This is categorizing with a pretty broad brush, it seems. I took a look at a random day, and the articles that are claimed to be AI are Waymo publishing a model for self driving cars and a post about scams in Apple News ads.
Sure, Waymo’s model is AI and the scams in Apple News are often AI generated, but this isn't some sort of monoculture of LLM articles.
Ironically, this article could have been improved by using an LLM to categorize things and also publishing the prompt, so we could inspect what it means to “be AI”.
The article already states that distinction. It doesn’t say that everything is written by AI, the heatmap even has a legend with different colors for “article about ai” and “written by ai”
You took a look at 5 posts made on one single day and then categorized all of the other days.. seems like you're good at categorizing with a broad brush yourself. Based on my experience as a pretty heavy HN consumer, I'm with the article author.
You took a look at all of one comment of a user and categorized them as unkind. Som'n som'n broad brush bad
Fwiw I think Waymo's self-driving AI model blog post and the piece about people using AI to mass produce ads on Apple News are fairly categorized. And the "50% of HN is or is about AI" tagline feels accurate enough.
There is so much more to tech than artificial intelligence (for that matter, there's is so much more to AI than LLMs) but even when having discussions about those other matters it feels like AI is still discussed
The Waymo one definitely is AI, the original point here was that it was AI not just LLMs. The Apple News one really is about fraud and AI as the mechanism is just a sign of the times.
The inclusion of the Apple News story is sufficient to know that probably the classification here isn’t too tight. It’s not really against the article author to point this out, either, just trying to add color for the forum here. It’s not always us vs them.
The reality is bots and bot-like behavior (honestly, much more likely the latter than the former from what I've heard IRL) are both equally effective in fucking up discourse on HN.
It absolutely has gone to shit. And a number of longtime readers and commenters have called this out as well.
The degradation started during COVID with the culture war crap (a plurality of accounts I bump into that comment and post are now created after 2021 - the fact that 0-indexing and source citations died out on HN is a good example of that shift from pre- to post-COVID HN especially becuase the only reason I started doing it was because JumpCrisscross and Raynier did it in the beforetimes), and then became a culture war after culture war. And at least in the Bay and/or amongst the younger YC crowd, most ignore HN after dealing with in-your-face toxicity or have never heard of it until they got into YC.
Hell, a deeply technical and deeply anti-AI buddy of mine got flamed on HN for supposedly using AI despite him not using any sort of AI on principal.
No one who matters care about HN anymore compared to 5 years ago, and those who are left are basically using it for the same reason a smoker continues to smoke despite feeling indifferent.
Anyhow, YCombinator owns the rights to all comments posted on HN but HN has been continually scraped by multiple repos and data stores outside of US and EU+ jurisdictions (go cry to your GDPR liaison Euro HNers who have increasingly become the plurality).
But Fortran goes skrt skrt (Gen Z/Zillenial for fast).
Honestly love Fortran as a language but really worried about how it's both the foundation of most core packages like Numpy and Scipy and is increasingly underresourced outside of physics departments.
> No one who matters care about HN anymore compared to 5 years ago, and those who are left are basically using it for the same reason a smoker continues to smoke despite feeling indifferent.
Any idea where they've moved on to? Asking for a friend...
> In the first half of the month, roughly 60% of the daily HN lineup was AI-related or AI-generated, tapering off to ~50% as we approached the end of the month. This is up from 40% in February.
This aligns with my much less scientific research. Every now and again, when I remember to do it, I will skim the top 10 stories and count how many are AI-related. I find it is very consistently in the 4-6 bracket. Probably a majority of the time, 5 out of 10 articles are AI-related, and very rarely does it drop below 4 out of 10. It also rarely goes over 6 out of 10, but I can remember it happening a couple of times.
In fact, the proportion of AI representation in the top 10 is so consistent (in my, again, unscientific experience) that I wonder if it is being moderated one way or the other.
Personally I find it boring that one topic gets so much discussion but who am I to argue. I do my bit by making sure to upvote non-AI stories that I find interesting.
A couple of years ago I considered making a simple regex-based web extension to block posts with common AI words: ChatGPT, Copilot, LLM, etc. There was one or two on every page and I didn't want to see them.
Over the next few years it flipped, most of the posts concern AI, and I scan for non-AI topics to engage with. I can't really blame HN for this, it's a massive upset in our industry and the world more broadly.
My reaction is: that's an appropriate amount. AI is probably the biggest thing to happen (in tech) in the last two decades, so I think it's fair that it's taking up a lot of the discourse.
Upvoted: I was going to write this exact reply. Everyone is a bit tired of hearing about AI, but it's advancing fast and it's the most important thing right now. As Demis says, if we get to ASI, we can solve all other issues with it.
Actually, it (LLM models) has been advancing slower for a while. This is pretty much in the mainstream. And the current approach is facing all sorts of obstacles that are not exclusively technological. The compute needs are no longer sustainable. Financially, big labs are drowning in debt and backlash against data centres has grown so much that it has become an electoral topic in at least one major country.
why exactly do we want ASI, and further, if we DID, why on Earth would we want it controlled by the likes of Sam Altman, Peter Thiel, Elon Musk, Marc Andreessen, and Larry Ellison?
Shit we’re onto thinking ASI is happening soon? I guess people hyped up AGI so much that it’s no longer thrilling to talk about despite not getting to that point either.
> if we get to ASI, we can solve all other issues with it
Jesus Christ this has to be bait. Added to the favourites to laugh about it when the lunacy and hype has died down in 5 years and the trough of disillusionment has truly settled in.
Or to laugh about it when ASI actually has arrived and we're waiting in line to be turned into biofuel for the Greater Good.
Everybody is literally pinning their hopes that God itself comes out of machine to save them from the exponentially accelerating pile of tech debt, and they're not even half joking about it. This is not engineering, this is utter madness.
I'd say it is engineering. There's a Moore's law like steady progress where the effective IQ goes up each year which will result in neither troughs nor biofuel but just smart AI.
Given human nature I also think it'll be kind of undramatic. Like when jet aviation transformed things people moan about check-in queues and maybe with ASI people will probably moan their robot butler is only 200 IQ when their neighbours is 205.
Feels more like another blockchain hype cycle to me.
Especially the part where we're always on the cusp of the entire world changing while the biggest change is just that huge sums of money changed hands, inducing psychosis in a large group of investors.
I know what you're saying, there is a lot of speculation going on in AI right now. The difference is there is hard evidence of AI capabilities accelerating now: we've gone from toy chatbots to sort-of useful coding helpers to agents solving novel math problems in 3.5 years. We could plateau any day and a lot of investors would be stuck holding the bag, or we could continue on that trajectory and they'll make crazy profits. With crypto, the bet was always "how much money will people keep sticking in this market?" which seems a lot more flawed to me.
The irony with blockchain is that it also had a bunch of real life use cases but the sheer extremity of the hype cycle managed to actually prevent them from being effectively brought to life.
There was a lot of genuinely useful tech that blockchain could have facilitated and all we were left with in the end was a mountain of scams and bitcoin.
I'm seeing something similar with AI where the genuinely useful stuff is getting washed out by the hype simply because it's a gradual improvement and not "disruptive". Meanwhile the stuff sold as disruptive like vibe coding is mostly bullshit.
What was the useful blockchain tech? It seemed to me to all be different ways of avoiding regulations or bypassing existing system or addressing issues that were already solved by using centralized systems.
it let you decentralize trust. as one example, I was hoping that as well as "sign in with google" I could sign in with a blockchain based decentralized identity service where nobody could one day decide to revoke my identity on a whim.
i remember my friends excitedly telling me how cross border container trade is still done with paperwork because nobody trusts each others' IT systems. I'm pretty sure it still is because they can't be convinced to trust blockchain systems either.
trustworthy voting is another application. ive yet to see a political party that runs an internal voting system that doesnt rely upon you trusting a central server run by a few people. this is crazy.
Funniest part of all of it is it’s all the same actors. Same hardware even. Just the crypto shills now talk about agentic capabilities & we’re approaching a singularity man! Can’t you see!!?
Yes, it's appropriate given the metric you have chosen ("big"ness), though maybe the word should then be "expected", since "appropriate" implies some kind of broader human-oriented concept of rightfulness/goodness.
You really have to expand the list to ridiculous levels before you get to something as big. We've already left humans behind in many ways (but in others we've failed to build LLMs that know how to find the gaps in implementation that humans are leaving out)
Cynics really say stuff like "but it's been 3 years and all they can claim is that it's changed text to speech, speech to text, photos to text, text to photos, programming, cybersecurity, research, mathematics, war...
"They lie about it being dangerous!" and simulataneously "They hacked a system with zero day exploits! Everyone involved should goto jail!"
Tech leaders are awful, but it is just that we've built a society that enables the most awful among us to benefit the most. It has never been the case that tech leaders are particularly amazing, they are just amazing at exploitation.
> You really have to expand the list to ridiculous levels before you get to something as big
Instead of expanding the list, it would be better to widen your vision. Email was more transformative than AI is right now. Unix, then Linux was more transformative. If you have a magic wand and remove AI, I think barely anyone would be impacted, but removing something like windows would be like erasing computers. And removing Linux would wipe most of the internet services.
Email in it's first five years barely extended beyond some research universities. And is it novel? Do you see email as somehow a difficult concept to imagine?
Unix in it's first five years reworked the way operating systems performed. I'd agree it's pretty important, but I don't see how it fundamentally changed the way we perform work in multiple disciplines (in its first 5 years). Same for Linux, but even less so.
We're in very early times, and generative AI is touching every corner of the workpace and education, for better and worse. Powerful technologies can have negative effects. Email has been negatively transformative in ways as well.
Part of the reason you can erase AI so effectively with so "little" effect is that it is new.
There's only a few months left for the core technologies of LLM, the transformer, to be a decade old (counting since the "Attention is all you need"). In a decade, we have gone from pure theory of computation (Turing paper in 1936) to the first general purpose computer (ENIAC in 1945). In 1971, the first email was sent over the network (ARPANET), by the early 1980s, we were looking at nationwide deployment of email systems.
> We're in very early times, and generative AI is touching every corner of the workpace and education, for better and worse.
Forcefully. By a mixture of free offerings and bundling AI in every tools with no disable button. It's still not an essential tech like email or Linux is (not even essential at the level of GitHub).
This is an entirely weird and bizarre conversation. The world today has network effects - specifically because of well, networking. It feels almost ridiculous comparing the impact of things like ENIAC to github because they aren't even in the same class. GIT and email (FYI, there is a predecessor to email in 1965 called MAIL) are specific implementations of systems that would have happened no matter what. GIT is awesome, email is kinda ok -- but that's beside the point.
> Forcefully. By a mixture of free offerings and bundling AI in every tools with no disable button. It's still not an essential tech like email or Linux is (not even essential at the level of GitHub).
We could survive without computers and the internet. We did it. I was there. We could go back. It's not essential. You've lived maybe 3 years with LLMs. No kidding they could disappear. As an old person, I have no idea why you think any other technology is any different. LLMs are new, that's the big difference as I said earlier.
> Forcefully.
So.... I hate mobile apps. I hate that TVs connect to the internet. I hate that we have cameras on every corner. I hate that we have cameras that watch us while driving.
Now, tell me about how this is different than all that.
> As an old person, I can tell you something: We could survive without computers and the internet. We did it.
As late as 2009, I was living without computers and internet. The most tech thingies in everyone home were the radio, the TV, the VCR (not everyone and half of them wasn't working properly) and a dumbphone.
I'm not saying that internet or any of the others is essential (in absolute). What I'm saying is that the supposedly "transformative" aspect of AI needs free offerings, dark patterns, and a mountain of hyping to just make it worthwhile to discuss in tech circles (removing the grifting on Linkedin). For everyone else, the most impact is the enormous amount of slop.
I paid for my computer. I'm paying for my internet connection and my electricity. I'm paying for my email service. Drop the free offering for any current AI service and see how much it is that transformative.
> As late as 2009, I was living without computers and internet. The most tech thingies in everyone home were the radio, the TV, the VCR (not everyone and half of them wasn't working properly) and a dumbphone.
The VCR? Hrm. What part of the world are you from? No computer and no internet? Hrm.
> Drop the free offering for any current AI service and see how much it is that transformative.
It is currently being paid for by millions of companies in Microsoft products, and I can tell you that many companies are happy with it.
Microsoft honestly has done a tremendous job integrating it with some of their products, less so in others. (Excel is a disaster, Teams has been pretty good)
- A successful surgery may require both a surgeon and a nurse, but that doesn't mean the nurse has nearly as much impact on the successful surgery as the surgeon.
- The lighting of the match is mainly caused by the striking of the match, not by the presence of oxygen, even though both are necessary for lighting the match.
So a cause being necessary for an effect doesn't imply that its impact on the effect is very large.
The ambiguity partly comes from denial. LLMs in the end, to be brutally honest, replace us. That is the direction it's headed and some people can't face the cold hard truth.
AI is not limited to the rigid construction plan of an organism. A model also doesn't have to fit inside a small skull or be limited to the daily energy intake of a human.
> AI is not limited to the rigid construction plan of an organism.
This is some AI psychosis shit.
Explain how you think extrapolating off of 4-5 years of growth(I’m being generous here) means the development is asymptomatic instead of reaching a plateau like every other aspect of physics.
Are you one of those people who believes Anthropics claim that they will have the revenue of the entire United States yearly for their IPO?
> AI is not limited to the rigid construction plan of an organism. A model also doesn't have to fit inside a small skull or be limited to the daily energy intake of a human.
We haven't seen any significant progress at all, much less extreme progress. You still can't use these stupid slop buckets for actual useful work, despite the massive amounts of resources being wasted on them.
There's also a part of me that misses the earlier days of Hacker News when the subject matter was much more eclectic and seemingly every day I'd have a chance to learn about something completely different.
And that's maybe what I miss. Nowadays Hacker News feels increasingly like another social media platform: mostly a lot of opinions, war stories, and self promotion. That feeling of serendipity that so characterized earlier ages of the Web is fading here, too.
It seems a bit self-reinforcing. AI makes almost zero (positive) impact on my day to day life and isn't actually something I interact with directly and the same holds true for most people I talk to.
There's a few of my friends in tech for whom AI has obviously massively changed how they work, but again not in a clear positive direction.
To me there seems to be a push to heavily prompte LLMs, even thought the benefits aren't all that clear yet. Some people are seeing massive improvements in how they work and it's understandable that they, like with blockchains or Kubernetes previously want to share their optimism. I just question if maybe these people are being to louder and have to much influence. There's a large silent majority for whom all the AI stuff is almost completely pointless.
