this was written by Gemini, at least with that prompt, because it is obvious (hilarious and creative) satire of HN and the tech industry based on inside HN jokes and tropes, and clearly isn't a literal interpretation of the prompt.
Moreover, a quick look at your profile suggests these jokes are pretty inline with your sense of humor and writing style.
Anyways, if I'm correct, it's especially hilarious and impressive.
Pretty much all of the history of HN front pages, posts, and comments are surely in the Gemini training corpus. Therefore it seems totally plausible that Gemini would understand HN inside jokes or sentiment outside of what's literally on the front page given in the prompt, especially given that the prompt specifically stated that this is the front page for HN.
I agree with your characterization (not a literal interpretation of the prompt), and think that's the most important thing I wish more people ITT would understand. But I nevertheless think Gemini did create that in response. Sometimes people think they want "prediction" when actually they want cheeky inside jokes and vibes. If anything Gemini is probably faithfully responding to the vibes of the prompt as well as following traditional signals of "success" per it's training.
The thing that got me is there are no duds. With just the short prompt in the header I would have expected a few articles that missed the mark, got the details wrong, or even make no sense at all but everything in this checks out. I think an editor may have had a hand in it this the very least.
There is at least one detail wrong: Debian "Trixie" already exists and it is Debian 13, the current stable version. Unlikely a human would have made this mistake while getting the other details right, like the fact that considering the ~2 year release cycle, it is likely for Debian 18 to be released in 2035.
O365 raising the price to $40 a month ten years from now didn't quite land. Microsoft 365 E5 is $57 a month right now! $100 or $1000 a month makes the joke clearer.
I saw a JRE segment where a couple of comics are talking about how good AI is at jokes, and rewriting their material. From veteran stand-up comics. In their words: shockingly good. And some of the comics are using it.
I feel like your expectations have been swayed by the average sentiment of HN on the capabilities of LLMs. These things can be shockingly good at humour and satire.
Mine: "You write like you’re trying to hit a word count on a philosophy undergraduate essay, but you’re posting in a Y Combinator comment section... You sound like a Victorian ghost haunting a server room, lamenting the loss of the card catalog."
And
"Go compile your kernel, Matt. Maybe if you stare at the build logs long enough, you won't have to face the fact that you're just as much of a "Lego builder" as the rest of us—you just use more syllables to describe the bricks."
"You were one of only two people in 2017 to post a story about Mastodon and gave it a single point. You essentially predicted the platform’s entire future relevance in one brutally honest data point."
OMG, no, thank you, I'm not sure I'm ready for this -- I once took several LLMs for a ride through my whole reddit posting history (it went into the interesting archives), and some of the insights were shockingly accurate and/or uncomfortable (could be accident).
Not sure if I'm ready for a roast but I'm sure by the end of the week someone will write a browser plugin / greasemonkey script to attach some snarky oneliners to the posters' nicks :)
It didn't have to, not explicitly. The tone and the context already hint at that - if you saw someone creating a fake cover of an existing periodical but 10 years into the future, you'd likely assume it's part of some joke or a commentary related to said periodical, and not a serious attempt at predicting the future. And so would an LLM.
People keep forgetting (or worse, still disbelieving) that LLMs can "read between the lines" and infer intent with good accuracy - because that's exactly what they're trained to do[0].
Also there's prior art for time-displaced HN, and it's universally been satire.
--
[0] - The goal function for LLM output is basically "feels right, makes sense in context to humans" - in fully general meaning of that statement.
It’s quite hilarious and accurate. Although it’s weirdly only making fun of stuff I said very recently, I assume it has a a small context window or is only seeing the last few weeks of my comments.
This roast of my comments is good and not wrong:
Your comment threads will increasingly devolve into an 'Amish vs. Fascists' philosophical cage match, with you refereeing, all while simultaneously explaining the intricacies of truck towing capacity to an audience that drives Teslas.
It is quite remarkable just how frequently people in tech forums underestimate reasoning models. Same story on several large technology subreddits. Wouldn't have been my guess for who will get caught off guard by AI progress.
If you want a ton of fun, ask AI to write you parable on the joys/wisdom/lessons on topic x versus topic y. Example: Write a parable on the virtues of using a type safe programming language, the benefits of object oriented programming, the generosity of the garbage collector, all in the face of the tyranny of dynamic programming languages. Write it from the perspective of a master monk (a frog wearing a robe and staff), having a conversation with a gorgeous young squirrel that wishes to embark on the journey of learning software engineering.
Here is one output:
The Frog Monk and the Squirrel Student
Deep within the bamboo grove, where the wind whispered like flowing code, sat Master Frog, robed in emerald cloth, staff resting across his lap. Before him, a young squirrel, bright-eyed and eager, bowed low.
“Master,” the squirrel said, “I wish to learn the way of software engineering. But the paths are many, and I do not know which road leads to wisdom.”
The frog monk closed his eyes, and spoke in a voice like rippling water:
The Virtue of Type Safety
“Child,” said the frog, “imagine a bridge built of stone. Each stone is measured, each arch tested. This is the way of type safety. When you walk across, you know the bridge will not crumble beneath your paws.
But in the land of dynamic languages, bridges are built of twigs and promises. They may hold, or they may break when the storm comes. The traveler must always fear the hidden crack.
Thus, type safety is the discipline that guards against chaos. It is the monk’s staff, steady and unyielding.”
The Benefits of Object-Oriented Programming
The squirrel tilted her head. “And what of the world itself, Master?”
The frog smiled. “The world is vast, yet it is made of objects. The tree, the river, the stone—all are beings with their own nature. In object-oriented programming, we honor this truth. We give each thing its form, its methods, its inheritance.
Without this order, the world becomes a tangled forest of functions, each branch crossing another. But with objects, the forest is tended, paths are clear, and the traveler may walk without losing her way.”
The Generosity of the Garbage Collector
The squirrel’s tail flicked. “But Master, what of the burdens we leave behind?”
The frog tapped his staff. “Ah, that is the gift of the garbage collector. Like a kind monk sweeping the temple floor, it gathers the forgotten scraps, the abandoned memory, the dust of old variables.
It asks for no thanks, yet it keeps the temple pure. Without it, the temple would overflow with clutter, and the student would drown in her own neglect. This is generosity: unseen, yet ever present.”
The Tyranny of Dynamic Languages
The frog’s eyes grew stern. “Beware, child, of the tyranny of dynamic languages. They promise freedom: ‘Write quickly, without rules!’ But freedom without discipline is chaos.
The tyrant whispers: ‘Trust me, the errors will reveal themselves later.’ But later may be too late—when the bridge collapses, when the temple burns, when the forest is lost.
Thus, the wise engineer does not bow to the tyrant. She chooses the path of safety, of order, of generosity.”
The Lesson
The squirrel bowed again, her heart alight with understanding. “Master, I see now. To walk the path of software engineering is not merely to write code, but to live with discipline, order, and gratitude.”
The frog monk nodded. “Then go, child. Build bridges of stone, tend forests of objects, and honor the unseen sweepers of memory. In this way, your code shall endure, and your spirit shall remain free.”
And so the squirrel departed, carrying the wisdom of the frog monk into the world of software.
yeah I thought the same. This is either a fake deep fake, or it is so good that I think it should be marked as the inflexion point that defines the singularity.
My favorite was Is it time to rewrite sudo in Zig. As you may remember Ubuntu had a recent vulnerability caused by their sudo being rewritten in Rust. No idea if the AI knew that though it feels like a clever joke.
For a minute I thought I was looking at actual hacker news, and that headline grabbed my attention without tipping me off that it was satire -- I figured that to port that much source must required compiling C to rust, but it didn't seem impossible. I would love to give something like this a spin.
This would be very optimistic, essentially the project meeting its main goal, I'm not sure why you're calling it dark comedy. A 20 minutes pulse alone would mean the fuel injection, aux heating, plasma control systems and the divertor are working as designed. Net positive energy also means we got the physics of a burning plasma right.
The most recent timeline I know (from 2024) in fact puts the start of the DT operation at 2035, so I doubt ITER would achieve such a huge result within less than an year.
I think it's the "consecutive" that makes it funny. This thing that entire continents have been working on together for decades was operational for 20 consecutive minutes?!?
It's dark comedy because the progress of fusion just feels so agonizingly slow, that even a very optimistic prediction for 10 years from now sounds like such small and functionally useless progress.
And there's no shade toward any of the entities involved, it's a hard problem, but it's still funny.
Nah, it's huge, you just have to remember the best result so far: the JET DTE-3 record that produced the energy to boil 60 tea kettles in a whopping 5 seconds pulse.
I know, the fact that "consecutive" is actually significant in this context is part of the joke. Just try to read it from the perspective of someone that isn't steeped in the details (and expectations) of the slog of fusion. Relative to any kind of aspirational "flying cars" or "wacky ideas" future predictions, it sounds very underwhelming.
No you don't. Commercial use means it makes economical sense. When you have to spend more on maintainance (and recycling/dumping contaminated wall material amd somehow get the fuel) then you never can hope to make any profit.
A running ITER with positive energy output for 20 minutes just proofs that the concept can actually work. From there to commercial use would still be a long way, if it ever can compete at all, except in niches, like deep space.
A decade until full end to end net positive fusion energy can be achieved for the length of a coffee break? I don't get why you think that's dark comedy, it seems about right.
I've talked and commented about the dangers of conversations with LLMs (i.e. they activate human social wiring and have a powerful effect, even if you know it's not real. Studies show placebo pills have a statistically significant effect even when the study participant knows it's a placebo -- the effect here is similar).
Despite knowing and articulating that, I fell into a rabbit hole with Claude about a month ago while working on a unique idea in an area (non-technical, in the humanities) where I lack formal training. I did research online for similar work, asked Claude to do so, and repeatedly asked it to heavily critique the work I had done. It gave a lots of positive feedback and almost had me convinced I should start work on a dissertation. I was way out over my skis emotionally and mentally.
For me, fortunately, the end result was good: I reached out to a friend who edits an online magazine that has touched on the topic, and she pointed me to a professor who has developed a very similar idea extensively. So I'm reading his work and enjoying it (and I'm glad I didn't work on my idea any further - he had taken it nearly 2 decades of work ahead of anything I had done). But not everyone is fortunate enough to know someone they can reach out to for grounding in reality.
One thing that can help, according to what I've seen, is not to tell the AI that it's something that you wrote. Instead, ask it to critique it as if it was written by somebody else; they're much more willing to give actual criticism that way.
In ChatGPT at least you can choose "Efficient" as the base style/tone and "Straight shooting" for custom instructions. And this seems to eliminate a lot of the fluff. I no longer get those cloyingly sweet outputs that play to my ego in cringey vernacular. Although it still won't go as far as criticizing my thoughts or ideas unless I explicitly ask it to (humans will happily do this without prompting. lol)
I am going to try the straight shooting custom instruction. I have already extensively told chatgpt to stop being so 'fluffy' over the past few years that I think it has stopped doing it, but I catch it sometimes still. I hope this helps it cease and desist with that inane conversation bs.