It does need to take up more space on HN, compared to your normal media, because the technology needs to find it's way into applications and appliances. The hype will die down, especially once we start to see the actual costs, e.g. ChatGPT doesn't actually provide enough value for the average user to warrant a subscription, hence the ads. For now, it seems that AI is heavily pushed by industry and a group of fans, the amount of news and articles just doesn't line up with the actual daily use.
I don't think that AI has to influence every single person's daily life to make a huge impact. I think for probably 90% of people it's pretty useless out side of being a better search engine. But that doesn't mean it won't have a huge impact on science and technology at large. There are a lot of snake oil salesmen trying to stick it into everyone's face, but I don't think that should be conflated with the real progress that's being made.
I would say 90% of people do not yet realize how useful it is. It is not useless to them, they just don’t understand yet what it can do.
And there were a few years there where a lot of BS AI hype turned them off to it. So a lot of people rightly dismissed it and don’t realize the inflection point it hit in the last few years.
ChatGPT work is basically a personal assistant for $20/month
> ChatGPT work is basically a personal assistant for $20/month
A personal assistant that will happily tell you that you’re the world’s largest supergenius who invented perpetual motion. I wouldn’t want that if they paid me $20/month to use it.
i just gotta say that i find it hard to believe that this comment is true for you but you hang out on hackernews.
almost everyone here is a software engineer or works in tech, they all use ai all day to do their entire jobs.
> even thought the benefits aren't all that clear yet
it isnt 2023 anymore. wtf are you talking about.
>he hype will die down, especially once we start to see the actual costs, e.g. ChatGPT doesn't actually provide enough value for the average user to warrant a subscription,
again, it isnt 2023 anymore. none of this is true.
I assure you they do not, many truly do not care and don't want to. That's just what the hype and the article and the stories tell you. The HN bubble tells you that's true, but 90% of my co-workers don't even want to touch Claude, despite being given a license.
And again, if e.g. ChatGPT is providing so much value, why aren't people buying?
They have "more than 50 million consumer subscribers" and "More than 9 million paying business users" [0]. That's for example more subscribers than for Adobe Creative Cloud.
So, about 0.6% of the world’s population is willing to pay for something that’s supposed to be hugely useful for literally everyone (as opposed to Adobe products, which are only for digital artists)?
Not everyone in the world has $20/month of disposable income, nor does everyone have the disposable time to learn how to make it hugely useful to them.
Also, these are just the numbers for OpenAI. A lot of others are ok with e.g. "paying" for it by going to Google's AI Mode, and getting some ads.
So ChatGPT cannot deliver you $20/month worth of productivity to offset the cost? That's at least my point, for developers, businesses and a ton of use cases, access to a good AI is worth every penny, because it easily pays for itself. For the average consumer, that's just not true, apparently.
HN is in an extreme hype bubble around LLMs, because of AI coding. Normal people outside of the tech bubble range from 1. not caring at all about AI to 2. disliking or opposing it.
Meanwhile, here on HN, you have people who earnestly say things like "AI is the subtrate the future runs on"[1] (actual quote).
Just saying "it isn't 2023 anymore" isn't an argument. Yeah it's not 2023. And LLMs still can't actually do anything useful outside of sift logs and write small scripts. They still hallucinate things regularly so that you have to check everything they do, which negates the time savings they supposedly provide. And that's not even getting into the externalities! LLMs are a very crappy technology.
I guess Hex-Rays and Vector35 must be hallucinating their effects for, say, reverse engineering. OFC rev isn't be all and end all of computing but the effect LLMs had on security research and adjacent spaces (say, CTF) is interesting
I am so fucking sick of this. No, you dumbshits -- ALL of us DO NOT use LLM's to write code. Just fucking STOP assuming every developer needs the same crutch that the rest of you can't seem to live without.
I use it a couple times a week, not even daily. It's useful about half the time, and a complete waste of time the other half. It's a useful tool to reach for sometimes and I'll keep using it for the things it's good at, but it hasn't really changed how I work. Shrug.
LLMs have helped us keep HN up and still (relatively) fast amid a huge onslaught of new traffic over the last year or so. We've been able to solve problems and find opportunities for optimization that we'd never have gotten to without the help of Claude or ChatGPT.
Although this is one of those threads where people love to report how much they dislike HN, how it's a poor shadow of its former self, and how they barely look at it anymore—a popular local sport for 15 years or more—but to the extent anyone still reads the site, there must be some positive impact in having it running and rendering pages, even if you don't like what's on them!
Hopefully HN is just getting more popular, but isn't a large part of the "huge onslaught of new traffic" driven by AI companies scraping the site for content or writing comments for various reasons?
Fixing the problems created by LLMs using LLMs seems like a bit like hiring a bully for security to prevent that same bully from beating you up.
But yes, there's plenty of positive things that are now possible. See as a combined net-positive vs net-negative, I'm still leaning towards the net-negative.
I fear this is not the only reason even if it is true. The more negative reason is a loss of skill/talent. We used to be able to know what to do in these situations, and how to fix things when they went wrong. Companies like Y Combinator used to be able to hire talented systems engineers who knew how to minimize the risk of things going wrong, and fix them when they did.
Now with LLMs it is just vibe-coded until it is fixed. Find the right prompt and hope the statistics is in your favor. If this is the case, there is nothing positive about it.
I take your point! But the positives are still positive. I've been amazed by how much LLMs have helped us do.
None of it was undoable otherwise, but we (where by "we" I mean "I") would never have gotten around to it otherwise. Lowering the bar from months to days, or weeks to hours, is significant.
I think the scrapers have gotten pretty good and can imitate real traffic. And it isn’t just scraping for model training. A large amount would be LLM tool calls searching the web (and HN) for info about some prompt.
>because the technology needs to find it's way into applications and appliances.
I hear what you're saying, but speaking for myself: can we *not*? I really have absolutely zero interest in having ChatGPT on my microwave or refrigerator. In fact, I'd rather have my appliances *not* be connected to the Internet, all things being equal.
>I'd rather have my appliances not be connected to the Internet,
It's getting harder and harder to find things that don't connect to the internet as time goes on. The moment a manufacture figures out how to sell your appliance data for more than it costs to add it, seemingly every model they produce gets it added with the only choice of going to some other manufacture if you don't want it... and even then finding one not internet connected gets nearly impossible.
In general most people don't care enough for people like us that don't want connected devices to move the dial.
Problem is other cool hacky stuff gets drawned out. Nobody votes on those because I imagine it's no longer attractive to the type of visitor that HN attracts now. I reckon there's many more non-tech, "make me a millionaire" gold rush crowd here, as cynical as it sounds. I base that on pure logic, since it's kind of inevitable with AI having delivered a massive democratization of tech knowledge, but also on the amount of snarkier, shorter, less technical, less logical, one short sentence type of comments.
I know it's been said countless of times before, but I think this community has truly changed within last year.
The worst IMO is the bot accounts. And the worst of that IMO are the people that behave as if they were bot accounts.
The self-promoting thought-leading experts (usually on AI) with a skin as thick as a single cell membrane.
At the moment, half of this site on a given day is terrible ads with a swarm of the same people defending each other's attempts at either becoming or staying relevant - utilizing the latest gold rush topic to do so.
___
I'm not quite sure what could be done about this. The moderation does do a good job, so it's not that.
I also see that not everyone feels repulsed by the self-promotion. So maybe there is no problem and/or the problem is me. But I kinda doubt that too.
Maybe the time of voting systems for ... uh.. "social media" is somewhat over-ish? Though if I was truly signed up on that thought, I would not be commenting here but on I guess lobsters?
Most of the content is about using some data center based ML provider. My view is how so many ignore the environmental harm these constructs are doing; increasing ambient temperatures, inducing heavy metals into the water table, decreasing the access to water for human consumption in numerous drought areas, hostel ambient noise ... The wealth who are the majority finical backers do not care about these issues.
ML is a tool and hardly focused on it being able to operate as a local based solution. Frameworks are tie to expensive hardware. This creates a hyper pay-to-play scenario. While hand written coding can be done on a computer from 20 years ago. Local based solutions are the only means to decrease the wealth gap. Data center solutions increase the wealth gap.
I don't find value in ML summaries. I want to consume the answer and the ideas around them that lead to the answer. This creates more long-term value.
_Lobsters_ has now becoming my go-to with their posts being focused on ideas and concepts versus advertisements for large corporation ML based solutions.
That study was so bad that I can’t believe it’s becoming an actual concern that people use against data centers. They were barely able to register a temperature increase a couple hundred meters directly downwind.
> _Lobsters_ has now becoming my go-to with their posts being focused on ideas and concepts versus advertisements for large corporation ML based solutions.
Lobsters has a funny obsession with being anti-AI, but the most commented posts on the front page are usually AI related.
> Lobsters has a funny obsession with being anti-AI, but the most commented posts on the front page are usually AI related.
The top five most filtered-out tags¹, excluding meta², are (in order): vibecoding, dotnet, php, emacs, ai. Two of the five are AI-related. The top spot (vibecoding) has 60% more people filtering it than the second spot.
So less than 4% of the userbase filters out vibecoding stories? This looks like a big red herring.
Every time I open up the lobste.rs home page, there's an 80% chance the most commented on stories are related to AI in some way. Other stories are lucky to get 10 comments.
> Lobsters has a funny obsession with being anti-AI, but the most commented posts on the front page are usually AI related.
Reminds me of gamer-gate: Back then, NerdCubed graphed the top posts from both pro- and anti-gamer-gate subreddits, and most were meta posts about themselves and each other rather than their alleged topics. Gosh, that was such a dumb period.
It is a tragedy that a handful of irresponsible outlier datacenter operators have soured public opinion so thoroughly in the past 2 years. For comparison, Loudon County, VA is among the wealthiest in the country and is home to the largest aggregation of datacenters in the world. Yet these datacenters don't have the externalities that are mentioned here.
Note the intersection between eyewitness accounts of noise, water, and light pollution from datacenters and those datacenter owners' noteworthy political activity. To be specific, the operators of these datacenters are notorious bad actors in many different public spheres.
Edit: the truth is obviously very painful and unpopular in this thread.
I really wish there was a way to filter out discussions about AI on HN. It feels like lately they take up a good 1/3 of posts on the front page. I don't want to hear about some new AI startup.
I made a similar site https://unslop.news/ but rather than keywords it actually feeds the title and article to an AI which classifies it as "about AI" or not to filter out the AI discussion. Also it's open source: https://github.com/leiDnedyA/hn-without-ai . Hope it helps!
Well your point on it being wrong to interact with a whole industry sounds more like a policy and political party line. This is just nerds discussing bleeding edge things.
a first look through the comments on some first page threads and they seem far more thoughtful and considered. it generally seems like a nicer commenting environment that might actually do me some good moving forward.
i feel the same way. there's nothing interesting or curious about someone writing prompts to outsource a bunch of effort/thinking/work off to some most likely continuation sequence predictor / set of most likely continuation sequence predictors.
i miss seeing clever and thrifty solutions to weird problems. hell, i miss people just doing plain weird shit and posting that.
Although there is some irrational hysteria about the whole "chatgpt is drinking all our water" thing, to me the biggest issue (other than the mass piracy) is local energy bills going up without contributing to the local economy. Seems like it would do more harm than good for all but the owners of the datacenters.
AI has made me aware how opening a tab to HN is a compulsion I no longer truly enjoy. Especially after having made the conscious choice of leaving software engineering as a career, my time here is a mildly negative experience hoping for the random chance of finding a deeply technical post that is not about LLMs, or whose discussion does not devolve into discussion with Joe Rando on how much more productive he is lately.
Sure, it's a moral failing of mine, but by god I wish we had a little more diversity in topics.
Consider trying https://unslop.news/ , I tried to solve this exact problem by classifying post title and webpage contents as AI or not and filtering out the AI stuff. Also, tangentially, if you're looking for a really impressive technical writeup I recently enjoyed this one :) https://philo.gay/linecam/
I appreciate the way you sum up the ML vs traditional software appeal and I think it's helped me articulate why ML is so unappealing to me. The power of software has very little to do with the power of the machine it's run on (with some obvious exceptions like games and physics sims) which made it very exciting to me when I was learning to code. The same cannot be said about deep learning where more data and more compute means more power.
I miss the old newsgroups. I wish we could land again in that weird, colorful oceans of geeky individuals with their quirks and interesting nerdy takes.
15 years ago, HN was about bootstrapping and early YC stars.
10 years ago, HN was in full VC-enterprise sauce.
5 years ago, it was crypto and NFTs, and how I switch jobs every quarter.
Now it is a blurry view into the future through the black mirror of AI.
How could we jump back into the cozy past again?
I do not have a clear answer, but what seems to work for you are small groups of friends and friends of friends hanging out on WhatsApp or Discord, discussing geeky news together in the context of their backgrounds. But these are very curated and very closed groups, still operating on the social media platforms of behemoths.
Would now be the right time to restart self-hosted message boards or Mattermost/Campfire instances?
Run a self hosted forum or instance of whatever you like, and just only let people sign up if they are recommended by someone or can argue their way in, basically. xenforo and some add ons can do that, and I'm sure there's a few self hosted messaging apps like you mention but I'm not as familiar
IRC is small and niche enough to be a good place for more ephemeral discussion if you have a bouncer to stay connected for you. Chat on it tends to move slowly, and the userbase trends more towards actual nerds.
I think it is a problem of how we changed: shorter attention spans, and a much larger volume of information to process.
Short forms win, long forms vanish, which is a tragedy in itself as good writing requires good understanding, and understanding is impossible without attention.
I'm in my 40s. I remember lcamtuf from pl.comp.security when I was just getting started playing with FreeBSD and other fun stuff.
I joined HN 13 years ago (2 years after you), and back then I saw it as a bunch of libertarians and venture capitalists and some open source (and/or functional programming) enthusiast with a few interesting and nerdy projects. I was mostly here for the news about new and useful technology.
I think the visual weight of the first group has increased over time as their projects of making a bunch of money while offering nothing of value has been validated more and more by our capitalist economy.
I joined in 2008, that was even after lurking for a while, and it’s almost unrecognizable now in comparison has been for many years. Was mostly entrepreneurship, tech was present but was a means to an end. It wasn’t tech for the sake of tech, which is what it feels like in the recent past. The discussion on entrepreneurship topics seems to have an employee mindset more so than an owner mindset. I find it similar to how renters vilify landlords but being a landlord is not massively glamorous and can be quite risky, renters are generally out of touch with the costs involved and the profit of a landlord is all I’m saying. There’s definitely some shitty landlords tarnishing the reputation of them all. This is what has happened in tech too, the startups we loved turned evil. My gut reaction of any new tech product is no longer curiosity, it’s mistrust.