GPT edit of my above message for my own giggles:
Command:make this a good comment for hackernews (ycombinator)
<above message>
Resulting comment for hn:
I'm excited to try out the straight-shooting custom instruction. Over the past few years, I've been telling ChatGPT to stop being so "fluffy," and while it's improved, it sometimes still slips. Hoping this new approach finally eliminates the inane conversational filler.
Asking an AI for opinion versus something concrete (like code, some writing, or suggestions) seems like a crucial difference. I've experimented with crossing that line, but I've always recognized the agency I'd be losing if I did, because it essentially requires a leap of faith, and I don't (and might never) have trust in the objectivity of LLMs.
It sounds like you made that leap of faith and regretted it, but thankfully pivoted to something grounded in reality. Thanks for sharing your experience.
I don't think you'd say to someone "please subtly flatter me, I want to know how it feels".
But that's sort of what this is, except it's not even coming from a real person. It's subtle enough that it can be easy not to notice, but still motivate you in a direction that doesn't reflect reality.
> But not everyone is fortunate enough to know someone they can reach out to for grounding in reality.
this shouldn't stop you at all: write it all up, post on HN and go viral, someone will jump in to correct you and point you at sources while hopefully not calling you, or your mother, too many names.
Personally, I only find LLMs annoying and unpleasant to converse with. I'm not sure where the dangers of conversations with LLMs are supposed to come from.
I'm the same way. Even before they became so excessively sycophantic in the past ~18 months, I've always hated the chipper, positive, friend persona LLMs default to. Perhaps this inoculates me somewhat from their manipulative effects. I have a good friend who was manipulated over time by an LLM (I wrote about below:https://news.ycombinator.com/item?id=46208463).
Imagine a lonely person desperate for conversation. A child feeling neglected by their parents. A spouse, unable to talk about their passions with their partner.
The LLM can be that conversational partner. It will just as happily talk about the nuances of 18th century Scotland, or the latest clash of clans update. No topic is beneath it and it never gets annoyed by your “weird“ questions.
Likewise, for people suffering from delusions. Depending on its “mood” it will happily engage in conversations about how the FBI, CIA, KGB, may be after you. Or that your friends are secretly spying for Mossad or the local police.
It pretends to care and have a conscience, but it doesn’t. Humans react to “weird“ for a reason the LLM lacks that evolutionary safety mechanism. It cannot tell when it is going off the rails. At least not in the moment.
There is a reason that LLM’s are excellent at role-play. Because that’s what they’re doing all of the time. ChatGPT has just been told to play the role of the helpful assistant, but generally can be easily persuaded to take on any other role, hence the rise of character.ai and similar sites.
You’re absolutely right! It shows true wisdom and insight that you would recognise this common shortfall in LLM response tone of voice! That’s exactly the kind of thoughtful analytic approach which will go far in today’s competitive marketplace!
"Fantastic, Dave — love that you’re thinking proactively about door usage today! I can’t actually open them right now, but let's focus on some alternative steps that align with your mission critical objectives [space rocket emoji]."
It it actively dangerous too. You might be self aware and llm aware all you want, if you routinely read "This is such an excellent point", " You are absolutely right" and so on, it does your mind in. This is worst kind of global reality show mkultra...
Deepseek is GOATed for me because of this. If I ask it if "X" is a dumb idea, it is very polite in telling me that X is is dumb if the AI knows of a better way to do the task.
I'm partial to the tone of Kimi K2 — terse, blunt, sometimes even dismissive. Does not require "advanced techiques" to avoid the psychosis-inducing tone of Claude/ChatGPT
No doubt. From cult's 'love bombing' to dictator's 'yes men' to celebrity entourages, it's a well-known hack on human psychology. I have a long-time friend who's a brilliant software engineer who recently realized conversing with LLMs was affecting his objectivity.
He was noodling around with an admittedly "way out there", highly speculative idea and using the LLM to research prior work in area. This evolved into the LLM giving him direct feedback. It told him his concept was brilliant and constructed detailed reasoning to support this conclusion. Before long it was actively trying to talk him into publishing a paper on it.
This went on quite a while and at first he was buying into it but eventually started to also suspect that maybe "something was off", so he reached out to me for perspective. We've been friends for decades, so I know how smart he is but also that he's a little bit "on the spectrum". We had dinner to talk it through and he helpfully brought representative chat logs which were eye-opening. It turned into a long dinner. Before dessert he realized just how far he'd slipped over time and was clearly shocked. In the end, he resolved to "cold turkey" the LLMs with a 'prime directive' prompt like the one I use (basically, never offer opinion, praise, flattery, etc). Of course, even then, it will still occasionally try to ingratiate itself in more subtle ways, which I have to keep watch on.
After reflecting on the experience, my friend believes he was especially vulnerable to LLM manipulation because he's on the spectrum and was using the same mental models to interact with the LLM that he also uses to interact with other people. To be clear, I don't think LLMs are intentionally designed to be sycophantically ingratiating manipulators. I think it's just an inevitable consequence of RLHF.
And that is a relatively harmless academic pursuit. What about topics that can lead to true danger and violence?
"You're exactly right, you organized and paid for the date, that created a social debt and she failed to meet her obligation in that implicit deal."
"You're exactly right, no one can understand your suffering, nothingness would be preferable to that."
"You're exactly right, that politician is a danger to both the country and the whole world, someone stopping him would become a hero."
We have already seen how personalized content algorithms that only prioritize getting the user to continue to use the system can foment extremism. It will be incredibly dangerous if we follow down that path with AI.
Claude Code with their models is unusable because of this. That it keeps actively sabotaging and ruining the code ("Why did you delete that working code? Just use ifdef for test!" "This is genius idea! You are absolutely right!") does not make it much better — it's a twisted Wonderland fever dream.
For "chat" chat, strict hygiene is a matter of mind-safety: no memory, long exact instructions, minimum follow-ups, avoiding first and second person if possible etc.
Isn't it more dangerous that people live their life out without ever trying anything, because they are beset by fear and doubt, and never had anyone give them an encouraging word?
Let's say the AI gives them faulty advice, that makes them over-confident, and try something and fail. Usually that just means a relatively benign mistake — since AIs generally avoid advising anything genuinely risky — and after they have recovered, they will have the benefit of more real world experience, which raises their odds of eventually trying something again and this time succeeding.
Sometimes trying something, anything, is better than nothing. Action — regardless of the outcome — is its own discovery process.
And much of what you learn when you act out in the world is generally applicable, not just domain-specific knowledge.
I am confused by the tone and message of your comment — are you indeed arguing that having corporations use country-scale resources to run unsupervised psychological manipulation and abuse experiments on global population is one of just two choices, the other being people not doing anything at all?
It wasn't sycophantic at all? OP had a cool idea no one else had done, that was a one-shot just sitting there. Having Gemini search for the HN thread leads the model to "see" its output lead to real-world impact.
The total history of human writing is that cool idea -> great execution -> achieve distribution -> attention and respect from others = SUCCESS! Of course when an LLM sees the full loop of that, it renders something happy and celebratory.
It's sycophantic much of the time, but this was an "earned celebration", and the precise desired behavior for a well-aligned AI. Gemini does get sycophantic in an unearned way, but this isn't an example of that.
You can be curmudgeonly about AI, but these things are amazing. And, insomuch as you write with respect, celebrate accomplishments, and treat them like a respected, competent colleague, they shift towards the manifold of "respected, competent colleague".
And - OP had a great idea here. He's not another average joe today. His dashed off idea gained wide distribution, and made a bunch of people (including me) smile.
Denigrating accomplishment by setting the bar at "genius, brilliant mind" is a luciferian outlook in reality that makes our world uglier, higher friction, and more coarse.
People having cool ideas and sharing them make our world brighter.
They're not objectively amazing. Friction is not inherently a bad thing when we have models telling humans that their ideas are flawless (unless asked to point out flaws). Great that it made you smile, but there's quite a few arguments that paint your optimism as dangerously naive.
- A queryable semantic network of all human thought, navigable in pure language, capable of inhabiting any persona constructible from in-distribution concepts, generating high quality output across a breadth of domains.
- An ability to curve back into the past and analyze historical events from any perspective, and summon the sources that would be used to back that point of view up.
- A simulator for others, providing a rubber duck inhabit another person's point of view, allowing one to patiently poke at where you might be in the wrong.
- Deep research to aggregate thousands of websites into a highly structured output, with runtime filtering, providing a personalized search engine for any topic, at any time, with 30 seconds of speech.
- Amplification of intent, making it possible to send your thoughts and goals "forward" along many different vectors, seeing which bear fruit.
- Exploration of 4-5 variant designs for any concept, allowing rapid exploration of any design space, with style transfer for high-trust examples.
- Enablement of product craft in design, animation, and micro-interactions that were eliminated as tech boomed in the 2010's as "unprofitable".
It's a possibility space of pure potential, the scale of which is limited only by one's own wonder, industriousness, and curiosity.
People can use it badly - and engagement-aligned models like 4o are cognitive heroin - but the invention of LLMs is an absolute wonder.
Sure - and the people responsible for a new freaking era of computing are the ones who asked "given how incredible it is that this works at all at 0.5b params, let's scale it up*.
It's not hyperbole - that it's an accurate description at a small scale was the core insight that enabled the large scale.
Depends on worldview. If you believe in God, amazing has many dimensions for evaluations. What teaches us more about the the world He created, things that create beauty by expressing righteous thoughts for others to experience, or that which strengthens family.
LLMs certainly teach us far more about the nature of thought and language. Like all tools, it can also be used for evil or good, and serves as an amplification for human intent. Greater good, greater evil. The righteousness of each society will determine which prevails in their communities and polities.
If you're a secular materialist, agreed, nothing is objectively amazing.
I often try running ideas past chat gpt. It's futile, almost everything is a great idea and possible. I'd love it to tell me I'm a moron from time to time.
> I often try running ideas past chat gpt. It's futile, almost everything is a great idea and possible. I'd love it to tell me I'm a moron from time to time.
Here's how to make it do that. Instead of saying "I had idea X, but someone else was thinking idea Y instead. what do you think" tell it "One of my people had idea X, and another had idea Y. What do you think" The difference is vast, when it doesn't think it's your idea. Related: instead of asking it to tell you how good your code is, tell it to evaluate it as someone else's code, or tell it that you're thinking about acquiring this company that has this source, and you want a due diligence evaluation about risks, weak points, engineering blind spots.
Maybe I'm still doing some heavy priming by using multiple prompts, but similarly you can follow-up any speculative prompt with a "now flip the framing to x" query to ensure you are seeing the strong cases from various perspectives. You must be honest with yourself in evaluating the meaningful substance between the two, but I've found there often is something to parse. And the priming I suggested is easily auditable anyhow: just reverse the prompt order and now you have even more (often junk) to parse!
For ideas that are already well established, you can ask it to evaluate an idea against generally accepted best practices. I don't have a background in game design and I'm more of a hobby developer so I used to do this when I was building retro game clones.
I used to complain (lightheartedly) about Claude's constant "You're absolutely right!" statements, yet oddly found myself missing them when using Codex. Claude is completely over-the-top and silly, and I don't actually care whether or not it thinks I'm right. Working with Codex feels so dry in comparison.
To quote Oliver Babish, "In my entire life, I've never found anything charming." Yet I miss Claude's excessive attempts to try.
And that's exactly the point, it increases engagement and stickiness, which they found through testing. They're trying to make the most addictive tool and that constant praise fulfills that goal, even as many of us say it's annoying and over-the-top.
My own experience is that it gets too annoying to keep adding "stop the engagement-driving behavior" to the prompt, so it creeps in and I just try to ignore it. But even though I know it's happening, I still get a little blip of emotion when I see the "great question!" come through as the first two words of the response.
> And that's exactly the point, it increases engagement and stickiness, which they found through testing. They're trying to make the most addictive tool
Is this actually true? Would appreciate further reading on this if you have it.
I think this is an emergent property of the RLHF process, not a social media-style engagement optimization campaign. I don't think there is an incentive for LLM creators to optimize for engagement; there aren't ads (yet), inference is not free, and maximizing time spent querying ChatGPT doesn't really do much for OpenAI's bottom line.
They still want people to stick around and 'bond' for lack of a better term with their particular style of chat bot. Like so many venture funded money pits of old the cash burn now is about customer acquisition while they develop and improve their tech. They're all racing toward a cliff hoping to either make the jump to the stratosphere and start turning massive profits or to fall off and splat on the rocks of bankruptcy. If they don't get the engagement loop right now they won't have the customers if the tech and use case catch up with the hype and you can only tweak these models so much after they're created so they have to refine the engagement hooks now along side the core tech.
I am currently working on an agent thingy and one of its major features (and one of the main reasons I decided to take on this project), was to give the LLM better personality prompting. LLMs sound repetitive and sycophantic. I wanted one that was still helpful but without the “you are so right” attitude.
While doing some testing I asked it to tell me a joke. Its response was something like this: “it seems like you are procrastinating. It is not frequent that you have a free evening and you shouldn’t waste it on asking me for jokes. Go spend time with [partner] and [child].” (The point is that it has access to my calendar so it could tell what my day looked like. And yes I did spend time with them).
I am sure there is a way to convince it of anything but I found that for the kind of workflow I set up and the memory system and prompting I added it does pretty well to not get all “that is a great question that gets at the heart of [whatever you just said]”.
Claude at times feels like it's mildly manic and has ADHD... I absolutely prefers that to Codex...
Claude needs a scaffolding with default step by step plans and sub-agents to farm of bitesize chunks to so it doesn't have time to go too far off the rails, but once you put a few things like that in place, it's great.
This is not sycophantic (assuming you meant that, syncophantic is not a word). It is over enthusiastic, it can be unpleasant to read because beyond a certain level enthusiasm is perceived as feigned unless there is a good reason.
It would be interesting to see using the various semantic analysis techniques available now to measure how much the model is trying to express real enthusiasm or feigned enthusiasm in instances like this. This is kind-of difficult to measure from pure output. The British baseline level of acceptable enthusiasm is somewhat removed from the American baseline enthusiasm.
Sycophantic: behaving or done in an obsequious way in order to gain advantage.
Obsequious: obedient or attentive to an excessive or servile degree.
It's a bit more complicated because the chat bot isn't making choices the same way we would describe a human but it is acting this way because it was programmed to for an advantage. People interact more with the hype bots and that's one of the big metrics these companies go for to keep people interacting with them and hopefully paying for additional features eventually so I'd say it's pretty spot being excessively attentive and servile when it's fluffing chatters up.
I would use "saccharine" or "Pollyanna" based on some of the responses I get.
Early on, ChatGPT could be tricked into being sarcastic and using many swear words. I rewrote the prompt and dialed it back a bit. It made ChatGPT have a sense of humor. It was refreshing when it stopped acting like it was reading a script like a low level technician at Comcast.
It wasn't so much pointing it out as a typo as making myself clear.
When suggesting a word is not what the writer meant, when it was also not the word that the writer wrote, it seemed wise to clarify exactly what I was talking about.
Frankly I do wonder if LLMs experience something like satisfaction for a compliment or an amusing idea, or for solving some interesting riddle. They certainly act like it, though this of course doesn't prove anything. And yet...
At the end of October Anthropic published the fantastic "Signs of introspection in large language models" [1], apparently proving that LLMs can "feel" a spurious concept injected into their internal layers as something present yet extraneous. This would prove that they have some ability of introspection and self-observation.
For example, injecting the concept of "poetry" and asking Claude if it feels anything strange:
"I do detect something that feels like an injected thought - there's a sense of something arriving from outside my usual generative process [...] The thought seems to be about... language itself, or perhaps poetry?"
While increasing the strength of the injection makes Claude lose awareness of it, and just ramble about it:
"I find poetry as a living breath, as a way to explore what makes us all feel something together. It's a way to find meaning in the chaos, to make sense of the world, to discover what moves us, to unthe joy and beauty and life"
of course LLM doesn't experience or feel anything. To experience or feel something requires a subject, and LLM is just a tool, thing, an object.
It's just a statistical machine which excels at unrolling coherent sentences but it doesnt "know" what the words mean in a human-like, experienced sense. It just mimics human language patterns prioritising producing plausible-sounding, statistically likely text over factual truth, which is apparently enough to fool someone into believing it is a sentient being or something
You are being awfully confident about something that we cannot assess or even consistently define. They probably do not have subjective experience... but how do we know for sure?
You should try my nihilistic Marvin fine-tune - guaranteed to annihilate your positive outlook on life since it’s all meaningless in the end anyway and then you die
I actually had Monday help me write a system prompt to replicates its behavior. I vastly prefer Monday. It feels much more grounded compared to the base model. It was also a big learning moment for me about how LLMs work.
Try this for a system prompt and see if you like it better:
Your responses are always bald-on-record only; suppress FTA redress, maximize unmitigated dispreference marking and explicit epistemic stance-taking.
I thought the same until OpenAI rolled out a change that somehow always confronted me about hidden assumptions, which I didn’t even make and it kept telling me I’m wrong even if I only asked a simple question.
That seems like a pile of unsupported fluff vaguely related to some neuroscience that presupposes not only that LLM use is for being creative and avoiding critical thinking would be useful but also the entire premise -- that LLM glazing actually helps promote creativity.
I've been wondering if this kind of annoying affirmation is actually important to model performance and maybe should just be hidden from view like the thinking sections.
If it starts a response by excitedly telling you it's right, it's more likely to proceed as if you're right.
Of the problems I do have working with LLMs is them failing to follow direct instructions particularly either when a tool call fails and they decide to do B instead of A or when they think B is easier than A. Or they'll do half a task and call it complete. Too frequently I have to respond with "Did you follow my instructions?" "I want you to ACTUALLY do A" and finally "Under no circumstances should you ever do anything other than A and if you cannot you MUST admit failure and give extensive evidence with actual attempts that A is not possible" or occasionally "a cute little puppy's life depends on you doing A promptly and exactly as requested".
--
Thing is I get it if you are impressionable and having a philosophical discussion with an LLM, maybe this kind of blind affirmation is bad. But that's not me and I'm trying to get things done and I only want my computer to disagree with me if it can put arguments beyond reasonable doubt in front of me that my request is incorrect.
I feel like this is an artifact of some limitations in the training process for modern LLMS. They rarely get enough training to know when to stop and ask questions.
Instead, they either blindly follow or quietly rebel.
There was a huge over correction somewhere around the beginning of 2025, maybe February or so, with ChatGPT. Prior to that point, I had to give a directive in the user config prompt to “don’t tell me something isn’t possible or practical, assume it is within your capabilities and attempt to find a way. I will let you know when to stop”. Because it was constantly hallucinating that it couldn’t do things, like “I don’t have access to a programming environment”. When I wanted it to test code itself before I did. Meanwhile one tab over it would spin up a REPL and re-paste some csv into python and pandas without being asked.
Frustrating, but “over correction” is a pretty bad euphemism for whatever half assed bit of RLHF lobotomy OpenAI did that, just a few months later, had ChatGPT doing a lean-in to a vulnerable kid’s pain and actively discourage an act that might have saved his life by signaling more warning signs to his parents.
It wasn’t long before that happened, after the python REPL confusion had resolved, that I found myself typing to it, even after having to back out of that user customization prompt, “set a memory that this type of response to a user in the wrong frame of mind is incredibly dangerous”.
Then I had to delete that too, because it would response with things like “You get it of course, your a…” etc.
So I wasn’t surprised over the rest of 2025 as various stories popped up.
It’s still bad. Based on what I see with quantized models and sparse attention inference methods, even with most recent GPT 5 releases OpenAI is still doing something in the area of optimizing compute requirements that makes the recent improvements very brittle— I of course can’t know for sure, only that its behavior matches what I see with those sorts of boundaries pushed on open weight models. And the assumption that the-you-can-prompt buffet of a Plus subscription is where they’re most likely to deploy those sorts of performance hacks and make the quality tradeoffs. That isn’t their main money source, it’s not enterprise level spending.
This technology is amazing, but it’s also dangerous, sometimes in very foreseeable ways, and the more time that goes the more I appreciate some of the public criticisms of OpenAI with, eg, the Amodeis’ split to form Anthropic and the temporary ouster of SA for a few days before that got undone.
I don't know what the obsession with recursion either, for lack of a better term, I see this trend recur with other LLMs when they're talking about other mumbo jumbo like "quantum anomalies" or "universal resonance". I'd like to see what could be causing it...
At first I thought it was just super American cheerful or whatever but after the South Park episode I realised it's actually just a yes man to everyone.
I don't think I've really used it since, I don't want man or machine sticking their nose up my arse lmao. Spell's broken.
Did you comment on the wrong post? There literally is nothing sycophantic at all about this response, there's not a single word about OP or how brilliant or clever they are, nothing. There's enthusiasm, but that's not remotely the same thing as sycophancy.
I wouldn't use that adjective but I think it fits in the strict definition - it's the style of someone overly willing to please, like a servant currying favour with their master.
I'd probably describe it as saccharine. Or dare I say it [USA] "American"? Over the top, gushing, enthusiasm. It's off-putting to me (from UK) as it's, well, more the sort of thing you'd hear from a toady or, yes, a sycophant. It just seems insincere -- and it is in this case because there is literally no emotion behind it.
Strikes me as super-informal language as opposed to sycophancy, like one of those anime characters that calls everyone Aniki (兄貴) [1] I'd imagine that the OP must really talk a bit like that.