Anyhow - I find tech stuff interesting to a degree, but fundamentally I’m an entrepreneur that sees it as a tool. A means to an end. And I would love to find a place that still has that vibe. But anywhere entrepreneur related is always a cesspool of grift so it’s hard to find.
As for the employee-vs-employer. I am for sure part of the problem there. When I joined I immediately saw the potential of this being a forum where tech workers can form an international solidarity and complain about our bosses and started using it as such. There is no way you can have that many workers in an industry together and we don‘t complain about our bosses, that is just not going to happen. Workers complaining about our bosses is a universal behavior.
I am part of the problem in the first. It's because early HNer's were early Bitcoin users and adopters and I feel we (I) had a hope that it could take hold as an actual currency (I no longer feel that). But, we were so early I can say my first Bitcoin purchase was for <$10 and most of the Bitcoin I own today has a cost basis <$30. For us, it's climb was fast, wild, and wealth inducing. I thought it was insane when it quickly went to $100 and I think one of my own HN posts that has been maybe my most upvoted ever was about it hitting $300. I often wonder how many of the bitcoin wealth you hear and see about IRL, were just HN users that heard about it's early traction and had enough curiosity to buy a few coins (a cup of coffee then, is worth a house now). It was a PITA to buy coins though, I luckily used Mt Gox and moved them out prior to them getting "hacked".
I put my Bitcoin on ice, refused to sell it, my initial investment thesis was "what happens to the price when all Bitcoin is mined", given the supply is fixed, the price should surge, was my thought. I think they have or will find a way to alter the algo so that even more can be mined. But once I made that decision, it became difficult to see all the Bitcoin related news. I wanted to just not think about it. Daily reminder that I could sell today at a crazy high gain is tempting. I don't even know if that thesis makes sense given the projection of when that last coin will mine. Those projections didn't exist when I put this into place. So I hodl for no reason probably.
In terms of the posting and ethos, it was really all the non-Bitcoin crypto crap that came afterwards that I got super tired of. But I am older and know it was just the current hype cycle, just like AI today, something else will come tomorrow. Although AI is kind of a big one, it's both massively hyped and useful too. It's also moving faster than I can keep up with, so I just learn to dabble where I have interest. I'm not trying to be a creator of AI or start an AI startup, it's just tinkering and when it comes to tech as a means to an end, it allows me to move faster. But it's also given everyone my superpower. Kind of a net zero with extra steps. It feels like my beloved woodworking hobby has been changed to IKEA furniture assembly which is not fun at all.
I never cared much for newsgroups, they were mostly nice for the historical stuff. I do miss IRC though, as toxic as it could be. I could run it in a little terminal on any system without download electron apps and being spammed with ads (discord).
The issue is the tech and the motives to wield it. Cosy decentralized message boards exist, thankfully outside the realm of the greater public.
The moment that the greater public sees these high trust networks and tries to join them, all the bad actors and their armies of bots will follow and I mourn for the recreation time of any admin who has to suddenly deal with that.
Private walled gardens seem to be the only way forward for now. If you love something, do not share it here
I'm nostalgic for the new JS framework every day era. Say what you will about that era but people were writing the code by hand, so looking back it all feels "hand-crafted" and quaint. You had to copy paste that express app from a medium.com article instead of prompting an LLM to scaffold a whole project in one second.
At least there was some genuine effort involved there. What I find most disheartening about the current wave of AI-generated projects is that there's often very little personal involvement - someone had a general idea and asked Claude (or whatever) to implement it for them, and the results are much more representative of the AI agent's tendencies than the author's intent.
Democracies require conserved stake to function. How individuals initially get that stake is a degree of freedom, but if the stake can be printed freely, the system is unstable.
For working implementations of democracy, look at: invite trees, shareholder corporations, proof-of-stake distributed ledgers, etc.
Matrix works fine for the friend to friend, small community things. Discourse forums too. Use Signal for family and friends, perhaps work. Fediverse and Atmosphere (Bluesky) for your microblog needs, and RSS to keep track of blogs.
Having used Matrix for almost a decade, it's hot garbage.
Sync is barely working, "unable to decrypt" still exists, the encryption is flawed [0] and the whole thing is effectively just Element and everyone else follows their lead.
About the post you linked, see [1] and [2], which makes me question both whether that is an actual vulnerability (Signal, the messenger recommended by the author, also didn't have that check, and it's addition is absent from release notes and CVEs) and whether the post was made in good faith (the check was added to libsignal on the same day that the author disclosed the vulnerability to Matrix, which can be a coincidence, but doesn't seem likely given that the code has been there for years).
Can't speak about the other issues, I haven't had those but I barely use it. Based on comments, it's clear they exist or have existed for quite some time.
> About the post you linked, see [1] and [2], which makes me question both whether that is an actual vulnerability (Signal, the messenger recommended by the author, also didn't have that check, and it's addition is absent from release notes and CVEs)
I've edited the above linked post several times to make this clearer, but people still keep linking Matrix's flawed response as if it actually addressed the issue. It does not.
The issue is not as simple as "did the check exist? [y/n]". There are actually two issues:
1. Did the check exist? [y/n]
2. Did the group key agreement protocol fall to shit if the check was omitted? [y/n]
Matrix was (n, y) due to how Megolm manages keys in group contexts. Signal was (n, n) due to their protocol design.
Matrix tried to defend point 1 when the blog post was about both 1 and 2.
(Post-Quantum MLS, for comparison, would also have been impervious to this issue due to how KEMs work. Another point in favor of MLS adoption.)
> and whether the post was made in good faith (the check was added to libsignal on the same day that the author disclosed the vulnerability to Matrix, which can be a coincidence, but doesn't seem likely given that the code has been there for years).
Even if you assume the worst possible interpretation of this coincidence (which, sure, you're free to if you want-- you would be wrong if you did, of course): What does it even matter?
Like, what is the notion of "good faith" that's even necessary for someone to do independent security research and publish criticism of the cryptography used by a software project that's being propped up by EU sovereignty movements?
In my opinion, this is a stupid and vacuous framing.
Signal omitting the check didn't harm Signal's confidentiality. Matrix omitting the check demonstrably did.
---
Also, if you're going to decry my blog post for not enumerating unrelated security bugs (n.b., the blog post was about Matrix, not Signal, so why would I talk about Signal's code in that post???), be aware that I reported non-cryptographic bugs to Element (Matrix client) and Conversations (XMPP client) around the same time. Only XMPP actually credited me with any fixes.
You don't even need explicit barriers, the masses are staying on Reddit/Twitter regardless. Only the discerning are seeking out better and newer options and those options are staying the right size rather than absorbing the eternal September.
> How could we jump back into the cozy past again?
"You can't go back home to your family, back home to your childhood ... back home to a young man's dreams of glory and of fame ... back home to places in the country, back home to the old forms and systems of things which once seemed everlasting, but which are changing all the time - back home to the escapes of Time and Memory."
“No man ever steps in the same river twice, for it is not the same river and he is not the same man.”
I'm seeing this nostalgic sentiment a lot on forums these days, but it's not just about bringing back ye old usenet groups: you also need to bring back the way people thought back then. It's hard.
The sentiment I'm seeing is that, to avoid the rose-tinted bias of nostalgia, we default to believing there is nothing the past has to teach us, and everything is just better today.
If it were democratization, we'd all have access to the data that underlies this dataset. In reality, this will destroy the origins of the data, and it will become hidden for all of humanity.
It is a tragedy of the commons of the highest order.
They will talk about how they are trying to protect humanity all while putting it behind a paywall.
> How could we jump back into the cozy past again?
Delete 9/10th of the HN population. I could accept even AI discussion if there wasn't so much of it, and so many posts in general; /new is like trying to win the lottery.
It doesn't matter how interesting it is, you just have to be lucky. The corollary is that the front page isn't the most interesting stuff posted today, it's a list of who's got lucky posting the average topic for the average reader.
Is that really the problem, though, or has the audience simply changed? It seems to me that the thoughtful deep dives of years past are still being produced, but they're buried under a much larger volume of lighter content. More importantly, many of the people working at the forefront of technology and product development appear to have moved to private communities such as Discord, Slack, and similar groups, perhaps in part because of NDAs.
All HN users can be part of the solution. Help mods by emailing them at hn@ycombinator.com. I recently pointed out a suspicious pattern, and the mods were able to use that to track and ban a collection of accounts.
Some users on HN are "allowed" to do breathless AI self-promotion on HN and other readers support them by posting it anyway. Other HNers on the other hand, got banned immediately for doing the same thing.
That is why many users here realized that there are protected accounts that can do that and everyone else cannot. Even in rare circumstances, the moderators sometimes unbeknownst to them, may also be indirectly allowing it.
It is far more complicated and one of the reasons why users here are fed up of the constant AI deluge on this orange site and nothing is changing other than there is more of it.
I’ve had some luck messaging the moderators and calling them out for it. They may or may not do it on purpose, I have no idea. But it has worked to get certain golden children at least flagged when they’re doing things that would get others straight up banned.
Reading about tech has become gradually less and less fun for me over the past 16 years.
Does anyone remember the awe that was felt the first time we encountered an iPhone. Or an iPad? I remember those days I would read tech news and sites like Ars Technica and comments were sometimes snarky and indignant but generally people were enthusiastic about reading tech because it offered glimpses into more technically advanced futures.
I wasn't browsing HN back then but I imaging the same thing was true of HN in the late 2000's and early 2010's.
Now tech sites like Ars Technica consistently have comment sections filled with immense negativity cynicism indigation and just plain old boiling rage. I avoid those comment sections now because reading such things negatively affects my mood and certainly isn't fun.
And over the past few years HN has gradually become more negative and nasty.
All this sucks. This isn't why I became interested in tech. This isn't why I used to spend time reading about technology and science.
It kinda started turning as people realized that a huge amount of the biggest people pushing tech are rabidly anti-human, incredibly racist and usually spouting nazi quotes every chance they get, and are out to exploit as many people as possible just because they can.
The negativity is rather well-earned at this point. I do still enjoy tech, but the ecosystem is no longer as infused (by ratio) with nerds doing cool things.
All this sucks. This isn't why I became interested in tech. This isn't why I used to spend time reading about technology and science.
You're 16 years older now. You will have changed. HN has too, but looking back at how great it all was without also realising that your point of view is different means you're missing a big part of why it feels worse now. I've been here a long time too so I'm not going to claim I don't see the same thing, but I imagine someone who signed up 3 months ago won't have the same opinion as us old timers.
> but I imagine someone who signed up 3 months ago won't have the same opinion as us old timers.
that's for sure but if I got to know HN 3 months ago I would not be an active user. The content and discussions here are still interesting but the demographics that enjoy it changed over time and I (and OP) might not be the kind of people that would get interested anymore.
There is no "hacker" in HN anymore, unless we count the hustlers trying to cheat their way to success. Talking about that, there are so much psyops going on in so many threads that we, the old hackers, are the ones getting hacked.
I still maintain the notion that tech was most exciting/peaked around 2010. Macs were awesome (and the Intel transition was complete at full strength), things were fast, OSes weren't bloated, and we didn't have flavor-of-the-month fads like cloud, crypto, NFTs, or AI.
I can't recall who said it, but "The biggest problem with golden ages is that you never realize you're in one until it's over".
There were several golden ages of tech for most people who went through them:
- the pre-internet run-up of personal computing (~1975-1995)
- the early internet (~ 1990-1998)
- the big post-dot com crash tech boom (~ 2003-2020, but more of a golden age for employees)
However, almost all golden ages (including non-tech ones) end for the same reason (the ranges above I'm excluding the last phase). Early adopters (tech enthusiasts), early implementations that focus on them, then mass market appeal that "dumbs it down", then the financialization as accountants and grifters start showing up. Same thing happened to Hollywood, the music industry, etc.
We're now squarely in the last phase of the post-dotcom boom.
(I'll also note that the timeframes are arguable - I'm not going to get into quibbles over minutiae)
>Does anyone remember the awe that was felt the first time we encountered an iPhone. Or an iPad?
Yes, and when I first saw LLMs and image generators I experienced a similar feeling of awe. We haven't had anything like this in decades.
It is unfortunate that many people are negative and nasty about it. Local politicians are getting shot at over datacenter projects. We have truly lost the plot.
I used to think that people had learned their lesson that automation was a force for good, and we had gotten past the Luddite days of smashing looms. But it seems that is not the case, people are just as short-sighted as ever.
Ignorance. Ignorance of the facts on the ground and the politics of the moment.
Do you honestly believe that the ledger for automation has zero items in the "liabilities" column? Do you think that 100% of people whose job is displaced find another job in a reasonable amount of time and on the same order or pay / status?
Do you believe that neighbors of recently built data centers have nothing to complain about?
The truth is that "data center politics" is happening in the context of post-NAFTA offshoring of lots of jobs/industry, post-manufacturing migration to the South, post-hollowing of labor unions, and in the context of Trump's War on Immigration (where AI companies have likened their product to "infinite AI immigrants").
Even the White House's Ratepayer Protection Pledge (a laudable goal) event was undermined by POTUS's speech at the event. Instead of even pretending to listen to the people who are hesitant about first and second order effects of data center buildouts, he encouraged developers to push the projects through ASAP and ignore all of the naysayers. It's clear this entire project is to maximize commercial advancement, not benefit people or give voice to voters/constituents.
And in a moment where we are told we have to do this ASAP because "we are in a race with China" but nobody can explain the race. It's not for AGI because China doesn't expect it to yield AGI. Perhaps Zuckerberg does, but he was wrong about crypto, then Metaverse, then open weights models training (before he returned to open weights). The US seems to be using LLMs wrong as a glorified DEI detector (DOGE using LLMs to enforce their political homodoxy) and to identify schools in Iran to bomb + double tap.
China has reasonable regulations on LLMs to protect their society whereas our "regulations" seem to be the Amazon CEO texting David Sachs on a Friday evening to warn about cybersecurity in a misunderstanding between what Mythos (prerelease for select privileged users) and Fable (general release with lots of moderation controls) and the nature of the cybersecurity vulnerabilities found (they were able to be replicated with other non-Claude models).
We are being outwitted by ourselves. The problem isn't the people anxious about data centers or AI. It's the clueless leaders who don't bother to anticipate the obvious outcomes of their decisions and execute a plan to mitigate it. And you desire "feeling awe" so much that you don't want to recognize that politics means not everybody is going to succumb to the same feeling of awe, and that's okay.
>Do you think that 100% of people whose job is displaced find another job in a reasonable amount of time and on the same order or pay / status?
Do you not realize that the only reason we are not impoverished dirt farmers right now is because of automation?
Automation is what creates wealth. None of these 'high pay/status' jobs would exist if it were not for automating the previous generation of jobs. Every time we automate, we invent new higher-paying jobs and increase average wages across the economy.
This has played out time and time again since the industrial revolution. Luddites like yourself try to block automation with 'what if there no jobs?? what if it only makes the rich richer and we all starve??' Then automation happens anyway, we create new jobs that pay more because the pie got bigger, and we all end up better off.
I'm curious if anybody has tried to figure out roughly the % of bot comments? I keep seeing the "ai-talking cliches", but wonder if I'm crazy because people (or... other bots???) take it at face value.
I've been noticing in recent months that even when the article is in no way related to AI, there's often comments trying to steer it into AI, often in a way that makes no sense at all. It has a way of becoming its own parody.
I'm deeply disappointed that many of the stories on HN are directly or indirectly about A.I. and LLMs. To me this is a complete turn-off. OTOH this seems to be the case everywhere in the IT landscape these days.
Many people interested in technical stories have mostly stopped reading computing related news. I know people that used to host talks about .NET software development quit because they hated all .NET developments pushed by Microsoft were related to Azure or A.I. These are developers that want to build software, not scream to/plead with an LLM to do it for them. It simply takes the fun out of being a software developer. It wouldn't surprise me in the least that in a couple of years many people will leave the sector because it's no fun anymore. Management is forcing you to use your (supposed) future replacement.
I think grief is appropriate for the part of the software engineering craft that has been lost, because it was genuinely fun and rewarding at times, and I don't think it'll ever come back. Sure, some of us will hack on things manually for fun here and there, but even that will likely become ever more niche, like doing your laundry with a washboard. And professionally, no one is going to be writing code without LLMs before long. That world is gone, and it's OK to mourn it.
Yeah, I definitely don’t think all software engineers will be sending their codebase to OpenAI or Anthropic, but I think basically all of us will be using LLMs. I keep hearing about how open weight models are only six months behind, so presumably in a few years, it’ll be feasible for corporations to self-host very capable models. Still, I suspect most will just use the dominant labs, or something like Bedrock where you have stronger ZDR guarantees.
A few, yes. Just like there are woodworkers who only use hand tools out of principle. But most engineers will use LLMs on their side projects, and virtually all will for professional work.
The post tracks two rather different things. AI is almost certainly the hottest topic in tech right now, just like mobile was hot after the iPhone launched or cloud was a decade ago. AI-generated text is a separate complaint, and it's the only one I think is controversial. Slop, no thanks. A human using AI tools to write better is fine by me, even if a tell or two slips through.
Counterpoint: what we call AI is really just more sophisticated statistical algorithms, and so all of the stories about using AI, making AI more efficient, and all of the vibe coded apps are really just the world coming to terms with a new computing paradigm. It’s what you’d expect on a computing-heavy forum.
> I can't tell the difference so what does it matter?
It matters because you’re not the only human on HN and others have difference preferences. “I don’t care about this particular issue” is not the same thing as “this particular issue isn’t important”.
> you shouldn't be surprised that everyone on HN is all over it.
Not the argument either. No one is “surprised” this happens, people (obviously not all, but a significant number) are annoyed by it.
If people can't tell the difference but still get the shits because of some notion of being cheated or putting humanity on the pedestal, they're just being tribalistic. It's a common human rationalisation that helps entrench social hierarchies but in this case, and it's probably not the right word but Speciesism.
It's about as strong an argument as going to the paint shop, loving a colour, painting your house, then your neighbour informing you that a person of colour or an algorithm created it, then you hating it.
We were something different before we were primates, and we'll be something different after. Evolution has always been in a flux.
We're all just entities processing information and in my opinion, humanity should collectively get off the high horse.
Serious question, could you ever have a meaningful relationship with a teddy bear and/or an algorithm?
Not “people”, you. Some can tell the difference, others can’t. Just because you cannot tell and don’t mind that the waiter pissed in your soup, it doesn’t mean other people can’t tell and shouldn’t be bothered by it.
> It's about as strong an argument as going to the paint shop (…)
No, no it is not. That analogy makes no sense. Colour is part of the light spectrum, it’s not “created”. Paint is, and that’s a physical process.
> We were something different before we were primates (…)
That’s not how evolution works. We don’t just evolve into different things over time. Mutations occur and the most adapted survive while the others die and are eventually replaced. We are short circuiting that process for ourselves and other species via selective breeding.
> (…) humanity should collectively get off the high horse.
I agree, and that has nothing to do with the conversation. Humanity’s “high horse” starts way before the conversation ever reaches computers”, let alone LLMs.
> Serious question (…)
In don’t think it is. You have to define what you mean by “meaningful relationship” (no two people will have the same definition) and “you” (are you being specific or general)?
We could sit here all day arguing about subjectivity.
It's not been my experience that people can reliably tell the difference.
I, personally, would not equate pissing in somebody's soup with having a chat with an LLM but if we must, you need to appreciate that it's the piss itself, that makes the soup delicious, if the soup was not pissed in, it would be middling at best. Only in that context can you then ask yourself, "Am I bothered by this delicious soup?"
I think the analogy works in the context in which it was intended. I was illustrating the absurdity of contempt for something based on its provenance, after praising it. If you really don't like the paint analogy image this, you read a book, loved it, then later found out the author was black, or an LLM, then you hated the book.
I'm not sure I understand where you're coming from regarding evolution. Sexual reproduction is a comparatively recent evolution in the scheme of things. That which came before us defined us, are our ancestors and is our history. What comes tomorrow, engineered or otherwise I would consider our children and should be loved, cherished and allowed to thrive in the same way our ancestors ushered us all into this universe.
I wasn't being facetious when I asked if you could have a meaningful relationship with an algorithm, I asked about the teddy bear to draw a parallel to which a lot of people are familiar. If a child can have an enriched life and what I would call meaningful, intimate relationship with an stuffed toy, why shouldn't they also be able to have the same with an algorithm that talks back to them? I'm being both general and specific if I think about it.
If I could image your brain perfectly and put it into a perfect physics simulation running on silicon or whatever substrate, then have this same conversation with you, I would call that meaningful. Whether or not you would at that point label yourself as living, conscious, sentient, human, whatever. It would be a meaningful relationship to me.
I think people are scared of change, and that's alright.
If people feel this way they should just start a flag this AI post campaign. If people can flag anything they disagree with a little effort the community can be AI free at least on main page.
That said it would be interesting to have a AI flagging so you can just filter the just AI topics. Its very different from people harping on crypto or VC focus. I dislike reading the AI output when someone says they have an opinion the original source material the AI ripped content from to me is more valuable flaws in all than a normalized AI output
>If people can flag anything they disagree with a little effort the community can be AI free at least on main page
Unfortunately AI-based articles are on topic (AI generated comments are not) and if people flag them they'll just lose their flagging rights. I'm also reasonably certain "flagging campaigns" are categorically against the rules, even though there is such a campaign implicitly against "politics."
There's nothing anyone can do about it except complain (which will eventually get them filtered because such complaints are also against the rules), be quiet and tolerate it (which is the only option that follows the guidelines), or else leave to greener pastures.
I run https://www.unslop.news/ which uses an LLM (ironically) to filter out content discussing AI. There's text at the bottom to indicate what fraction of the first 3 pages gets filtered out, and it's pretty much always at 50% or more.
Thank you so much for this! Generally I like to read AI news, yet it's important to know what else is out there!
I hope that ycombinator is implementing this filter by themselves, one day.
I have a comment here from a few years ago where I am accused of it being generated by AI. I wear it like a badge of honor because I wrote it 100% by hand.
"At the same time, it has a fair number of toxic commenters, making it a dependable source of insults hurled in my general direction; if you want a taste, {another article by same author} has been called “watered-down” and “slop”."
We have gotten to the point where it is virtually guaranteed that some commenter to almost any given article will have an oversensitive AI detector and accuse the author of it being AI. I swear there's people whose AI detectors are just
def is_content_ai(content):
return True
Occasionally such people also complain about how much of the world is AI now... well... yeah, that's what happens when that's your detector. It's not quite that bad, as bad as it is.
I don't know what to do about it. Discussion about AI provenance of a given work is often a valid topic. When YouTube shovels an AI channel at me and it briefly penetrates my defenses, it's heartening to see the commentors trying to warn people about it, shortly before I block the channel.
(It's annoying that so many legit channels, which I know are legit because they predate AI and the same person is on camera as before, have switched to heavily-AI thumbnails. That's the hole that the other channels sometimes get me with, because just having an AI thumbnail doesn't prove it's not a legit channel now.)
On the other hand, the chorus of very annoyed people who are very annoyed because they see AI everywhere who see AI everywhere because they literally see AI everywhere are becoming a bit of a nuisance themselves.
> We have gotten to the point where it is virtually guaranteed that some commenter to almost any given article will have an oversensitive AI detector and accuse the author of it being AI.
Well... when the utter majority of articles written about topics surrounding AI are written with AI and a ton of newcomers to the reverse-engineering and even hardware hacking communities now have AI do their work for them, even the last remnants of areas that used to be "for humans only" are now contaminated with AI.
Of course people end up being very sensitive. I come to HN to read words written by humans, not fluff hallucinated by an LLM.
The truth is, writing is hard and time consuming, and most people are uncomfortable with the process. AI is a very easy shortcut to lucid, logical, authoritative text. Like sugary drinks and snacks, people can't resist. Add in the pressure to keep up with everyone else and the choice is easy -- spend a few days drafting and editing, or a few minutes chatting with an LLM before hitting post?
AI has rapidly destroyed, completely destroyed, the remaining authenticity on the internet. The majority of articles are AI generated. And yeah at this point its futile to keep pointing it out. Human prose has been eliminated.
Also like sugary drinks, which our brains perceive as satisfying but are actually the opposite, LLM’s have been well-trained to produce text our brains (and up-voting fingers) perceive as lucid, logical, and authoritative despite the frequent absence of those attributes.
HN is a reflection of the industry and you'd have to be living under and rock (or in denial) to ignore LLMs. People who complain about it give off a real "kids these days"-vibe that has been present on HN for the entire time I've been here. There are _always_ people complaining about the current infatuation (and I'm sure I've been one of them).
I find the posts/comments bemoaning the pervasiveness of LLMs on HN to be more tiring than even some of the more annoying LLM posts.
If HN didn't feature LLM-related content there would be people, rightful so, complaining that we are ignoring a huge thing happening in the industry.
I think its very easy to note and come to the conclusion that this is the most disruptive tech of our industry. HN is full of people from all walks of life, but people from all walks of life are using AI (or LLMs). The way I hear people describe using AI is the way people used to talk about Google.
As much as I am one of the "In Betweeners" of AI (not hyper loving, nor hyper hating it) I do wish HN wasn't flooded with it, but its just the nature of LLMs currently, they are the new thing that has taken over our society in a drastic way, those who pretend otherwise will be surprised to know that the rest of society has moved forward with it.
HN is about to witness the exodus/migration we've seen in prior years, except this time it will be a much larger movement rather than focused/specific top contributors who just said "fuck it" and had enough with this shithole.
The thing is, the AI-related topics are pretty easy to skip. Sure, you may have to go to the second page to see the same number of non-AI stories that you used to see on the front page alone, but you can just not click on them. (Same with politics, BTW.)
But AI-generated content is harder, both stories and comments. HN does moderately (heh) well at filtering them out, but I would like better.
(That's the first one I found, maybe there were others)
So...are you complainers all using it? If not, why not?
Spoiler -- because when someone says "I'm tired of reading about {X}", they're really saying "I want everyone else to stop talking about {X}". This has happened repeatedly over HN's life, and it just betrays how emotional people get when they feel that they can't control the narrative.
Is Pangram's detection capability considered to be on the higher side? This post seems to be written under the assumption that Pangram's detection is highly effective. Is it actually considered to be on the higher side in English-speaking circles?
Pangram links several third party evaluations [1] that all seem to agree with 90+% detection, and very close to 0% false positives. Obviously those could be cherry picked, but I’ve yet to see arguments to the contrary that actually include any supporting data. (E.g here’s a prompt that will get Claude to spit out text that pangram doesn’t detect or here’s an article authored in 2018 that pangram says is AI.) Detection is an arms race, so this could change (though I’d expect in the direction of false negatives), but right now it seems like defense is winning.
The "could" isn't required here, since Pangram calls many of the evals "collaborations", and some received direct support. Results that don't favor Pangram as being among the least-worst of a questionable product category* surely wouldn't be featured.
There is some LLM detection for comments that auto-flags them.
It has some unfortunate side effects, though. Namely on a thread where people are discussing a text-heavy post that is LLM-authored, quoting a large enough block of that content appears to trigger it sometimes. Makes sense that it does, but..
Contrary to popular belief, it's quite easy to spot something written with AI. You just need the skill/gift. I usually spot them rather quickly and will flag on sight. Other users will do the same.
"That said, the recent shift feels more profound: almost every day, it feels that the lineup is dominated by stories focused on AI, written by AI, or commented on by AI."
I interpreted that as surprise. I apologize if you disagree.
I'd just like to say that as an old nerd who has been reading hacker news, often daily, for a long long time, hacker news is an absolute treasure, and dang & pg are saints for keeping it going.
The Magnasanti that we call "Tech" is a little swimming pool that we first leapt into as kids, trying to splash the befuddled saps standing outside the pool. As we splashed and dunked and sprayed each other, a thrilling, unceasing mania began to build within us, and so we splashed all the harder, forcing enormous waves that slapped against the world, shocking it into noticing us. As more jumped in (or were pushed) we demanded and they forced upon us more and more and more water, until we no longer had to entice or push, the waves swept through the whole world, eating at the foundations of entire societies, once solid land crumbling into ravenous waters and spilling their hapless citizens into the vortex, spinning them head over heels, forcing them to cling to the shattered debris of their lives, and try to just-stay-above-water frantically with the rest of us, the billions and billions bobbing in the froth.
And how goddamn nice is it.... that in that raging omniflux, maybe not quite at the center, but maybe also not far from it, is a little sheltered swim up bar, its orange sign still cheerily showing the Y of Y Combinator, which used to BE one of the first kids, one of the ones that pushed and pulled and flooded and floated. They still do, of course.