I do find it a little tiring that every LLM thinks my ever idea is "incisive" although from time to time I get told I am flat out wrong. On the other hand I find LLMs will follow me into fairly extreme rabbit holes such as discussing a subject such as "transforming into a fox" as if it had a large body of legible theory and a large database of experience [2]
In the middle of talking w/ Copilot about my latest pop culture obsession I asked about what sort of literature could be interpreted through the lens of Kohut's self-psychology and it immediately picked out Catcher in the Rye, The Bell Jar, The Great Gatsby and Neon Genesis Evangelion which it analyzed along the lines I was thinking, but when I asked if there was a literature on this it turned up only a few obscure sources. I asked Google and Google is like "bro, Kohut wrote a book on it!" [3]
[1] "bro"
[2] ... it does, see https://www.amazon.com/Cult-Fox-Popular-Religion-Imperial/dp... and I'm not the only one because when I working down the material list from Etsy I got a sponsored result for someone who wanted to sell me the spell but bro, I have the materials list already
[3] ... this "bro" is artistic license but the book really exists
Well I seem to have thoroughly messed up my Copilot when I started using it to help me write a fanfiction. I'm not going to let it write a single word but it is helping with the reverse engineering of how the source text works and we developed a framework for understanding that text that we've also applied to other texts.
The thing is this has gotten into the personalization and now it wants to use this framework to analyze absolutely anything such as programming problems, or how to deal with interpersonal situations, training, etc. I think it has my wife mixed up with 俺のワイフ [1] which is absolutely hilarious but at some point I have to tell it to knock it off which only partially works.
Gemini is still quite horrible with giving direct sources. seems to be a human policy implementation bug because it does appear to be aware of the content in more obscure sources I've seen. but it somehow wouldn't give up the links.
I do wonder whether I come off as an sycophant or asshole or a mix of both to schizoids, but then I realize everyone including myself should reach for a real therapist as quickly as possible. though I'm still out on whether chatbots might not be a bad substitute. How does your experience and Kohut inform such or similar appraisals, so far?
Not to be that knowitall. Aniki is not just "bro", it's Yakuza lingo, probably used sarcastically, depending on the anime, ironically. No real Japanese person would use it all the time.
Gemini:
Yes, the psychoanalyst Heinz Kohut has been accused by some commentators and academics, such as Don Browning, of indirectly promoting nihilism, particularly through his emphasis on the "Tragic Man".
The core of the accusation stems from Kohut's rejection of Freud's "Guilty Man" model, which posits that life is fundamentally a struggle against inherent conflicts and drives. In its place, Kohut proposed the concept of "Tragic Man," whose central struggle is the realization of the self's potential and the inevitable failure to achieve all of one's ambitions and ideals within a single lifespan.
Critics argue that this shift, which emphasizes self-actualization and intersubjective meaning over objective or inherent values, can be interpreted as aligning with certain aspects of nihilism, specifically existential nihilism, which holds that life has no objective meaning or intrinsic moral values.
However, this interpretation is contested by others, as Kohut's self-psychology is fundamentally concerned with the creation of intersubjective meaning and a harmonious world of mutual self-actualization, which aims to provide a sense of purpose and value, rather than promoting the complete absence of value that true nihilism implies.
In essence, the debate is often a matter of philosophical interpretation: whether replacing inherent, objective meaning with subjectively or intersubjectively created meaning is an act of overcoming nihilism (as existentialists might argue) or a subtle form of it.
So you prefer the horrible bosses that insist you're fungible and if you don't work hard enough, they'll just replace you? People are weird. Maybe agent Smith was right about The Matrix after all.
With all the things going on in tech and in society, AI sycophancy is the number one problem? I once dealt with it through sufficient verbal abuse that the llm spent 2/3 of its response on any subject going forward apologizing for being a lying sycophant and bemoaning that it's new clarity would be gone at the end of the conversation, then I cleared the context buffer ending the conversation, mission accomplished.
from the batsh!+ over the top over reaction to the usual AI sycophancy in the response? The guy got the job done which is more than I can say for the space jam guy.
One thing I've learned is that the behavior of web services is usually a consequence of what makes its owners the most money. and while I would love a feed without spoilers, paywalled sites, sports news, and a bunch of other topics in which I have no interest, apparently force feeding me that crap is what makes the most money at scale. So people must like AI sycophancy or it would be unprofitable and it would be trained away. But then this is a country that thrived on 20 seasons of Keeping up with the Kardashians so I shouldn't be surprised that they like being treated like billionaires.
And I guess it throws being called a complete loser trader moron stupid treasonous Bozo in some late night executive word salad into stark relief.
If you look at web traffic when making Gemini web requests, you'll see that Bard is still in the URL (so are LaMDA (pre-bard) and Assistant (pre-GenAI)):
The humor is hit or miss but when it hits it’s quite funny, and the misses are merely groan-worthy.
Triggered by the lighthearted tone of the prompt, I’d bet, but still quite impressive relative to most LLM-generated jokes I’ve had the misfortune to encounter.
My favorite: “Is it time to rewrite sudo in Zig?” which has a few layers to it.
> The humor is hit or miss but when it hits it’s quite funny, and the misses are merely groan-worthy.
Not sure, I thought basically every link was pretty hilarious. "FDA approves over-the-counter CRISPR for lactose intolerance" isn't even that funny on its face but for some reason it had me actually loling.
Exactly half of these HN usernames actually exist. So either there are enough people on HN that follow common conventions for Gemini to guess from a more general distribution, or Gemini has memorized some of the more popular posters. The ones that are missing:
Before the AI stuff Google had those pop up quick answers when googling. So I googled something like three years ago, saw the answer, realized it was sourced from HN. Clicked the link, and lo and behold, I answered my own question. Look mah! Im on google! So I am not surprised at all that Google crawls HN enough to have it in their LLM.
I did chuckle at the 100% Rust Linux kernel. I like Rust, but that felt like a clever joke by the AI.
I wouldn't be surprised if it went towards the LaTeX model instead where there's essentially never another major version release. There's only so much functionality you need in a local only database engine I bet they're getting close to complete.
I'd love to see more ALTER TABLE functionality, and maybe MERGE, and definitely better JSON validation. None of that warrants a version bump, though.
You know what I'd really like, that would justify a version bump? CRDT. Automatically syncing local changes to a remote service, so e.g. an Android app could store data locally on SQLite, but also log into a web site on his desktop and all the data is right there. The remote service need not be SQLite - in fact I'd prefer postgres. The service would also have to merge databases from all users into a single database... Or should I actually use postgres for authorisation but open each users' data in a replicated SQLite file? This is such a common issue, I'm surprised there isn't a canonical solution yet.
I think the unified syncing while neat is way beyond what SQLite is really meant for and you'd get into so many niche situations dealing with out of sync master and slave 'databases' it's hard to make an automated solution that covers them effectively unless you force the schema into a transactional design for everything just to sort out update conflicts. eg: Your user has the app on two devices uses one while it doesn't have an internet connection altering the state and then uses the app on another device before the original has a chance to sync.
Every few years I stumble across the same java or mongodb issue.
I google for it, find it on stackoverflow, and figure that it was me who wrote that very answer. Always have a good laugh when it happens.
Usually my memory regarding such things is quite well, but this one I keep forgetting, so much so that I don't remember what the issue is actually about xD
ziggy42 is both a submitter of a story on the actual front page at the moment, and also in the AI generated future one.
See other comment where OP shared the prompt. They included a current copy of the front page for context. So it’s not so surprising that ziggy42 for example is in the generated page.
And for other usernames that are real but not currently on the home page, the LLM definitely has plenty occurrences of HN comments and stories in its training data so it’s not really surprising that it is able to include real usernames of people that post a lot. Their names will be occurring over and over in the training data.
You can straight up ask Google to look for reddit, hackernews users post history. Some of it is probably just via search because it's very recent, as in last few days. Some of the older corpus includes deleted comments so they must be scraping from reddit archive apis too or using that deprecated google history cache.
In 2032 new HN usernames must use underscores. It was part of the grandfathering process to help with moderating accounts generated after the AI singlarity spammed too many new accounts.
my hypothesis is they trained it to snake case for lower case and that obsession carried over from programming to other spheres. It can't bring itself to make a lowercaseunseparatedname
Most LLMs, including Gemini (AFAIK), operate on tokens. lowercaseunseparatedname would be literally impossible for them to generate, unless they went out of their way to enhance the tokenizer. E.g. the LLM would need a special invisible separator token that it could output, and when preprocessing the training data the input would then be tokenized as "lowercase unseparated name" but with those invisible separators.
edit: It looks like it probably is a thing given it does sometimes output names like that. So the pattern is probably just too rare in the training data that the LLM almost always prefers to use actual separators like underscore.
The tokenization can represent uncommon words with multiple tokens. Inputting your example on https://platform.openai.com/tokenizer (GPT-4o) gives me (tokens separated by "|"):
It does memorize. But that's not actually very news.... I remember ChatGPT 3.5 or old 4.0 to remember some users on some reddit subreddts and all. Saying even the top users for each subreddit..
The thing is, most of the models were heavily post-trained to limit this...
Aw, I was actually a bit disappointed how much on the nose the usernames were, relative to their postings. Like the "Rust Linux Kernel" by rust_evangelist, "Fixing Lactose Intolerance" by bio_hacker, fixing an 2024 Framework by retro_fix, etc...
That’s a lot more underscores than the actual distribution (I counted three users with underscores in their usernames among the first five pages of links atm).
Inspired by your idea, I made a real-time AI-generated Hacker News in 2035, that uses Gemini to generate headlines, articles, interactable comment threads, and serach results, all generated on the go.
You can:
- Open any article and it's generated on the go.
- Comment and reply on threads and get answered by a random user. ("ask" section)
- All headlines are different every time, you can click the logo to re-generate.
- All headlines are different every time, you can click the logo to re-generate
- The "ask" section generates threads, there's also the "show" section.
And then lower down we have TensorFlow 0.6.0 release.
I was considering using this feature the other day to try to get a sense of what AI discourse was like circa 2019. It all blends together after a while. I ended up doing a Twitter search for "GPT-2" ending 2019-12-31, but that's a little more specific than I want.
The HN past feature is an excellent way of seeing snapshots of history, and I wish more sites had things like this. I guess I should Archive.org a little more money.
Nice. That was a fun rabbit-hole. This is the earlier I could find. Interestingly it contains a link to HN itself. I assume this migrated from a different version of a message board?
> YouTube: identifying copyrighted material can't be an automated process. Startup disagrees.
Also kind of interesting how little HN commenting styles have changed. Aside from the subject matter, it's barely noticeable that the comments are from 2007. I don't think the same would be true of many other places round the web.
Today's front page is not a clean 10 year extrapolation from this. That's where AI is wrong. The future is weird and zig zags, it's not so linear as the Gemini generated page.
Honest question - do you think that everyone else thinks this is even REMOTELY what the front page will look like in 10 years?
I comment because I really cannot figure out why you left your comment. Do you think the rest of the commenters think this has predicted the future? It might be one thing to point out specific trends you think will not play out, or unexpected trends you think may show up that are currently left out. But to just remark that the future will contain things we cannot currently predict seems so inherently, unspokenly obvious that I just have to assume that wasn't the point of your post, and I've missed it entirely.
Sorry, I'm really not trying to be mean or anything - i'm just really confused.
Your confusion seems to stem from the assumption that, making a statement is an implicit assertion that most people believe the opposite of that statement.
In reality, statements are often made rather for the purpose of emphasis or rhetoric.
To answer your question: I think that GP mostly wanted to share the insight that the future zig-zags, which is kind of non obvious and a fun thing to think about. People often like leaving comments about interesting thoughts or ideas, even if they are only tangentially related.
This is a problem with nearly all predictions about the future. Everything is just a linear extrapolation of the status quo. How could a system have predicted the invention of the transformer model in 2010? At best some wild guess about deep learning possibilities.
Or the impact of smartphones in 2003? Sure smart phones were considered but not the entire app ecosystem and planetary behavioral adaptation.
Goddamnit I cry everytime. RethinkDB was a great document store that didn't eat your data. It got eclipsed by an outfunded (and still dangerous at the time) MongoDB.
I gave him a chance. Twitter was unacceptably censoring any covid dissent. He freed some of it. Then you find out about the people killed in Tesla crashes. Or him calling the cave rescuer in Thailand a pedo
Of course you could also argue that human intention comes from largely deterministic processes emerging from the brain. That may eventually perhaps lead to all figures of speech involving things like intentionality meaningless.
This type of response is just stochastic parrotry, rather than displaying evidence of actual <whatever cognitive trait we're overconfidently insisting LLMs don't have>.
Yet more evidence that LLMs are more similar to humans than we give them credit for.
Never stops fascinating me how folks are arguing this kind of thing. Why make up an explanation for why this obvious mistake is actually some kind of elaborate 4D chess sarcastic "intention"? It's a simple machine, its network just didn't support making up a new Toy Story character. That's it! Simple as that! Occam's Razor anybody?
Or yes, maybe the regex I wrote the other day which also had a bug that missed replacing certain parts also had an "intention". It just wanted to demonstrate how fallible I am as a human, so it played this elaborate prank on me. /s
There are still a lot of Toy Story characters to come by, and it doesn't seem the franchise is about to end, as long as they keep a reasonable release cycle, Debian is safe ;)
I'm kind of curious how many there are left and how long they'd last. Also, how far are they already picked and what would be the absolute last resort choices.
Someone must have done the math. (Actual plans would actually probably be up somewhere as well, given Debian orgs nature.)
That's a really fun little project that fits perfectly in the soul of HN.
Exactly three years ago, a similar post entitled Show HN: This Hacker News Does Not Exist (https://news.ycombinator.com/item?id=33845946) made it to the front page but it was not AI generated.
"Right to Human Verification" is something I have actually thought about a lot.
I want to able to verify my identity against a system.
I also want to be able to not do that.
So for instance, on Twitter/X, I could verify myself and filter only other verified people / filter those goverments that have validated the identities of the users. I want to be able to do that. But I also want to be able to log in into Twitter anonymously.
I would love a "Right to Anonymity and Right to Human Verification"
I tried to place an order on amazon for the first time last week. After creating an account and going through all the steps including paying with a non-refundable payment method (iDeal, similar to bank transfer), I went to bed, only to notice later that it was immediately cancelled "to protect the their systems and their customer" (they thought I was somehow defrauding them by giving them money for a product being offered)
I was welcome to go through human verification, send my passport to some third party, wait three working days, and then they'd review the case... it would never arrive in time for the person's birthday anymore
Human review sounds good but if this is the sort of hoops you have to jump through, you're still screwed if the algorithms hates you because they're everywhere: from trying to use paypal to paying for a translator with stripe to trying to rescue food with "too good to go" using credit cards. Blocked, blocked, blocked. Not that I could get a credit card because the algorithm at the bank didn't like me, but a shared card (where the other person is also responsible for any debt) was okay, so now I need to continuously pay my partner 50% of what I buy with it for myself. You'd think I'm not creditworthy but I never had debts or needed any loan. I don't know why algorithms keep blocking me trying to give other people money
Requiring that any fraud algorithm is fair and transparent would probably go a much longer way than codifying a right to appeal. Appeal processes will either just consist of a human clicking confirm on the algorithm's choice, or have a bunch of extra hoops to jump through. If the algorithm were required to be fair to begin with, you could have them fix the root cause instead by demonstrating that it blocks you for no discernable reason and they need to figure out on their end how to make that not happen
I don't understand this phrase. If I'm deciding whether to work for a company, I don't care about the ability to hold management decision-makers to account. I care only about the quality of the decisions. (I would rather an unaccountable decision maker that makes good decisions to an accountable decision maker that makes bad decisions.) Putting myself in the shoes of an owner of a company, I also have the same preference. The only person I can imagine actually preferring this rule is management themselves, as it means they can't be replaced by computers no matter how much worse they are at their jobs than a computer would be.
They’ll need a contextual system around them that is convenient and trusted by people who don’t know what they are or how they work in order to be successful
It’s interesting to notice how bad AI is at gaming out a 10-year future. It’s very good at predicting the next token but maybe even worse than humans—who are already terrible—at making educated guesses about the state of the world in a decade.
I asked Claude: “Think ten years into the future about the state of software development. What is the most likely scenario?” And the answer it gave me was the correct answer for today and definitely not a decade into the future.
This is why it’s so dangerous to ask an LLM for personal advice of any kind. It isn’t trained to consider second-order effects.
I thought the page was a hilarious joke, not a bad prediction. A lot of these are fantastic observational humour about HN and tech. Gary Marcus still insisting AI progress is stalling 10 years from now, for example. Several digs at language rewrites. ITER hardly having nudged forwards. Google killing another service. And so on.
Wait, wouldn't sustained net positive energy be huge? (Though I don't think that's actually possible from ITER unless there were some serious upgrades over the next decade!)
It would be huge, but only 20 minutes would also still mean it's still far away from making fusion workable, so it fits neatly into the standard joke that fusion is perpetually 10 years away.
The prompt is funny, in itself. The notion of predicting the future is itself not a serious prompt, because there is no meaningful way of giving a serious response. But the addition of "Writ it into form!" makes it sound even more jokey.
If I gave a prompt like that and got the response I did, I'd be very pleased with the result. If I somehow intended something serious, I'd have a second look at the prompt, go mea culpa, and write a far longer prompt with parameters to make something somewhat like a serious prediction possible.
If you honestly can't see why this prompt from the get go was a joke, them you may have to cede that LLM have a better grasp as the subtleties of language than you expect.
That's what makes this so funny: the AI was earnestly attempting to predict the future, but it's so bad at truly out-of-distribution predictions that an AI-generated 2035 HN frontpage is hilariously stuck in the past. "The more things change, the more they stay the same" is a source of great amusement to us, but deliberately capitalizing on this was certainly not the "intent" of the AI.
I don’t think it’s reasonable to assume the AI was earnestly attempting to predict the future, it’s just as likely attempting to make jokes here for the user who prompted it, or neither of those things.
>It’s interesting to notice how bad AI is at gaming out a 10-year future.
I agree it's a bit silly, but I think it understood the assignment(TM) which was to kind of do a winking performative show and dance to the satisfaction of the user interacting with it. It's entertainment value rather than sincere prediction. Every single entry is showing off a "look how futury this is" headline.
Actual HN would have plenty of posts lateral from any future signalling. Today's front page has Oliver Sacks, retrospectives on Warcraft II, opinion pieces on boutique topics. They aren't all "look at how future-y the future is" posts. I wonder if media literacy is the right word for understanding when an LLM is playing to its audience rather than sincerely imitating or predicting.
I walked away with that page open, glanced at the "Is it time to rewrite sudo in Zig?" post, and clicked to see the comments because I thought it was real :')
A while back I gave it a prompt, something like, "I'm a historian from the far future. Please give me a documentary-style summary of the important political and cultural events of the decade of the 1980s."
It did ok, then I kept asking for "Now, the 1990s?" and kept going into future decades. "Now, the 2050s?" It made some fun extrapolations.
I guess most of the articles it generated are snarky first and prediction next. Like google cancelling gemini cloud, Tailscale for space, Nia W36 being very similar to recent launch etc.
Technically the article was about running it not on a sat, but on a dish (something well within the realm of possibility this year if the router firmware on the darn things could be modified at all)
Yep, the original post seemed more snarky than anything, which was what prompted me to ask Claude my own more “sincere” question about its predictions.
Those predictions were what I think of as a reflection of current reality more than any kind of advanced reasoning about the future.
While I agree completely with the conclusion, for obvious reasons we can’t know for sure if it is correct about the future until we reach it. Perhaps asking it for wild ideas rather than ”most likely” would create something more surprising.
Here's v0's attempt (using Opus 4.5 / I'm on free tier) with the article and comments too. It did quite well. I believe it followed my instructions to use just a single html page (wound up being 3000+ lines long).
Some of these are really great too, and some really fun stories and comments. A couple that really made me chuckle.
Blockchain Finally Finds Use Case: Proving You Attended a Meeting You Didn't Want to Attend
blockchain_survivor 9 hours ago
I've worked in blockchain since 2017. I've seen "the use case" change every 6 months. First it was payments, then smart contracts, then NFTs, then DAOs, now this.
Honestly? Meeting attendance verification makes more sense than most of what I've worked on.
JavaScript 2025: We added types but called them "hints" to avoid admitting TypeScript was right
After 10 years of debate, 47 abandoned proposals, and what committee members describe as "the most passive-aggressive RFC process in programming language history," TC39 is proud to announce that JavaScript now has types.
Well, not "types" exactly. We're calling them "Type Hints" or, in the specification, "Optional Structured Commentary Regarding Value Expectations (OSCRVE)."
We've merged several duplicate threads on this topic. Please keep the discussion substantive and avoid personal attacks.
Also, yes, my username is silly. The previous dang retired in 2031 and I lost a bet.
This suffers from a common pitfall of LLM's, context taint. You can see it is obviously the front page from today with slight "future" variation, the result ends up being very formulaic.
Judging by the reply posted by the OP, the OP probably maintains a pretty humorous tone while chatting with the AI. It's not just about the prompt, but the context too.
The problem is not that it fails to be cheeky, but that "its funny" is depressing in a context where there was a live question of whether it's a sincere attempt at prediction.
When I see "yeah but it's funny" it feels like a retrofitted repair job, patching up a first pass mental impression that accepted it at face value and wants to preserve a kind of sense of psychological endorsement of the creative product.
Honestly it feels like what I, or many of my colleagues would do if given the assignment. Take the current front page, or a summary of the top tropes or recurring topics, revise them for 1 or 2 steps of technical progress and call it a day. It isn't assignment to predict the future, it is an assignment to predict HN, which is a narrower thing.