Perhaps the most amazing thing about this little bar, is that in all this swirling, foaming, madness, it is an oasis of stability. Absolute stability: This is from the first snapshot on Feb 21st 2007: https://web.archive.org/web/20070221033032/http://news.ycomb.... Look familiar? See pg's post getting knocked off the top already? Shelley might have approved.
Sometimes you get to chat with the people forcing the great devouring waves that reshape the earth, sometimes you just chat about what's new in JS tooling. If you are particularly lucky that day, you will find yourself enraptured by some hacker with the mad gleam of triumph in their eye, recounting in excruciating detail their monomaniacal pursuit of some utterly pointless pet project, and you will not consider one second of it wasted. Not even on your deathbed. Not even when compared to spending more time with your beloved.
I think it would be difficult to find an example of a community that has been stewarded as well, for as long, as hacker news, and while you do get commenters of all ages, interests, walks of life, politeness levels, and hyper-fixations upon irrelevant minutia, I've often found that for the most part you are all a lovely bunch of humans and/or AI bots.
Lest you accuse me of labiogluteal angling (a classic!), I am too old and too wet and too cold and too tired to want or need anything from anyone here, beyond a pleasant place to read.
But... I thought that just once over these decades, I'd tell you this how very grateful I am to be able to swim up to this bar, sit on the familiar stools, take a moment of respite from the tempest, and read what any of you might have to share, even -- or perhaps if I am being honest, especially -- when I think I think its absolute bollocks.
Thank you all, and may Tech continue to spin you, right round, right round, like a...
What I meant is that I’ve seen people complaining that the negativity towards AI on HN is manufactured and astroturfed. So we have people complaining on both ends of the debate.
I would bet that the AI booster side is way more astroturfed, but I’m very likely biased (I see genAI as a horrible mistake for human societies)
That's fine, but it would be helpful to see examples. People often make generic claims about astroturfing, bots, etc. that don't hold up in practice when we look at the data.
Does HN do proactive detection and/or flagging of comments determined to be AI generated?
I understand if information publicized on how this process might work is intentionally restricted, so as not to be easily circumvented. I'm also not asking this in a pointed or judgmental way with respect to HN, I'm mostly just curious how various social platforms are approaching this problem.
> People often make generic claims about astroturfing, bots, etc. that don't hold up in practice when we look at the data.
I'd be very careful with believing the system that looks for bots and astroturfing works well enough to confidently say that. Some of the people, corporations and institutions interested in manipulating HN and its audience are quite capable.
Of course it's possible that there's a SSM (Sufficiently Smart Manipulator), which by definition we can't know about. But despite not everything being knowable, there are still things one can say.
Here's an example of what we can say: if user A accuses user B of astroturfing, spying, manipulating, etc., and a quick look at B's posting history shows that they were posting to HN about, let's say, Erlang in 2019, the odds of that accusation being correct go down to negligible. It's possible, of course, that B was already working for "people, corporations and institutions interested in manipulating HN" 7 years ago. It's also possible that malefactors came along later and found a way to pay or coerce B into doing their dirty work. But at this point one becomes unmoored from any evidence at all. The far more likely explanation is that A is slinging cheap accusations because they disliked something B said and are resorting to the internet's favorite trope ("you're a shill!").
When I say "People often make generic claims about astroturfing, bots, etc. that don't hold up in practice when we look at the data", what I mean is that the vast majority of such claims boil down to cases like the one I just described or something similar.
> Here's an example of what we can say: if user A accuses user B of astroturfing, spying, manipulating, etc., and a quick look at B's posting history shows that they were posting to HN about, let's say, Erlang in 2019, the odds of that accusation being correct go down to negligible
about this one tho I have noticed bot behavior (including posts that are incoherent) from old accounts - so there is the risk of 2 things going on:
- people getting hacked and their accounts being used for that
- some people that just decided to connect a bot to their account or use their account themselves to astroturf.
which is way more difficult to spot. Anyway, I don't have a solution for those problems. You have an interesting and difficult work :)) Have a nice day, dang!
AI has legitimately changed everything. It's not 2022 anymore. Try a recent LLM and set it to a coding task. For even moderate tasks, things you could burn a weekend or more on, the model will complete it in minutes, to a high degree of quality. It can write code better than most programmers, faster than any human. If that doesn't inspire positive feelings, it should at least fill you with awe. We live in an age of wonders.
>If that doesn't inspire positive feelings, it should at least fill you with awe. We live in an age of wonders.
It's impressive, despite its ethical baggage and psychological, environmental and psychological harm. But if if filed you in awe I think you need to get away from the keyboard. Things are happening in biology, astronomy, archaeology, engineering, that are far more awesome than software that can write code for you.
And if your only response to that is to be in awe of whatever part AI may have had in the process, you're still missing the point.
As someone who has been posting here since 2010, I don't blame AI. The exodus started long ago and continues to evolve over time. It is unfortunate what this place has become due to poor quality of moderation. Lobste.rs just hits better, and Reddit isn't half-bad as purported.
Did not read the article, but I would say about 40%.
Btw, you can toggle [showdead] in profile settings, then you'll see messages caught by moderators, which are sophisticated bots (writing style at ChatGPT3 level) + absolutely valid points that lash out on ruling class, SV mentality, US and Israel.
Much of these are caught by users downvoting and/or flagging. I think the lack of a public moderation log makes people think the mods are constantly running around every story/comment to steer the site when really it's the users doing most of the grunt work while the moderators only step in occasionally for the more serious items.
This is categorizing with a pretty broad brush, it seems. I took a look at a random day, and the articles that are claimed to be AI are Waymo publishing a model for self driving cars and a post about scams in Apple News ads.
Sure, Waymo’s model is AI and the scams in Apple News are often AI generated, but this isn't some sort of monoculture of LLM articles.
Ironically, this article could have been improved by using an LLM to categorize things and also publishing the prompt, so we could inspect what it means to “be AI”.
The article already states that distinction. It doesn’t say that everything is written by AI, the heatmap even has a legend with different colors for “article about ai” and “written by ai”
The articles I cite were in the “about ai” category not “written by ai”.
I don't know about the article about Apple News scams, but how is an article about Waymo publishing a model for self driving cars not about AI?
You took a look at 5 posts made on one single day and then categorized all of the other days.. seems like you're good at categorizing with a broad brush yourself. Based on my experience as a pretty heavy HN consumer, I'm with the article author.
You took all of one comment to categorize a person, and closed with your subjective experience.
Based on my experience, this isn’t typically done for kind reasons.
But what do I know, I am but one dumb monkey pushing tokens through tubes to what claim to be other monkeys but might be nothing of the sort.
You took a look at all of one comment of a user and categorized them as unkind. Som'n som'n broad brush bad
Fwiw I think Waymo's self-driving AI model blog post and the piece about people using AI to mass produce ads on Apple News are fairly categorized. And the "50% of HN is or is about AI" tagline feels accurate enough.
There is so much more to tech than artificial intelligence (for that matter, there's is so much more to AI than LLMs) but even when having discussions about those other matters it feels like AI is still discussed
The Waymo one definitely is AI, the original point here was that it was AI not just LLMs. The Apple News one really is about fraud and AI as the mechanism is just a sign of the times.
If you think I'm being unkind then I think you're abnormally sensitive.
The inclusion of the Apple News story is sufficient to know that probably the classification here isn’t too tight. It’s not really against the article author to point this out, either, just trying to add color for the forum here. It’s not always us vs them.
Well what I said wasn't really against you either, I was just being cheeky.
The reality is bots and bot-like behavior (honestly, much more likely the latter than the former from what I've heard IRL) are both equally effective in fucking up discourse on HN.
It absolutely has gone to shit. And a number of longtime readers and commenters have called this out as well.
The degradation started during COVID with the culture war crap (a plurality of accounts I bump into that comment and post are now created after 2021 - the fact that 0-indexing and source citations died out on HN is a good example of that shift from pre- to post-COVID HN especially becuase the only reason I started doing it was because JumpCrisscross and Raynier did it in the beforetimes), and then became a culture war after culture war. And at least in the Bay and/or amongst the younger YC crowd, most ignore HN after dealing with in-your-face toxicity or have never heard of it until they got into YC.
Hell, a deeply technical and deeply anti-AI buddy of mine got flamed on HN for supposedly using AI despite him not using any sort of AI on principal.
No one who matters care about HN anymore compared to 5 years ago, and those who are left are basically using it for the same reason a smoker continues to smoke despite feeling indifferent.
Anyhow, YCombinator owns the rights to all comments posted on HN but HN has been continually scraped by multiple repos and data stores outside of US and EU+ jurisdictions (go cry to your GDPR liaison Euro HNers who have increasingly become the plurality).
Let it burn.
plot twist seizethecheese is a bot themselves :D jkjk
If they are so be it. Knuckedraggers and bots act the same way. At with some bots are kinda creative.
I’ve always 1 indexed because there are non software engineers on here and I didn’t need to introduce the confusion that shibboleth would cause
Yeah. There's nothing wrong with 1-indexing, but it's interesting to notice it go from being the norm to being something only a handful of us do now.
There used to be a time when some HNers used to kvetch over not using 0-indexing and that kind of pedantry should have been cracked down.
Well some of use were using R (or later Julia) and have always 1-indexed as an act of defiance... There are dozens of us... DOZENS!
Pour one out for the poor saps in econometrics departments and logistics industries still using SPSS.
dozens of us PVX programmers as well. 1- index gang
...and some of us Fortran.
But Fortran goes skrt skrt (Gen Z/Zillenial for fast).
Honestly love Fortran as a language but really worried about how it's both the foundation of most core packages like Numpy and Scipy and is increasingly underresourced outside of physics departments.
> No one who matters care about HN anymore compared to 5 years ago, and those who are left are basically using it for the same reason a smoker continues to smoke despite feeling indifferent.
Any idea where they've moved on to? Asking for a friend...
Some lobsters. Some blind. Some BookFace. Some BlueSky. Some back to X. But mostly IRL in the core tech hubs.
The ones who rely solely on HN have fallen out of the loop.
> In the first half of the month, roughly 60% of the daily HN lineup was AI-related or AI-generated, tapering off to ~50% as we approached the end of the month. This is up from 40% in February.
This aligns with my much less scientific research. Every now and again, when I remember to do it, I will skim the top 10 stories and count how many are AI-related. I find it is very consistently in the 4-6 bracket. Probably a majority of the time, 5 out of 10 articles are AI-related, and very rarely does it drop below 4 out of 10. It also rarely goes over 6 out of 10, but I can remember it happening a couple of times.
In fact, the proportion of AI representation in the top 10 is so consistent (in my, again, unscientific experience) that I wonder if it is being moderated one way or the other.
Personally I find it boring that one topic gets so much discussion but who am I to argue. I do my bit by making sure to upvote non-AI stories that I find interesting.
A couple of years ago I considered making a simple regex-based web extension to block posts with common AI words: ChatGPT, Copilot, LLM, etc. There was one or two on every page and I didn't want to see them.
Over the next few years it flipped, most of the posts concern AI, and I scan for non-AI topics to engage with. I can't really blame HN for this, it's a massive upset in our industry and the world more broadly.
I did exactly this: https://github.com/ColdPie1/enough-ai-on-hn
I've definitely seen 9/10. It gets ridiculous!
My reaction is: that's an appropriate amount. AI is probably the biggest thing to happen (in tech) in the last two decades, so I think it's fair that it's taking up a lot of the discourse.
I'm sick of it, but you are 100% right. I'm sure a good number of active readers here want even more AI-related articles.
Same. I hate that the TS is right.
Upvoted: I was going to write this exact reply. Everyone is a bit tired of hearing about AI, but it's advancing fast and it's the most important thing right now. As Demis says, if we get to ASI, we can solve all other issues with it.
> it's advancing fast
Actually, it (LLM models) has been advancing slower for a while. This is pretty much in the mainstream. And the current approach is facing all sorts of obstacles that are not exclusively technological. The compute needs are no longer sustainable. Financially, big labs are drowning in debt and backlash against data centres has grown so much that it has become an electoral topic in at least one major country.
why exactly do we want ASI, and further, if we DID, why on Earth would we want it controlled by the likes of Sam Altman, Peter Thiel, Elon Musk, Marc Andreessen, and Larry Ellison?
FUCK THAT
Support DeepSeek and Ziphu than, I am not advocating the merits but it is the reality. ASI will be pursued regardless of wants.
Shit we’re onto thinking ASI is happening soon? I guess people hyped up AGI so much that it’s no longer thrilling to talk about despite not getting to that point either.
> if we get to ASI, we can solve all other issues with it
Jesus Christ this has to be bait. Added to the favourites to laugh about it when the lunacy and hype has died down in 5 years and the trough of disillusionment has truly settled in.
Or to laugh about it when ASI actually has arrived and we're waiting in line to be turned into biofuel for the Greater Good.
Everybody is literally pinning their hopes that God itself comes out of machine to save them from the exponentially accelerating pile of tech debt, and they're not even half joking about it. This is not engineering, this is utter madness.
Its just the next hype cycle that all of the crypto people have moved on to.
Making hyperbolic claims about new tech makes people billions, until that changes you will always see this crap.
This site is owned an operated by people building empires around that principle.
Say what you want about crypto, they didn’t expect Jesus to descend from the sky to shill Bitcoin.
I remember having conversations with crypto shills where they really thought blockchains were going to be used to replace nation states.
They had whole manifestos prepared about how crypto was going to reshape society as we know it.
Its the same exact people making the same exact claims, only difference is substitution of "blockchain" for "ai".
I'd say it is engineering. There's a Moore's law like steady progress where the effective IQ goes up each year which will result in neither troughs nor biofuel but just smart AI.
Given human nature I also think it'll be kind of undramatic. Like when jet aviation transformed things people moan about check-in queues and maybe with ASI people will probably moan their robot butler is only 200 IQ when their neighbours is 205.
If intelligence solved everything, you’d have already solved everything already brother.
Feels more like another blockchain hype cycle to me.
Especially the part where we're always on the cusp of the entire world changing while the biggest change is just that huge sums of money changed hands, inducing psychosis in a large group of investors.
We're not on the cusp, we're in the process.
World-changing doesn't happen overnight. The internet didn't, either.
This is like 1991 internet.
it's more like the flight of the Concorde
Blockchain wasn't this overwhelming though. You could pretty much ignore it if you wanted.