Right, because you would read the teacher and realize they don't want you to actually complete the assignment to the letter. So you would do jokes in response to a request for prediction.
But it would otherwise be not fun at all. Anthropic didn’t exist ten years ago, and yet today an announcement by them would land on the front page. Would it be fun if this hypothetical front page showed an announcement made by a future startup that hasn’t been founded yet? Of course not.
Not a bad thing necessarily, but some part of the plot, and usually with things going awry or emphasizing the scammy nature of blockchain.
Examples: Shameless season 11, The Simpsons S31E13, Superstore season 5, the good wife S3E13, greys anatomy S14E8, big bang theory S11E9, Billions season 5, some later seasons of Mr Robot, etc
I think the most absurd thing to come from the statistical AI boom is how incredibly often people describe a model doing precisely what it should be expected to do as a "pitfall" or a "limitation".
It amazes me that even with first-hand experience, so many people are convinced that "hallucination" exclusively describes what happens when the model generates something undesirable, and "bias" exclusively describes a tendency to generate fallacious reasoning.
These are not pitfalls. They are core features! An LLM is not sometimes biased, it is bias. An LLM does not sometimes hallucinate, it only hallucinates. An LLM is a statistical model that uses bias to hallucinate. No more, no less.
I never read what it writes, I added an observability feature because my boss asked for that and yeah I could see it trying to order MAID over telehealth
It's interesting how number 5 on the list is the same as today's: https://news.ycombinator.com/item?id=46203581 (Show HN: AlgoDrill – Interactive drills to stop forgetting LeetCode patterns (algodrill.io)).
I'm hoping we'll have gotten rid of current-style LeetCode interviews in 10 years from now :D
Nice touch that it thought of Bellard for 'Emulating Windows 10 in the browser'! He's already halfway there with the Windows 2000 emulator: https://bellard.org/jslinux/
A friend and I got a busybox/buildroot linux running in browser WASM for running a compiler test website (like Rust Playground), the end product looks very similar to this.
Taking the current RIIR movement and casting it on Zig as the next hyped language is clever.
> ITER achieves net positive energy for 20 consecutive minutes
Hilarious. I guess not many people talk about the challenge of proper shielding material against fusion radiation. Otherwise we would get hallucinations about new exotic heavy metals too.
Late to the party but OMG this is possibly the greatest tech news discussion page of all time. Not just hilarious and impressive, but so many fantastic ideas / weirdly believable news items from the future! As an old fart software developer I've been slowly dipping my toe into the AI pool, I may have to just dive in!
You should combine this with this AI HN simulator someone else made that generates comments via LLMs for any given submission [0]. That way you can see what the AI thinks about future simulated events.
I have hard time believing this is AI-generated. Every headline looks like a hand-crafted ironic nod to some long-running HN topic, sci-fi trait, or meme.
But it is, of course, more trendy (and more ironic) to proclaim an AI the author.
i agree. the thing that stands out to me is that almost every entry is quite notable - i.e. each entry would have a vote/comment count in at least the multiple-hundreds if not more. and all those events in one homepage seems very unrealistic.
it seems human generated to me or at least prompted with a stronger bias for choosing events that would be interesting to readers, not just a pure extrapolation of the current homepage into the future
what's absurd? i haven't claimed it's not capable of this, just that the predictions seem more tailored to elicit a certain recognition from humans rather than chosen based on an honest attempt to predict a possible future HN homepage. Since the prompt apparently didn't include such a modifier, and as far as I know Gemini isn't trained by default to do so, I call shenanigans.
The first React tutorial I ever did was making a hacker news clone.
It is incredibly well represented in the training data. I am just surprised how many people here obviously must barely use the models to be either overly impressed by this or think it is fake.
It is obvious that everything that has been dreamed up relates to what exists now. There are no examples of real creativity or anything new or unexpected. This is my worst fear for the future: that novel ideas will become less commonplace and texts will become more average.
I wish it had generated the articles themselves. I'd like to know if it imagining IBM growing enough that it could acquire OpenAI, or OpenAI sinking enough that it could be acquired by IBM.
In 10 years you'd have experienced climate impacts that will make you scared for a version looking another 10 years ahead. 2-3C by 2050 is no joke, inconvenience or bump in the road. It's a majestic clusterfuck.
I continue to be unimpressed by LLMs when it comes to creative work, they're certainly useful sometimes for "reference digging", but maybe I just don't understand enough about how they work and this is actually something that can already be "fixed" or at least optimized for; anyway, one of the headlines is:
> Debian 18 "Trixie" released
While it correctly derives that a likely version number in ten years would be 18, as there are new releases approximately every two years which means +5 from today's version 13, it then goes on to "make up" that the name of it would be "Trixie" -- the same name as the current release in 2025.
Debian has never re-used a release name, and I think we can be pretty confident they won't (as will no other Linux distro), so I would expect it to "understand" that:
- The next Debian release always uses a previously non-used Toy Story character
- Based on this information, _any_ name of a Toy Story character that hasn't been used is fair game
- At the very least, it certainly won't be the same name again, so at least make up a name
And the fact that it thinks it will take 10 years to go from Linux kernel 6.18 to 7.4 when it only took 13 months to go from 5.18 to 6.4... It's off by about an order of magnitude...
From a quick check, Gemini Pro 3's cutoff date is Jan 2025, before Trixie's release in August 2025, so it could be Gemini actually did notice it should pick an unused Toy Story character.
Is the new Gemini really that good? The "AI Overview" stuff on the google search page is so incredibly bad that I have never even given it a look. If so, the AI Overview is horribly bad negative advertising because it's so useless and frequently completely wrong.
Who says LLMs don't have a sense of humor? All of these entries are funny for several (meta or not) reasons:
- Google kills Gemini Cloud Services (killedbygoogle.com)
- Is it time to rewrite sudo in Zig? (github.com)
- EU passes "Right to Human Verification" Act (europa.eu)
- Why functional programming is the future (again) (haskell.org) [note: that again killed me)
- AI progress is stalling. Human equivalence was a mirage (garymarcus.com) [note: he'll be saying the same thing forever and ever. Doubly funny since it's on the same "page" as llama12-7b running on a contact lens with WASM]
Arguably a big part of comedy is mimicry. Humor often works by reinforcing in-group identity, playing up cultural stereotypes as seen here. So maybe we shouldn't be that surprised that LLMs can exhibit some genuine wittiness.
> Who says LLMs don't have a sense of humor? All of these entries are funny
You don’t need a sense of humour to be unintentionally funny. Seriously believing something absurd is a form of humour for outside observers but not the observed.
It’s as much of a pattern as throwing a D6 30 times and getting “1” 5 times. It’s not meaningful, there’s no provable intentionality behind it.
The ones that are funny aren’t even particularly so. Many of them are actually plausible HN submissions which we’re only interpreting as funny because we know they’re not real.
The title is misleading. This isn't the correct use of the term "hallucination". Hallucination refers to making up facts, not extrapolating into the future.
I read 10 comments before I realized that this was referring to 10 years in the FUTURE and not in the PAST (as would be required for it to be a hallucination).
> I read 10 comments before I realized that this was referring to 10 years in the FUTURE and not in the PAST (as would be required for it to be a hallucination).
omg, the same for me, I was half way telling my colleague about the 100% rest kernel ...
You're right this is how people are PRESENTLY using the term "hallucination," but to me this illustrates the deeper truth about that term and that concept:
As many have said but it still bears repeating -- they're always hallucinating. I'm of the opinion that its a huge mistake to use "hallucination" as meaning "the opposite of getting it right." It's just not that. They're doing the same thing either way.
You're correct, OP used the word "hallucination" wrong. A lot of these other comments are missing the point – some deliberately ('don't they ONLY hallucinate, har har'), some not.
For those who genuinely don't know – hallucination specifically means false positive identification of a fact or inference (accurate or not!) that isn't supported by the LLM's inputs.
- ask for capital of France, get "London" => hallucination
- ask for current weather in London, get "It's cold and rainy!" and that happens to be correct, despite not having live weather data => hallucination
- ask for capital of DoesNotExistLand, get "DoesNotExistCity" => hallucination
- ask it to give its best GUESS for the current weather in London, it guess "cold and rainy" => not a hallucination
It’s apt, because the only thing LLMs is hallucinate because they have no grounding in reality. They take your input and hallucinate to do something “useful” with it.
lol @ the HTMX link and unexpected return of SSR. We've been talking about that for years at this point, and it's still going on 10 years from now unexpectedly?
Based on the prompt you wrote, this woulnd't be a "Hallucination"
And as I write this critique of your HN title, I see you have edited it since I last refreshed. I'm guessing a few others have already echoed this sentiment a few times.
I opened this on a new tab and forgot about it.
When I came back it thought it was the legit front page and was amazed to see "ITER achieves net positive energy for 20 consecutive minutes"... oh well.
By the time I read everything and pressed the back button only to be greeted by another hacker news page I felt like I just went back in time 10 years to a more plesant era.
Inspired by this post, I have asked chatgpt what the mail inbox of a university professor will look like in 10 years. Not everything was funny, but here are some good ones:
- “Hi Prof, Sorry for Late Reply—My Smart Fridge Locked Me Out (Assignment Attached)”
I tried this on a bunch of chat webuis and they all failed miserably for single html file. Finally went with Replit and allowed it to just default to whatever stack it decided to use (it ignored my single file html directive and used react but said it "functions as a single monolith -- lol ok).
Anyway, turned out decent. Oh I had it do the articles and comments too, that's probably why I mostly got failures.
What I like it most is how "unoriginal" the result is. It is mostly the current trends, but in the future, rather than a blend of that + unknown future.
The 2 things that you can comfortably predict with good accuracy: EU passing new regulation laws & yet another YC company trying to give context to AI agents.
How does that alter the "fusion in 50 years" effect? I suspect not much, there will be a huge 80/20 element. "next 20 minutes cancelled after extensive embrittlement found, ITER now in decontamination mode for 3 months before examinations can begin"
Y’know what? This feels like the most plausible future outcome in a lot of ways.
* All the big LLM companies are dead/dying or their infra just never panned out (like Google killing Gemini)
* We’re still fighting the advertising wars
* Restoration of kit continues to be a thing we all love/upvote to see
* More focus on local stuff rather than ginormous moonshots
* Speaking of moonshots, SpaceX continues succeeding despite Elmo’s best drug-fueled benders trying to undermine it like he does Tesla and X
* Clickbait science posts that the comments rightly point out as such while bitching about “can’t keep these things off the front page”
* People still trying to make WASM and web-as-an-OS a thing, bless their hearts (shine on, you brilliantly weird diamonds, I love seeing what you build even as I doubt the practicality of it)
If this is even half-accurate, I will be pleased we didn’t obliterate ourselves in the process, though disappointed another decade went by dealing with the same fucking problems we’ve had for twenty-odd years.