I know what you're saying, there is a lot of speculation going on in AI right now. The difference is there is hard evidence of AI capabilities accelerating now: we've gone from toy chatbots to sort-of useful coding helpers to agents solving novel math problems in 3.5 years. We could plateau any day and a lot of investors would be stuck holding the bag, or we could continue on that trajectory and they'll make crazy profits. With crypto, the bet was always "how much money will people keep sticking in this market?" which seems a lot more flawed to me.
The irony with blockchain is that it also had a bunch of real life use cases but the sheer extremity of the hype cycle managed to actually prevent them from being effectively brought to life.
There was a lot of genuinely useful tech that blockchain could have facilitated and all we were left with in the end was a mountain of scams and bitcoin.
I'm seeing something similar with AI where the genuinely useful stuff is getting washed out by the hype simply because it's a gradual improvement and not "disruptive". Meanwhile the stuff sold as disruptive like vibe coding is mostly bullshit.
What was the useful blockchain tech? It seemed to me to all be different ways of avoiding regulations or bypassing existing system or addressing issues that were already solved by using centralized systems.
it let you decentralize trust. as one example, I was hoping that as well as "sign in with google" I could sign in with a blockchain based decentralized identity service where nobody could one day decide to revoke my identity on a whim.
i remember my friends excitedly telling me how cross border container trade is still done with paperwork because nobody trusts each others' IT systems. I'm pretty sure it still is because they can't be convinced to trust blockchain systems either.
trustworthy voting is another application. ive yet to see a political party that runs an internal voting system that doesnt rely upon you trusting a central server run by a few people. this is crazy.
You have absolutely nailed it.
Funniest part of all of it is it’s all the same actors. Same hardware even. Just the crypto shills now talk about agentic capabilities & we’re approaching a singularity man! Can’t you see!!?
Yes, it's appropriate given the metric you have chosen ("big"ness), though maybe the word should then be "expected", since "appropriate" implies some kind of broader human-oriented concept of rightfulness/goodness.
I would say, AI is probably the biggest thing to happen in tech, full stop.
LLMs went from a toy to highly capable (along many dimensions) in just a few years.
Which makes it very likely that in a few years more, we will be seeing crazy capabilities that leave us humans increasingly behind.
You really have to expand the list to ridiculous levels before you get to something as big. We've already left humans behind in many ways (but in others we've failed to build LLMs that know how to find the gaps in implementation that humans are leaving out)
Cynics really say stuff like "but it's been 3 years and all they can claim is that it's changed text to speech, speech to text, photos to text, text to photos, programming, cybersecurity, research, mathematics, war...
"They lie about it being dangerous!" and simulataneously "They hacked a system with zero day exploits! Everyone involved should goto jail!"
Tech leaders are awful, but it is just that we've built a society that enables the most awful among us to benefit the most. It has never been the case that tech leaders are particularly amazing, they are just amazing at exploitation.
> "They lie about it being dangerous!" and simulataneously "They hacked a system with zero day exploits! Everyone involved should goto jail!"
FWIW:
The simultaneous "it's stupidly weak" and "it's dangerously strong" is a well known propaganda technique.
> You really have to expand the list to ridiculous levels before you get to something as big
Instead of expanding the list, it would be better to widen your vision. Email was more transformative than AI is right now. Unix, then Linux was more transformative. If you have a magic wand and remove AI, I think barely anyone would be impacted, but removing something like windows would be like erasing computers. And removing Linux would wipe most of the internet services.
Email in it's first five years barely extended beyond some research universities. And is it novel? Do you see email as somehow a difficult concept to imagine?
Unix in it's first five years reworked the way operating systems performed. I'd agree it's pretty important, but I don't see how it fundamentally changed the way we perform work in multiple disciplines (in its first 5 years). Same for Linux, but even less so.
We're in very early times, and generative AI is touching every corner of the workpace and education, for better and worse. Powerful technologies can have negative effects. Email has been negatively transformative in ways as well.
Part of the reason you can erase AI so effectively with so "little" effect is that it is new.
There's only a few months left for the core technologies of LLM, the transformer, to be a decade old (counting since the "Attention is all you need"). In a decade, we have gone from pure theory of computation (Turing paper in 1936) to the first general purpose computer (ENIAC in 1945). In 1971, the first email was sent over the network (ARPANET), by the early 1980s, we were looking at nationwide deployment of email systems.
> We're in very early times, and generative AI is touching every corner of the workpace and education, for better and worse.
Forcefully. By a mixture of free offerings and bundling AI in every tools with no disable button. It's still not an essential tech like email or Linux is (not even essential at the level of GitHub).
This is an entirely weird and bizarre conversation. The world today has network effects - specifically because of well, networking. It feels almost ridiculous comparing the impact of things like ENIAC to github because they aren't even in the same class. GIT and email (FYI, there is a predecessor to email in 1965 called MAIL) are specific implementations of systems that would have happened no matter what. GIT is awesome, email is kinda ok -- but that's beside the point.
> Forcefully. By a mixture of free offerings and bundling AI in every tools with no disable button. It's still not an essential tech like email or Linux is (not even essential at the level of GitHub).
We could survive without computers and the internet. We did it. I was there. We could go back. It's not essential. You've lived maybe 3 years with LLMs. No kidding they could disappear. As an old person, I have no idea why you think any other technology is any different. LLMs are new, that's the big difference as I said earlier.
> Forcefully.
So.... I hate mobile apps. I hate that TVs connect to the internet. I hate that we have cameras on every corner. I hate that we have cameras that watch us while driving.
Now, tell me about how this is different than all that.
> As an old person, I can tell you something: We could survive without computers and the internet. We did it.
As late as 2009, I was living without computers and internet. The most tech thingies in everyone home were the radio, the TV, the VCR (not everyone and half of them wasn't working properly) and a dumbphone.
I'm not saying that internet or any of the others is essential (in absolute). What I'm saying is that the supposedly "transformative" aspect of AI needs free offerings, dark patterns, and a mountain of hyping to just make it worthwhile to discuss in tech circles (removing the grifting on Linkedin). For everyone else, the most impact is the enormous amount of slop.
I paid for my computer. I'm paying for my internet connection and my electricity. I'm paying for my email service. Drop the free offering for any current AI service and see how much it is that transformative.
> As late as 2009, I was living without computers and internet. The most tech thingies in everyone home were the radio, the TV, the VCR (not everyone and half of them wasn't working properly) and a dumbphone.
The VCR? Hrm. What part of the world are you from? No computer and no internet? Hrm.
> Drop the free offering for any current AI service and see how much it is that transformative.
It is currently being paid for by millions of companies in Microsoft products, and I can tell you that many companies are happy with it.
Microsoft honestly has done a tremendous job integrating it with some of their products, less so in others. (Excel is a disaster, Teams has been pretty good)
I would say, AI is probably the biggest thing to happen in tech, full stop.
I would say, the transistor was probably the biggest thing to happen in tech, full stop.
I think you will change your mind when AI will become smarter than any human.
Even if that happens, it depends on the transistor, and thus is less big of an innovation than the transistor.
- A successful surgery may require both a surgeon and a nurse, but that doesn't mean the nurse has nearly as much impact on the successful surgery as the surgeon.
- The lighting of the match is mainly caused by the striking of the match, not by the presence of oxygen, even though both are necessary for lighting the match.
So a cause being necessary for an effect doesn't imply that its impact on the effect is very large.
I think microprocessors were the biggest thing, until the internet, then the web, open source, system on a chip, virtualization, cloud, ..
When each of those things came in, there was definitely some stumbling but I remember things being less ambiguous about being good than LLMs.
> I think microprocessors were the biggest thing, until the internet, then the web, open source, system on a chip, virtualization, cloud, ..
and none of them changed society more than the smartphone (starting with the iphone).
The ambiguity partly comes from denial. LLMs in the end, to be brutally honest, replace us. That is the direction it's headed and some people can't face the cold hard truth.
Or are tired of the hyperboles and straight up lies. Will you hand over your bank account to an agent? Can it grow food and sew clothes?
If it can't yet, it will very soon. In less than five years probably.
And what is justifying such belief?
reality as we know it
Extrapolation of the extreme AI progress we have seen in the last few years.
My 3 month old son is now twice as big as when he was born.
He’s on track to weigh 7.5 trillion pounds by the age of 10.
There’s nothing proving that this technology won’t be an S curve growth like every other before it.
AI is not limited to the rigid construction plan of an organism. A model also doesn't have to fit inside a small skull or be limited to the daily energy intake of a human.
> AI is not limited to the rigid construction plan of an organism.
This is some AI psychosis shit.
Explain how you think extrapolating off of 4-5 years of growth(I’m being generous here) means the development is asymptomatic instead of reaching a plateau like every other aspect of physics.
Are you one of those people who believes Anthropics claim that they will have the revenue of the entire United States yearly for their IPO?
You ignored my previous explanation.
Which previous explanation did I ignore?
The only one I saw upon review was
> Extrapolation of the extreme AI progress we have seen in the last few years.
I don’t think that’s a viable extrapolation hence my mockery about a 3 month old being trillions of pounds by the time they were 10 years old.
If there was another explanation please let me know so I can evaluate it.
> AI is not limited to the rigid construction plan of an organism. A model also doesn't have to fit inside a small skull or be limited to the daily energy intake of a human.
We haven't seen any significant progress at all, much less extreme progress. You still can't use these stupid slop buckets for actual useful work, despite the massive amounts of resources being wasted on them.
You are being delusional. LLMs are extremely good at programming and mathematics, as well as many other tasks.
With an actual good programmer at the wheel, so it’s like saying SO is good at programming or GitHub is good at programming.
It still is a toy.
There's a part of me that agrees.
There's also a part of me that misses the earlier days of Hacker News when the subject matter was much more eclectic and seemingly every day I'd have a chance to learn about something completely different.
And that's maybe what I miss. Nowadays Hacker News feels increasingly like another social media platform: mostly a lot of opinions, war stories, and self promotion. That feeling of serendipity that so characterized earlier ages of the Web is fading here, too.
It seems a bit self-reinforcing. AI makes almost zero (positive) impact on my day to day life and isn't actually something I interact with directly and the same holds true for most people I talk to.
There's a few of my friends in tech for whom AI has obviously massively changed how they work, but again not in a clear positive direction.
To me there seems to be a push to heavily prompte LLMs, even thought the benefits aren't all that clear yet. Some people are seeing massive improvements in how they work and it's understandable that they, like with blockchains or Kubernetes previously want to share their optimism. I just question if maybe these people are being to louder and have to much influence. There's a large silent majority for whom all the AI stuff is almost completely pointless.
It does need to take up more space on HN, compared to your normal media, because the technology needs to find it's way into applications and appliances. The hype will die down, especially once we start to see the actual costs, e.g. ChatGPT doesn't actually provide enough value for the average user to warrant a subscription, hence the ads. For now, it seems that AI is heavily pushed by industry and a group of fans, the amount of news and articles just doesn't line up with the actual daily use.
I don't think that AI has to influence every single person's daily life to make a huge impact. I think for probably 90% of people it's pretty useless out side of being a better search engine. But that doesn't mean it won't have a huge impact on science and technology at large. There are a lot of snake oil salesmen trying to stick it into everyone's face, but I don't think that should be conflated with the real progress that's being made.
I would say 90% of people do not yet realize how useful it is. It is not useless to them, they just don’t understand yet what it can do.
And there were a few years there where a lot of BS AI hype turned them off to it. So a lot of people rightly dismissed it and don’t realize the inflection point it hit in the last few years.
ChatGPT work is basically a personal assistant for $20/month
> ChatGPT work is basically a personal assistant for $20/month
A personal assistant that will happily tell you that you’re the world’s largest supergenius who invented perpetual motion. I wouldn’t want that if they paid me $20/month to use it.
Current Opus and Fable simply don't do this. You post is stupid.
There’s more than 2 LLMs in the world. Now whose post is stupid?
i just gotta say that i find it hard to believe that this comment is true for you but you hang out on hackernews.
almost everyone here is a software engineer or works in tech, they all use ai all day to do their entire jobs.
> even thought the benefits aren't all that clear yet
it isnt 2023 anymore. wtf are you talking about.
>he hype will die down, especially once we start to see the actual costs, e.g. ChatGPT doesn't actually provide enough value for the average user to warrant a subscription,
again, it isnt 2023 anymore. none of this is true.
> they all use ai all day to do their entire jobs
I assure you they do not, many truly do not care and don't want to. That's just what the hype and the article and the stories tell you. The HN bubble tells you that's true, but 90% of my co-workers don't even want to touch Claude, despite being given a license.
And again, if e.g. ChatGPT is providing so much value, why aren't people buying?
> why aren't people buying?
They have "more than 50 million consumer subscribers" and "More than 9 million paying business users" [0]. That's for example more subscribers than for Adobe Creative Cloud.
[0] https://openai.com/index/scaling-ai-for-everyone/
> more than 50 million consumer subscribers
So, about 0.6% of the world’s population is willing to pay for something that’s supposed to be hugely useful for literally everyone (as opposed to Adobe products, which are only for digital artists)?
Not everyone in the world has $20/month of disposable income, nor does everyone have the disposable time to learn how to make it hugely useful to them.
Also, these are just the numbers for OpenAI. A lot of others are ok with e.g. "paying" for it by going to Google's AI Mode, and getting some ads.
> nor does everyone have the disposable time to learn how to make it hugely useful to them.
True, I’d rather spend my time to get useful skills whose usefulness cannot be destroyed by a single corporation deciding to pull the rug.
> A lot of others are ok with e.g. "paying" for it by going to Google's AI Mode, and getting some ads.
Since everyone who uses Google the search engine is forced to use Google AI, we can’t get meaningful statistics for those.
So ChatGPT cannot deliver you $20/month worth of productivity to offset the cost? That's at least my point, for developers, businesses and a ton of use cases, access to a good AI is worth every penny, because it easily pays for itself. For the average consumer, that's just not true, apparently.
HN is in an extreme hype bubble around LLMs, because of AI coding. Normal people outside of the tech bubble range from 1. not caring at all about AI to 2. disliking or opposing it.
Meanwhile, here on HN, you have people who earnestly say things like "AI is the subtrate the future runs on"[1] (actual quote).
1: https://news.ycombinator.com/item?id=49417949
Just saying "it isn't 2023 anymore" isn't an argument. Yeah it's not 2023. And LLMs still can't actually do anything useful outside of sift logs and write small scripts. They still hallucinate things regularly so that you have to check everything they do, which negates the time savings they supposedly provide. And that's not even getting into the externalities! LLMs are a very crappy technology.
I guess Hex-Rays and Vector35 must be hallucinating their effects for, say, reverse engineering. OFC rev isn't be all and end all of computing but the effect LLMs had on security research and adjacent spaces (say, CTF) is interesting
https://github.com/HexRaysSA/ida-nexus
https://sidekick.binary.ninja/
> they all use ai all day to do their entire jobs
I am so fucking sick of this. No, you dumbshits -- ALL of us DO NOT use LLM's to write code. Just fucking STOP assuming every developer needs the same crutch that the rest of you can't seem to live without.
I use it a couple times a week, not even daily. It's useful about half the time, and a complete waste of time the other half. It's a useful tool to reach for sometimes and I'll keep using it for the things it's good at, but it hasn't really changed how I work. Shrug.
LLMs have helped us keep HN up and still (relatively) fast amid a huge onslaught of new traffic over the last year or so. We've been able to solve problems and find opportunities for optimization that we'd never have gotten to without the help of Claude or ChatGPT.
Although this is one of those threads where people love to report how much they dislike HN, how it's a poor shadow of its former self, and how they barely look at it anymore—a popular local sport for 15 years or more—but to the extent anyone still reads the site, there must be some positive impact in having it running and rendering pages, even if you don't like what's on them!
Hopefully HN is just getting more popular, but isn't a large part of the "huge onslaught of new traffic" driven by AI companies scraping the site for content or writing comments for various reasons?
Fixing the problems created by LLMs using LLMs seems like a bit like hiring a bully for security to prevent that same bully from beating you up.
But yes, there's plenty of positive things that are now possible. See as a combined net-positive vs net-negative, I'm still leaning towards the net-negative.
I fear this is not the only reason even if it is true. The more negative reason is a loss of skill/talent. We used to be able to know what to do in these situations, and how to fix things when they went wrong. Companies like Y Combinator used to be able to hire talented systems engineers who knew how to minimize the risk of things going wrong, and fix them when they did.
Now with LLMs it is just vibe-coded until it is fixed. Find the right prompt and hope the statistics is in your favor. If this is the case, there is nothing positive about it.
I take your point! But the positives are still positive. I've been amazed by how much LLMs have helped us do.
None of it was undoable otherwise, but we (where by "we" I mean "I") would never have gotten around to it otherwise. Lowering the bar from months to days, or weeks to hours, is significant.
that new traffic likely was from AI and their scrapers. So no, not much of a net positive.
I mean, scraper traffic should be pretty easy to tell.
Now, AI posting bots are much more difficult.
I think the scrapers have gotten pretty good and can imitate real traffic. And it isn’t just scraping for model training. A large amount would be LLM tool calls searching the web (and HN) for info about some prompt.
>because the technology needs to find it's way into applications and appliances.
I hear what you're saying, but speaking for myself: can we *not*? I really have absolutely zero interest in having ChatGPT on my microwave or refrigerator. In fact, I'd rather have my appliances *not* be connected to the Internet, all things being equal.
>I'd rather have my appliances not be connected to the Internet,
It's getting harder and harder to find things that don't connect to the internet as time goes on. The moment a manufacture figures out how to sell your appliance data for more than it costs to add it, seemingly every model they produce gets it added with the only choice of going to some other manufacture if you don't want it... and even then finding one not internet connected gets nearly impossible.
In general most people don't care enough for people like us that don't want connected devices to move the dial.
What are you doing in tech that has not been impacted by AI?
Problem is other cool hacky stuff gets drawned out. Nobody votes on those because I imagine it's no longer attractive to the type of visitor that HN attracts now. I reckon there's many more non-tech, "make me a millionaire" gold rush crowd here, as cynical as it sounds. I base that on pure logic, since it's kind of inevitable with AI having delivered a massive democratization of tech knowledge, but also on the amount of snarkier, shorter, less technical, less logical, one short sentence type of comments.
I know it's been said countless of times before, but I think this community has truly changed within last year.
Bigger than the Metaverse or monkey NFTs?
> AI is probably the biggest thing to happen (in tech) in the last two decades, so I think it's fair that it's taking up a lot of the discourse.
Almost. the iphone was launched 19 years ago. It was (and still is) much more relevant than the current LLM-based AI.
The worst IMO is the bot accounts. And the worst of that IMO are the people that behave as if they were bot accounts. The self-promoting thought-leading experts (usually on AI) with a skin as thick as a single cell membrane.
At the moment, half of this site on a given day is terrible ads with a swarm of the same people defending each other's attempts at either becoming or staying relevant - utilizing the latest gold rush topic to do so.
___
I'm not quite sure what could be done about this. The moderation does do a good job, so it's not that.
I also see that not everyone feels repulsed by the self-promotion. So maybe there is no problem and/or the problem is me. But I kinda doubt that too.
Maybe the time of voting systems for ... uh.. "social media" is somewhat over-ish? Though if I was truly signed up on that thought, I would not be commenting here but on I guess lobsters?
AI has turned me away from HN more and more.
Most of the content is about using some data center based ML provider. My view is how so many ignore the environmental harm these constructs are doing; increasing ambient temperatures, inducing heavy metals into the water table, decreasing the access to water for human consumption in numerous drought areas, hostel ambient noise ... The wealth who are the majority finical backers do not care about these issues.
ML is a tool and hardly focused on it being able to operate as a local based solution. Frameworks are tie to expensive hardware. This creates a hyper pay-to-play scenario. While hand written coding can be done on a computer from 20 years ago. Local based solutions are the only means to decrease the wealth gap. Data center solutions increase the wealth gap.
I don't find value in ML summaries. I want to consume the answer and the ideas around them that lead to the answer. This creates more long-term value.
_Lobsters_ has now becoming my go-to with their posts being focused on ideas and concepts versus advertisements for large corporation ML based solutions.
> increasing ambient temperatures,
That study was so bad that I can’t believe it’s becoming an actual concern that people use against data centers. They were barely able to register a temperature increase a couple hundred meters directly downwind.
> _Lobsters_ has now becoming my go-to with their posts being focused on ideas and concepts versus advertisements for large corporation ML based solutions.
Lobsters has a funny obsession with being anti-AI, but the most commented posts on the front page are usually AI related.
One AI psychotic's "barely able to register" is another person's "increasing ambient temperature".
> Lobsters has a funny obsession with being anti-AI, but the most commented posts on the front page are usually AI related.
The top five most filtered-out tags¹, excluding meta², are (in order): vibecoding, dotnet, php, emacs, ai. Two of the five are AI-related. The top spot (vibecoding) has 60% more people filtering it than the second spot.
¹ https://lobste.rs/filters
² In the sense of being about the website itself, not Facebook’s parent company.
Wait, how is Emacs AI-related aside from being written in Lisp?
It’s not. Neither is php or dotnet. I meant to write “two of the five”, not “three of the five”. Fixed. Sorry about the confusion.
I'd argue dotnet is about AI. The official blogs and conferences have been hijacked by AI products, much like VS Code.
So less than 4% of the userbase filters out vibecoding stories? This looks like a big red herring.
Every time I open up the lobste.rs home page, there's an 80% chance the most commented on stories are related to AI in some way. Other stories are lucky to get 10 comments.
> Lobsters has a funny obsession with being anti-AI, but the most commented posts on the front page are usually AI related.
Reminds me of gamer-gate: Back then, NerdCubed graphed the top posts from both pro- and anti-gamer-gate subreddits, and most were meta posts about themselves and each other rather than their alleged topics. Gosh, that was such a dumb period.
It is a tragedy that a handful of irresponsible outlier datacenter operators have soured public opinion so thoroughly in the past 2 years. For comparison, Loudon County, VA is among the wealthiest in the country and is home to the largest aggregation of datacenters in the world. Yet these datacenters don't have the externalities that are mentioned here.
Note the intersection between eyewitness accounts of noise, water, and light pollution from datacenters and those datacenter owners' noteworthy political activity. To be specific, the operators of these datacenters are notorious bad actors in many different public spheres.
Edit: the truth is obviously very painful and unpopular in this thread.
I really wish there was a way to filter out discussions about AI on HN. It feels like lately they take up a good 1/3 of posts on the front page. I don't want to hear about some new AI startup.
Having a set of tag presets would help somewhat for filtering.
Yeah, I like the way lobste.rs does it with the vibecoding tag
HN has been strongly adverse to tags for quite some time - for legitimate reasons. I don't think we'll see it anytime soon.
For reasons. The legitimacy of which are debatable.
Lobste.rs seems to get along fine with them.
Personally, I've forked the hnrss project to create two separate feeds (ironically, quickly implemented using AI).
It just a basic filter based on keywords, but it helps.
I made a similar site https://unslop.news/ but rather than keywords it actually feeds the title and article to an AI which classifies it as "about AI" or not to filter out the AI discussion. Also it's open source: https://github.com/leiDnedyA/hn-without-ai . Hope it helps!
Thank you for this. Bookmarked.
https://www.unslop.news/
https://hcker.news/?ai=exclude
Writing to remember this
If you click the timestamp on a comment, you can bookmark it.
1/3? Try 2/3 or more. And the technical articles I used to enjoy reading are outright gone.
This is an entirely different site from what it was just a few years ago, even if the orange tint remains the same.
But you still have the downvotes and the flagging in case you fall out of line. Oh, and the "You're posting too fast, please slow down."
There is a really easy way to do this. Ask an AI to write a browser plugin to filter out the AI discussions.
Ask AI to save you from AI.
Well your point on it being wrong to interact with a whole industry sounds more like a policy and political party line. This is just nerds discussing bleeding edge things.
Bleeding something, alright.
I'm sure some of the career chemists et al at IG Farben and BASF felt similarly in 1945.
There is no amoral middle ground where any of this is detached from its context. Context is everything.
Do you have lobsters invites?
Sure do - invite sent to the email indicated by your HN profile.
Thanks!
i wouldn't mind one as well if at all possible?
a first look through the comments on some first page threads and they seem far more thoughtful and considered. it generally seems like a nicer commenting environment that might actually do me some good moving forward.
I don't even care about the environmental aspect- it's just so...un-hackerlike and boring. Where are all the crazy ones, the misfits, the rebels?
Not in Silicon Valley tech lol!
Your parent is referring to a nostalgia for the Silicon Valley hacker ethos from the 1980s.
> "Where are all the crazy ones, the misfits, the rebels?"
They've been driven underground or entirely out of software by the mundanes but pointing that out draws huffy accusations of "gatekeeping".
i feel the same way. there's nothing interesting or curious about someone writing prompts to outsource a bunch of effort/thinking/work off to some most likely continuation sequence predictor / set of most likely continuation sequence predictors.
i miss seeing clever and thrifty solutions to weird problems. hell, i miss people just doing plain weird shit and posting that.
A lot of those environmental claims are not accurate, or substantially overblown.
Heavy metals in the water table is a new one I haven't heard before, and I'm skeptical.
Most of this is motivated reasoning by people who don't like AI for other reasons and are just looking for justifications to shut it down.
Although there is some irrational hysteria about the whole "chatgpt is drinking all our water" thing, to me the biggest issue (other than the mass piracy) is local energy bills going up without contributing to the local economy. Seems like it would do more harm than good for all but the owners of the datacenters.
I might migrate if I could could an invite, but I haven't been able to find anyone on the inside.
Edit: Thank you kind stranger. I'm in.
I got you - invite sent to the email on your HN profile.
If you got any more, I'd appreciate one too. My email is my username at 'pm' dot 'me'
Invite sent, be kind & have fun
How about me? If you’d be so kind. Thanks!
I used to use Lobsters, but I was banned for making a stupid joke about emojis.
> AI has turned me away from HN more and more.
AI has made me aware how opening a tab to HN is a compulsion I no longer truly enjoy. Especially after having made the conscious choice of leaving software engineering as a career, my time here is a mildly negative experience hoping for the random chance of finding a deeply technical post that is not about LLMs, or whose discussion does not devolve into discussion with Joe Rando on how much more productive he is lately.
Sure, it's a moral failing of mine, but by god I wish we had a little more diversity in topics.
Consider trying https://unslop.news/ , I tried to solve this exact problem by classifying post title and webpage contents as AI or not and filtering out the AI stuff. Also, tangentially, if you're looking for a really impressive technical writeup I recently enjoyed this one :) https://philo.gay/linecam/
I appreciate the way you sum up the ML vs traditional software appeal and I think it's helped me articulate why ML is so unappealing to me. The power of software has very little to do with the power of the machine it's run on (with some obvious exceptions like games and physics sims) which made it very exciting to me when I was learning to code. The same cannot be said about deep learning where more data and more compute means more power.
Democratization of access has its victims.
I miss the old newsgroups. I wish we could land again in that weird, colorful oceans of geeky individuals with their quirks and interesting nerdy takes.
15 years ago, HN was about bootstrapping and early YC stars.
10 years ago, HN was in full VC-enterprise sauce.
5 years ago, it was crypto and NFTs, and how I switch jobs every quarter.
Now it is a blurry view into the future through the black mirror of AI.
How could we jump back into the cozy past again?
I do not have a clear answer, but what seems to work for you are small groups of friends and friends of friends hanging out on WhatsApp or Discord, discussing geeky news together in the context of their backgrounds. But these are very curated and very closed groups, still operating on the social media platforms of behemoths.
Would now be the right time to restart self-hosted message boards or Mattermost/Campfire instances?
Run a self hosted forum or instance of whatever you like, and just only let people sign up if they are recommended by someone or can argue their way in, basically. xenforo and some add ons can do that, and I'm sure there's a few self hosted messaging apps like you mention but I'm not as familiar
IRC is small and niche enough to be a good place for more ephemeral discussion if you have a bouncer to stay connected for you. Chat on it tends to move slowly, and the userbase trends more towards actual nerds.
I do not think this is a technology problem.
I think it is a problem of how we changed: shorter attention spans, and a much larger volume of information to process.
Short forms win, long forms vanish, which is a tragedy in itself as good writing requires good understanding, and understanding is impossible without attention.
I'm in my 40s. I remember lcamtuf from pl.comp.security when I was just getting started playing with FreeBSD and other fun stuff.
I still use a forum like that. You don't need an enormous community to have good discussions - 100 active users creates quite a lot.
It's a community issue, not a technical one, but you have to intentionally create the community.
Yes exactly. People have changed too. We're not going back. But we can hope that something new and better can one day appear.
I joined HN 13 years ago (2 years after you), and back then I saw it as a bunch of libertarians and venture capitalists and some open source (and/or functional programming) enthusiast with a few interesting and nerdy projects. I was mostly here for the news about new and useful technology.