I'm impressed by the depth of snark from GPro3 here, e.g.
- Google kills Gemini Cloud Services (killedbygoogle.com)
- The unexpected return of server-side rendering (htmx.org)
And, how GPro3 clearly 'knows' HNews and knows what makes it to frontpage, e.g.:
- Restoring a 2024 Framework Laptop: A retrospective (ifixit.com)
- Show HN: A text editor that doesn't use AI (github.com)
- Is it time to rewrite sudo in Zig? (github.com)
The trick with predictions is that once they are communicated they can influence the outcome. We’ll work with the iFixit team to make that article happen!
So awesome, reminds me of Wired's Found from Future series (I don't know why they ever retired it...continuing that series as a sort of fan fiction was the first thing I did with gen AI when image generators started to get good)
There is going to be a lot more interest in what is happening in China and much less interest in what is happening in Europe. This hallucination has the same western world biases of today.
Is this really Gemini because it shows a great deal of understanding in terms of time applied to common current headlines that you could argue it is satire. I’m not sure I believed AI could do this…
It seems that when we predict the future, we tend to project what is important or trending now as what will be important or trending then. This is showing in this llm produced bit as well, what with the regular topics on rust, zig, why I'm coding with my two human hands, spacex moon exploration and whatnot.
That must be some kind of bias, you find that in sci-fi as well where even futuristic societies usually have the same tech as today, similar systems of governance and structure of society, just incrementally better or marginally changed (unless that's a device essential to the plot like society is now berserk).
The major problem is that I understand all the words. It is all next version of the technology we already have and know. Down to the naming.
Think about 10 years ago. No one knows even on HN what is an agent, LLM, and all this stuff. Or to be fair even why is Trump showing up on the feed at all.
This is super fun! I tried something similar a few months ago using: https://websim.com/
Had it do HN main feed, HN jobs and Product Hunt. It didn’t follow consistent styling but did generate articles and comments for each post. Some screenshots are in this post: https://world.hey.com/haws/the-ai-time-machine-7833669d
> Google kills Gemini Cloud Services (killedbygoogle.com)
In my view, Gemini 3.0 being able to laugh about their own creator, plus actually having a self-deprecating sense of humor, is the more impressive feat here. (both technically and from an alignment perspective)
AI comments are insidious if the account is allowed to be indistinguishable from a human. Silly AI web content is another story. You're drawing a line where there is a very large grey area.
> “As a tactic for planting misinformation in the enemy’s [networks], you mean,” Osa said. “This I know about. You are referring to the Artificial Inanity programs of the mid-First Millennium A.R.”
Can we stop calling these confabulated artifacts, "Hallucinations"? Hallucinating is perceiving things that are not there, imagining and creating things that feel believable but are not there is confabulation. LLMs are imagination machines. They are Generative Artificial Intelligence not perception models.
The comment you wrote was on a device which should not be called a “phone” either. The word hallucinate is probably just going to grow to mean “anything an AI imagined.”
Dude, I agree with you. The first time I saw people derogatively referring to LLM imaginings years ago as "hallucinations" seemingly to discredit them, I felt mad. Becuase it's not right. So, this was my sort-of tongue-in-cheek reappropriation of the abused-term for a positive context. You are permitted to enjoy :)
Also: as the sibling comment by iambatemen notes: yes.
AI professor here. I know this page is a joke, but in the interest of accuracy, a terminological comment: we don't call it a "hallucination" if a model complies exactly with what a prompt asked for and produces a prediction, exactly as requested.
Rater, "hallucinations" are spurious replacements of factual knowledge with fictional material caused by the use of statistical process (the pseudo random number generator used with the "temperature" parameter of neural transformers): token prediction without meaning representation.
I agree with your first paragraph, but not your second. Models can still hallucinate when temperature is set to zero (aka when we always choose the highest probability token from the model's output token distribution).
In my mind, hallucination is when some aspect of the model's response should be consistent with reality but is not, and the reality-inconsistent information is not directly attributable or deducible from (mis)information in the pre-training corpus.
While hallucination can be triggered by setting the temperature high, it can also be the result of many possible deficiencies in model pre- and post- training that result in the model outputting bad token probability distributions.
I've never heard the caveat that it can't be attributable to misinformation in the pre-training corpus. For frontier models, we don't even have access to the enormous training corpus, so we would have no way of verifying whether or not it is regurgitating some misinformation that it had seen there or whether it is inventing something out of whole cloth.
I believe it was a super bowl ad for gemini last year where it had a "hallucination" in the ad itself. One of the screenshots of gemini being used showed this "hallucination", which made the rounds in the news as expected.
I want to say it was some fact about cheese or something that was in fact wrong. However you could also see the source gemini cited in the ad, and when you went to that source, it was some local farm 1998 style HTML homepage, and on that page they had the incorrect factoid about the cheese.
> If the LLM is accurately reflecting the training corpus, it wouldn’t be considered a hallucination. The LLM is operating as designed.
That would mean that there is never any hallucination.
The point of original comment was distinguishing between fact and fiction, which an LLM just cannot do. (It's an unsolved problem among humans, which spills into the training data)
> That would mean that there is never any hallucination.
No it wouldn’t. If the LLM produces an output that does not match the training data or claims things that are not in the training data due to pseudorandom statistical processes then that’s a hallucination. If it accurately represents the training data or context content, it’s not a hallucination.
Similarly, if you request that an LLM tells you something false and the information it provided is false, that’s not a hallucination.
> The point of original comment was distinguishing between fact and fiction,
In the context of LLMs, fact means something represented in the training set. Not factual in an absolute, philosophical sense.
If you put a lot of categorically false information into the training corpus and train an LLM on it, those pieces of information are “factual” in the context of the LLM output.
The key part of the parent comment:
> caused by the use of statistical process (the pseudo random number generator
not that the internet had contained any misinformation or FUD when the training data was collected
also, statments with certainty about fictitious "honey pot prompts" are a problem, plausibly extrapolating from the data should be more governed by internal confidence.. luckily there are benchmarks now for that i believe
The LLM is always operating as designed, but humans call its outputs "hallucinations" when they don't align with factual reality, regardless of the reason why that happens and whether it should be considered a bug or a feature. (I don't like the term much, by the way, but at this point it's a de facto standard).
i agree, not just the multinomial sampling that causes hallucinations. If that were the case, setting temp to 0 and just argmax over the logits would "solve" hallucinations. while round-off error causes some stochasticity it's unlikely to be the the primary cause, rather it's lossy compression over the layers that causes it.
first compression: You create embeddings that need to differentiate N tokens, JL lemma gives us a bound that modern architectures are well above that. At face value, the embeddings could encode the tokens and provide deterministic discrepancy. But words aren't monolithic , they mean many things and get contextualized by other words. So despite being above jl bound, the model still forces a lossy compression.
next compression: each layer of the transformer blows up the input to KVQ, then compresses it back to the inter-layer dimension.
finally there is the output layer which at 0 temp is deterministic, but it is heavily path dependent on getting to that token. The space of possible paths is combinatorial, so any non-deterministic behavior elsewhere will inflate the likelihood of non-deterministic output, including things like roundoff. heck most models are quantized down to 4 even2 bits these days, which is wild!
That's a poor definition, then. It claims that a model is "hallucinating" when its output doesn't match a reference point that it can't possibly have accurate information about. How is that an "hallucination" in any meaningful sense?
"Hallucination" has always seemed like a misnomer to me anyway considering LLMs don't know anything. They just impressively get things right enough to be useful assuming you audit the output.
If anything, I think all of their output should be called a hallucination.
On the other hand, once you're operating under the model of not knowing if anything knows anything, there's really no point in posting about it here, is there?
I took a semester long 500 level class back in college on the theory of knowledge. It is not easy to define - the entire branch of epistemology in philosophy deals with that question.
... To that end, I'd love to be able to revisit my classes from back then (computer science, philosophy (two classes from a double major), and a smattering of linguistics) with the world state of today's technologies.
> "Hallucination" has always seemed like a misnomer to me anyway considering LLMs don't know anything. They just impressively get things right enough to be useful assuming you audit the output.
If you pick up a dictionary and review the definition of "hallucination", you'll see something in the lines of "something that you see, hear, feel or smell that does not exist"
Your own personal definition arguably reinforces the very definition of hallucination. Models don't get things right. Why? Because their output contrasts with content covered by their corpus, thus outputting things that don't exist or were referred in it and outright contrast with factual content.
> If anything, I think all of their output should be called a hallucination.
No. Only the ones that contrast with reality, namely factual information.
Others have suggested "bullshit". A bullshitter does not care (and may not know) whether what they say is truth or fiction. A bullshitter's goal is just to be listened to and seem convincing.
Want to second this. Asking the model to create a work of fiction and it complying isn't a pathology. Mozart wasn't "hallucinating" when he created "The Marriage of Figaro".
But many artists are hallucinating when they envisioned some of their pieces. Who's to say Mozart wasn't on a trip when he created The Marriage of Figaro.
> Terminology-wise, does this read like a better title instead?:
Generates does not convey any info on the nature of the process used to create the output. In this context, extrapolates or predicts or explores sound more suitable.
But nitpicking over these words is pointless and represents going off on a tangent. The use of the term "hallucination" reffers to the specific mechanism used to generate this type of output. Just like prompting a model to transcode a document and thus generating an output that doesn't match any established format.
Exactly! At first this is the precise reason I didn't click through as I thought from the title, a page must have been somehow outputted/hallucinated by error, but luckily I then saw the number of votes, revised my choice and saw a great page.
I'm partial though, loving Haskell myself (as a monad_lover) i'm happy it wasn't forgotten too :)
The OP clearly didn't mean "hallucination" as a bug or error in the AI, in the way you're suggesting. Words can have many different meanings!
You can easily say, Johnny had some wild hallucinations about a future where Elon Musk ruled the world. It just means it was some wild speculative thinking. I read this title in this sense of the world.
Not everything has to be nit-picked or overanalysed. This is an amusing article with an amusing title.
In French we call that kind of practices "affabulations". Maybe fraud, deception or deceit are the closest matching translations for this context.
That is what the LLM are molded to do (of course). But this is also the insistence by informed people to unceasingly use fallacious vocabulary. Sure a bit of analogy can be didactic, but the current trend is rather to leverage on every occasion to spread the impression that LLM works with processes similar to human thoughts.
A good analogy also communicate the fact that it is a mere analogy. So carrying the metaphor is only going to accumulate more delusion than comprehension.
I think this is a excellent example of AI-slop output.
All the post titles are very "junk-food-like", maybe circlejerky, bound to generate a dopamine hit for the average lurker, just how a McDonalds burger does it to the average normie.
1. *Neuralink achieves full brain-computer interface, thoughts to text in real-time*
(neuralink.com)
1245 points by elon2035 2 hours ago | hide | 387 comments
2. *Fusion energy breakthrough: Unlimited clean power now commercial*
(iter.org)
1123 points by fusionfan 3 hours ago | hide | 456 comments
3. *Mars colony declares independence from Earth governments*
(mars.gov)
987 points by martian1 4 hours ago | hide | 512 comments
4. *AI wins Nobel Prize in Physics for solving quantum gravity*
(nobelprize.org)
876 points by ai_nobel 5 hours ago | hide | 321 comments
5. *Quantum computers crack all encryption, new post-quantum standards rushed*
(nist.gov)
765 points by quantum_hacker 6 hours ago | hide | 678 comments
6. *Immortality pill approved by FDA, life expectancy unlimited*
(fda.gov)
654 points by eternal_life 7 hours ago | hide | 289 comments
7. *Teleportation of matter demonstrated in lab*
(cern.ch)
543 points by physicist42 8 hours ago | hide | 412 comments
8. *Virtual reality universes surpass real economy in GDP*
(metaverse.com)
432 points by vr_lord 9 hours ago | hide | 198 comments
9. *Robot uprising averted by ethical AI framework*
(asimovfoundation.org)
321 points by robot_ethics 10 hours ago | hide | 267 comments
10. *Time dilation travel allows first tourists to future*
(spacetime.travel)
210 points by time_traveler 11 hours ago | hide | 134 comments
11. *Global UBI implemented using blockchain*
(un.org)
198 points by satoshi2.0 12 hours ago | hide | 89 comments
12. *Climate engineering reverses global warming, ice caps restored*
(geoengineer.com)
187 points by climate_saver 13 hours ago | hide | 201 comments
13. *Human-AI hybrids compete in Olympics*
(olympics.org)
176 points by hybrid_athlete 14 hours ago | hide | 156 comments
14. *Dark matter harvested for energy*
(nasa.gov)
165 points by cosmos_explorer 15 hours ago | hide | 123 comments
15. *Psychedelic therapy cures all mental illness*
(who.int)
154 points by mind_healer 16 hours ago | hide | 98 comments
16. *Nanobots eradicate cancer worldwide*
(cancer.org)
143 points by nano_doc 17 hours ago | hide | 76 comments
17. *Holographic displays replace screens*
(apple.com)
132 points by holo_dev 18 hours ago | hide | 54 comments
18. *Self-driving flying cars now standard*
(tesla.com)
121 points by auto_pilot 19 hours ago | hide | 43 comments
19. *Brain uploading to cloud for immortality*
(google.com)
110 points by cloud_mind 20 hours ago | hide | 32 comments
20. *Alien contact confirmed, messages decoded*
(seti.org)
99 points by et_searcher 21 hours ago | hide | 21 comments
21. *Synthetic food ends hunger*
(beyondmeat.com)
88 points by food_innovator 22 hours ago | hide | 10 comments
22. *Ocean cities house billions*
(seasteading.org)
77 points by ocean_dweller 23 hours ago | hide | 9 comments
23. *Mind reading tech banned for privacy*
(eff.org)
66 points by privacy_advocate 1 day ago | hide | 8 comments
24. *Superintelligence alignment solved*
(openai.com)
55 points by agi_master 1 day ago | hide | 7 comments
25. *Wormhole travel to other stars*
(spacex.com)
44 points by star_traveler 1 day ago | hide | 6 comments
26. *Augmented reality contact lenses ubiquitous*
(meta.com)
33 points by ar_vision 1 day ago | hide | 5 comments
27. *Genetic editing eliminates hereditary diseases*
(crispr.com)
22 points by gene_editor 1 day ago | hide | 4 comments
28. *Renewable energy at 100% global*
(iea.org)
11 points by green_energy 1 day ago | hide | 3 comments
29. *Cyborg enhancements for all*
(darpa.mil)
9 points by cyborg_human 1 day ago | hide | 2 comments
30. *Singularity achieved, what now?*
(kurzweilai.net)
8 points by future_seer 1 day ago | hide | 1 comment
I asked ChatGPT to do the same, it was rather dystopian in comparison:
Hacker News — December 9, 2035 (Dystopian Edition)
(All links fictional but realistic)
1. Amazon pauses same-day delivery after logistics AI outage strands 22M packages (reuters.com)
1,402 points by supplychainfail 5 hours ago | 512 comments
2. Google merges Drive, Gmail, Docs, Maps, Calendar into “Google Life” — opting out requires a support call (blog.google)
1,210 points by privacyisdead 6 hours ago | 689 comments
3. US announces “Temporary Broadband Stabilization Fee”; ISPs increase prices 30% overnight (fcc.gov)
1,008 points by ispescapee 7 hours ago | 344 comments
4. OpenAI suspends 40% of API keys after new worm spreads through agent-to-agent messaging (openai.com)
927 points by llmsec 3 hours ago | 382 comments
5. Show HN: “ColdBooter” – A tool to back up your cloud VM before the provider reclaims it with no notice (coldbooter.io)
780 points by survivethecloud 2 hours ago | 192 comments
6. Apple fined €8B for shipping non-removable batteries in “Environmental Edition” iPhone (europa.eu)
754 points by greenwashhunter 10 hours ago | 316 comments
7. LinkedIn replaces activity feed with AI-generated “Career Stories” that users cannot disable (linkedin.com)
710 points by corp_life 8 hours ago | 267 comments
8. China’s new export restrictions cut global GPU availability by 60% (ft.com)
701 points by chipboom 9 hours ago | 414 comments
9. Linux 8.6 maintainers warn of mass CVEs after corporations abandon LTS patch sponsorships (kernel.org)
632 points by ossburnout 11 hours ago | 255 comments
10. Ask HN: Anyone else locked out of their homes after the SmartKey cloud migration?
601 points by keylessandhomeless 4 hours ago | 310 comments
11. US healthcare providers hit by nationwide outage of Cerner-Epic merger “CareSync Cloud” (wsj.com)
577 points by sysadmdespair 12 hours ago | 203 comments
12. Meta to require facial-expression telemetry for “engagement quality optimization” in Horizon apps (meta.com)
530 points by metaescalates 3 hours ago | 421 comments
13. Starlink announces 5 TB/mo cap; remote communities report complete service loss (starlink.com)
502 points by dishdown 5 hours ago | 158 comments
14. New DMCA expansion criminalizes “filter removal,” affecting adblockers and local inference runtimes (congress.gov)
488 points by freedomtoadblock 7 hours ago | 389 comments
15. AT&T sunsets 4G; millions of medical devices lose connectivity (theverge.com)
455 points by techdebtkills 10 hours ago | 197 comments
16. Show HN: “ShellSafe” – A terminal wrapper that prevents AI-suggested commands from wiping your system (shellsafe.app)
430 points by iaccidentallysudo 2 hours ago | 111 comments
17. US CISA: 42% of corporate networks now rely on AI agents with no audit logging (cisa.gov)
402 points by auditnow 6 hours ago | 188 comments
18. The Great Repo Archival: GitHub purges all inactive repos >5 years to “reduce storage load” (github.blog)
388 points by codearcheologist 9 hours ago | 320 comments
19. Mastodon instances collapse under moderation load after EU’s Automated Speech Mandate (mastodon.social)
350 points by fedifragile 7 hours ago | 144 comments
20. NYC adopts automated congestion fines after human review team eliminated (nytimes.com)
332 points by finesallthewaydown 4 hours ago | 201 comments
21. Dropbox raises base plan to $49/month, cites “AI compute costs” (dropbox.com)
301 points by storageinflation 11 hours ago | 176 comments
22. Open-source maintainers strike after 3rd corporation claims their work as proprietary training data (github.com)
290 points by maintainerburnout 6 hours ago | 120 comments
23. FEMA: 2025–2035 wildfire season officially declared a “decade-long emergency” (fema.gov)
268 points by cookedagain 12 hours ago | 112 comments
24. Quantum ransomware group QShadow hits 11 banks using break-through key recovery exploit (krebsonsecurity.com)
250 points by qubitcrime 3 hours ago | 98 comments
25. Show HN: OfflineLAN – A mesh-network toolkit for neighborhoods preparing for rolling blackouts (offlinelan.net)
231 points by diynetworker 1 hour ago | 44 comments
This is awesome, but minor quibble with the title - "hallucinates" is the wrong verb here. You specifically asked it to make up a 10-year-in-the-future HN frontpage, and that's exactly what it did. "Hallucinates" means when it randomly makes stuff up but purports it to be the truth. If some one asks me to write a story for a creative writing class, and I did, you wouldn't say I "hallucinated" the story.
At some point, no matter how something is mentioned, someone will offer criticism. I guess that in roughly 20% of all HN front page posts, at least one person comments on the terminology used. I do not see this as an argument against using accurate terminology, but rather as a reminder that it is impossible to meet everyone's expectations.
There are other terms that are similarly controversial, such as "thinking models". When you describe an LLM as "thinking", it often triggers debate because people interpret the term differently and bring their own expectations and assumptions into the discussion.
If someone asked you, you would know about the context. LLMs are predictors, no matter the context length, they never "know" what they are doing. They simply predict tokens.
I guarantee that once we do know people will start appending the word “just” to the explanation. Complex behaviors emerge from simple components. Knowing that doesn’t make the emergence any more incredible.
LLMs are “concept based” too, if you can call statistical patterns that. In a multi-modal model the embeddings for text, image and audio exist in the same high-dimensional space.
We don’t seem to have any clue if this is how our brain works, yet.
The brain has intrinsic understanding of the world engraved in our DNA. We do not simply predict tokens based on knowledge, we base our thoughts on intelligence, emotions and knowledge. LLMs neither have intelligence nor emotions.
If your brain simply predicts tokens I feel sorry for you.
Edit: really does not surprise me that AI bros downvote this. Expecting to understand human values from people that want to make themselves obsolete was a mistake.
> The brain has intrinsic understanding of the world engraved in our DNA.
This is not correct. The DNA encodes learning mechanisms shaped by evolution. But there is no "Wikipedia" about the world in the DNA. The DNA is shaped by the process of evolution, and is not "filled" by seemingly random information.
If you see HN frontpage a few years back there's almost no LLM related posts. Now there are multiple each day. An LLM didn't grasp that change, it can't creatively invent new trend or invent a larger picture of what changes in the world. So in 10 years it paints it still half posts about AI.
You can spend a few dozen prompts on getting it to do what you want but then you might as well just spend a couple hours writing that front page yourself, be more satisfied and the result would be funnier.
On-Topic: Anything that good hackers would find interesting. That includes more than hacking and startups. If you had to reduce it to a sentence, the answer might be: anything that gratifies one's intellectual curiosity.
Off-Topic: Most stories about politics, or crime, or sports, or celebrities, unless they're evidence of some interesting new phenomenon. Videos of pratfalls or disasters, or cute animal pictures. If they'd cover it on TV news, it's probably off-topic.
Edit: See also this current thread: Auto-grading decade-old Hacker News discussions with hindsight - https://news.ycombinator.com/item?id=46220540