I think the visual weight of the first group has increased over time as their projects of making a bunch of money while offering nothing of value has been validated more and more by our capitalist economy.
I joined in 2008, that was even after lurking for a while, and it’s almost unrecognizable now in comparison has been for many years. Was mostly entrepreneurship, tech was present but was a means to an end. It wasn’t tech for the sake of tech, which is what it feels like in the recent past. The discussion on entrepreneurship topics seems to have an employee mindset more so than an owner mindset. I find it similar to how renters vilify landlords but being a landlord is not massively glamorous and can be quite risky, renters are generally out of touch with the costs involved and the profit of a landlord is all I’m saying. There’s definitely some shitty landlords tarnishing the reputation of them all. This is what has happened in tech too, the startups we loved turned evil. My gut reaction of any new tech product is no longer curiosity, it’s mistrust.
Anyhow - I find tech stuff interesting to a degree, but fundamentally I’m an entrepreneur that sees it as a tool. A means to an end. And I would love to find a place that still has that vibe. But anywhere entrepreneur related is always a cesspool of grift so it’s hard to find.
I specifically remember there being a similar complaint about bitcoin back in the day, and I think somebody even SHOW HN-ed an HN front page without news about bitcoin (I went searching but failed... instead I found these two ASK HNs: https://news.ycombinator.com/item?id=20385116 https://news.ycombinator.com/item?id=5601876).
As for the employee-vs-employer. I am for sure part of the problem there. When I joined I immediately saw the potential of this being a forum where tech workers can form an international solidarity and complain about our bosses and started using it as such. There is no way you can have that many workers in an industry together and we don‘t complain about our bosses, that is just not going to happen. Workers complaining about our bosses is a universal behavior.
I am part of the problem in the first. It's because early HNer's were early Bitcoin users and adopters and I feel we (I) had a hope that it could take hold as an actual currency (I no longer feel that). But, we were so early I can say my first Bitcoin purchase was for <$10 and most of the Bitcoin I own today has a cost basis <$30. For us, it's climb was fast, wild, and wealth inducing. I thought it was insane when it quickly went to $100 and I think one of my own HN posts that has been maybe my most upvoted ever was about it hitting $300. I often wonder how many of the bitcoin wealth you hear and see about IRL, were just HN users that heard about it's early traction and had enough curiosity to buy a few coins (a cup of coffee then, is worth a house now). It was a PITA to buy coins though, I luckily used Mt Gox and moved them out prior to them getting "hacked".
I put my Bitcoin on ice, refused to sell it, my initial investment thesis was "what happens to the price when all Bitcoin is mined", given the supply is fixed, the price should surge, was my thought. I think they have or will find a way to alter the algo so that even more can be mined. But once I made that decision, it became difficult to see all the Bitcoin related news. I wanted to just not think about it. Daily reminder that I could sell today at a crazy high gain is tempting. I don't even know if that thesis makes sense given the projection of when that last coin will mine. Those projections didn't exist when I put this into place. So I hodl for no reason probably.
In terms of the posting and ethos, it was really all the non-Bitcoin crypto crap that came afterwards that I got super tired of. But I am older and know it was just the current hype cycle, just like AI today, something else will come tomorrow. Although AI is kind of a big one, it's both massively hyped and useful too. It's also moving faster than I can keep up with, so I just learn to dabble where I have interest. I'm not trying to be a creator of AI or start an AI startup, it's just tinkering and when it comes to tech as a means to an end, it allows me to move faster. But it's also given everyone my superpower. Kind of a net zero with extra steps. It feels like my beloved woodworking hobby has been changed to IKEA furniture assembly which is not fun at all.
I never cared much for newsgroups, they were mostly nice for the historical stuff. I do miss IRC though, as toxic as it could be. I could run it in a little terminal on any system without download electron apps and being spammed with ads (discord).
The issue is the tech and the motives to wield it. Cosy decentralized message boards exist, thankfully outside the realm of the greater public.
The moment that the greater public sees these high trust networks and tries to join them, all the bad actors and their armies of bots will follow and I mourn for the recreation time of any admin who has to suddenly deal with that.
Private walled gardens seem to be the only way forward for now. If you love something, do not share it here
I'm nostalgic for the new JS framework every day era. Say what you will about that era but people were writing the code by hand, so looking back it all feels "hand-crafted" and quaint. You had to copy paste that express app from a medium.com article instead of prompting an LLM to scaffold a whole project in one second.
Is this comment... satire?
>You had to copy paste that express app from a medium.com article instead of prompting an LLM to scaffold a whole project in one second.
You mean they had to import it from NPM and compile it from Typescript, right? I don't think anyone's copied and pasted JS since JQuery was a thing.
At least there was some genuine effort involved there. What I find most disheartening about the current wave of AI-generated projects is that there's often very little personal involvement - someone had a general idea and asked Claude (or whatever) to implement it for them, and the results are much more representative of the AI agent's tendencies than the author's intent.
> Democratization of access has its victims.
Democracies require conserved stake to function. How individuals initially get that stake is a degree of freedom, but if the stake can be printed freely, the system is unstable.
For working implementations of democracy, look at: invite trees, shareholder corporations, proof-of-stake distributed ledgers, etc.
I thought we were all about inclusion these days. I'd be wary of sharing such retrograde sentiments, you might hurt an LLM.
Matrix works fine for the friend to friend, small community things. Discourse forums too. Use Signal for family and friends, perhaps work. Fediverse and Atmosphere (Bluesky) for your microblog needs, and RSS to keep track of blogs.
Having used Matrix for almost a decade, it's hot garbage.
Sync is barely working, "unable to decrypt" still exists, the encryption is flawed [0] and the whole thing is effectively just Element and everyone else follows their lead.
[0]: https://soatok.blog/2026/02/17/cryptographic-issues-in-matri...
Element in particular is embarrassingly bad. (Consistently, for the past 5 years, through the present)
About the post you linked, see [1] and [2], which makes me question both whether that is an actual vulnerability (Signal, the messenger recommended by the author, also didn't have that check, and it's addition is absent from release notes and CVEs) and whether the post was made in good faith (the check was added to libsignal on the same day that the author disclosed the vulnerability to Matrix, which can be a coincidence, but doesn't seem likely given that the code has been there for years).
Can't speak about the other issues, I haven't had those but I barely use it. Based on comments, it's clear they exist or have existed for quite some time.
[1] https://blog.erinshepherd.net/2026/02/non-contributory-keys-...
[2] https://matrix.org/blog/2026/02/analysis-of-reported-issues-...
> About the post you linked, see [1] and [2], which makes me question both whether that is an actual vulnerability (Signal, the messenger recommended by the author, also didn't have that check, and it's addition is absent from release notes and CVEs)
I've edited the above linked post several times to make this clearer, but people still keep linking Matrix's flawed response as if it actually addressed the issue. It does not.
The issue is not as simple as "did the check exist? [y/n]". There are actually two issues:
1. Did the check exist? [y/n]
2. Did the group key agreement protocol fall to shit if the check was omitted? [y/n]
Matrix was (n, y) due to how Megolm manages keys in group contexts. Signal was (n, n) due to their protocol design.
Matrix tried to defend point 1 when the blog post was about both 1 and 2.
(Post-Quantum MLS, for comparison, would also have been impervious to this issue due to how KEMs work. Another point in favor of MLS adoption.)
> and whether the post was made in good faith (the check was added to libsignal on the same day that the author disclosed the vulnerability to Matrix, which can be a coincidence, but doesn't seem likely given that the code has been there for years).
Even if you assume the worst possible interpretation of this coincidence (which, sure, you're free to if you want-- you would be wrong if you did, of course): What does it even matter?
Like, what is the notion of "good faith" that's even necessary for someone to do independent security research and publish criticism of the cryptography used by a software project that's being propped up by EU sovereignty movements?
In my opinion, this is a stupid and vacuous framing.
Signal omitting the check didn't harm Signal's confidentiality. Matrix omitting the check demonstrably did.
---
Also, if you're going to decry my blog post for not enumerating unrelated security bugs (n.b., the blog post was about Matrix, not Signal, so why would I talk about Signal's code in that post???), be aware that I reported non-cryptographic bugs to Element (Matrix client) and Conversations (XMPP client) around the same time. Only XMPP actually credited me with any fixes.
https://gultsch.social/@daniel/116125064866469085
I think you answered your own question, simply un-democratize access. Go places with high barriers of entry.
You don't even need explicit barriers, the masses are staying on Reddit/Twitter regardless. Only the discerning are seeking out better and newer options and those options are staying the right size rather than absorbing the eternal September.
You forgot to mention the "JS-framework-of-the-month" years :-D
also known as the good old days
Some weeks ago, a colleague laughed at me when I asked: "btw, do we use jQuery somewhere in our web app?" - he said, its no more a thing today.
Is that true?
(Im out of webdev for years)
jQuery is certainly still used in a lot of production apps and they still release updates.
But most new web apps will use a flavor of React.
I bet half the devs who say 'we don't use it', probably have it on their site and don't realize it.
On the topic, I agree smaller groups are the only way to achieve this.
Here is a protptype of how easy it is to setup a forum on onionv3 service. Just a few commands and your off.
https://gitlab.com/here_forawhile/torum
Discord is actively torpedoing small, niche, or marginalized communities. Not a safe place.
Torpedoing how?
the composition of HN has changed to be a lot more leftist and less capitalist, and I don't think that'll change anytime soon
> How could we jump back into the cozy past again?
"You can't go back home to your family, back home to your childhood ... back home to a young man's dreams of glory and of fame ... back home to places in the country, back home to the old forms and systems of things which once seemed everlasting, but which are changing all the time - back home to the escapes of Time and Memory."
Well sometimes you can
“No man ever steps in the same river twice, for it is not the same river and he is not the same man.”
I'm seeing this nostalgic sentiment a lot on forums these days, but it's not just about bringing back ye old usenet groups: you also need to bring back the way people thought back then. It's hard.
The sentiment I'm seeing is that, to avoid the rose-tinted bias of nostalgia, we default to believing there is nothing the past has to teach us, and everything is just better today.
Eternal September: AI edition
"Democratization of access has its victims."
If it were democratization, we'd all have access to the data that underlies this dataset. In reality, this will destroy the origins of the data, and it will become hidden for all of humanity.
It is a tragedy of the commons of the highest order.
They will talk about how they are trying to protect humanity all while putting it behind a paywall.
> How could we jump back into the cozy past again?
Delete 9/10th of the HN population. I could accept even AI discussion if there wasn't so much of it, and so many posts in general; /new is like trying to win the lottery.
It doesn't matter how interesting it is, you just have to be lucky. The corollary is that the front page isn't the most interesting stuff posted today, it's a list of who's got lucky posting the average topic for the average reader.
Is that really the problem, though, or has the audience simply changed? It seems to me that the thoughtful deep dives of years past are still being produced, but they're buried under a much larger volume of lighter content. More importantly, many of the people working at the forefront of technology and product development appear to have moved to private communities such as Discord, Slack, and similar groups, perhaps in part because of NDAs.
Usenet is still here. I'm using it on a regular basis. The point is very few others are.
HN really needs an AI filter. Most posts are AI company promotion now.
All HN users can be part of the solution. Help mods by emailing them at hn@ycombinator.com. I recently pointed out a suspicious pattern, and the mods were able to use that to track and ban a collection of accounts.
Some users on HN are "allowed" to do breathless AI self-promotion on HN and other readers support them by posting it anyway. Other HNers on the other hand, got banned immediately for doing the same thing.
That is why many users here realized that there are protected accounts that can do that and everyone else cannot. Even in rare circumstances, the moderators sometimes unbeknownst to them, may also be indirectly allowing it.
It is far more complicated and one of the reasons why users here are fed up of the constant AI deluge on this orange site and nothing is changing other than there is more of it.
I’ve had some luck messaging the moderators and calling them out for it. They may or may not do it on purpose, I have no idea. But it has worked to get certain golden children at least flagged when they’re doing things that would get others straight up banned.
All the god damn pelican spam
I wrote a browser extension to auto-hide AI-related articles. It makes HN much more readable: https://github.com/ColdPie1/enough-ai-on-hn/
If you're interested in a solution, I maintain a reader that performs client-side filtering from a given view. Here's a view with an AI filter I try to keep up to date: https://logicnow.io/hn/index.html?view=https://raw.githubuse...
It's not just HN, have you seen mainstream news recently? It's just a hot topic and I think it's normal for people to talk about those a lot.
If quantum computing ever goes mainstream, I'm sure you'll see a lot of articles about that as well.
The programming Reddit forbade AI posts. Unfortunately the result is NOT better level posts, but a wasteland instead.
Reddit had become a wasteland before the LLM era began.
Forbidding AI posts is a net positive, not negative.
Reading about tech has become gradually less and less fun for me over the past 16 years.
Does anyone remember the awe that was felt the first time we encountered an iPhone. Or an iPad? I remember those days I would read tech news and sites like Ars Technica and comments were sometimes snarky and indignant but generally people were enthusiastic about reading tech because it offered glimpses into more technically advanced futures.
I wasn't browsing HN back then but I imaging the same thing was true of HN in the late 2000's and early 2010's.
Now tech sites like Ars Technica consistently have comment sections filled with immense negativity cynicism indigation and just plain old boiling rage. I avoid those comment sections now because reading such things negatively affects my mood and certainly isn't fun.
And over the past few years HN has gradually become more negative and nasty.
All this sucks. This isn't why I became interested in tech. This isn't why I used to spend time reading about technology and science.
It kinda started turning as people realized that a huge amount of the biggest people pushing tech are rabidly anti-human, incredibly racist and usually spouting nazi quotes every chance they get, and are out to exploit as many people as possible just because they can.
The negativity is rather well-earned at this point. I do still enjoy tech, but the ecosystem is no longer as infused (by ratio) with nerds doing cool things.
All this sucks. This isn't why I became interested in tech. This isn't why I used to spend time reading about technology and science.
You're 16 years older now. You will have changed. HN has too, but looking back at how great it all was without also realising that your point of view is different means you're missing a big part of why it feels worse now. I've been here a long time too so I'm not going to claim I don't see the same thing, but I imagine someone who signed up 3 months ago won't have the same opinion as us old timers.
> but I imagine someone who signed up 3 months ago won't have the same opinion as us old timers.
that's for sure but if I got to know HN 3 months ago I would not be an active user. The content and discussions here are still interesting but the demographics that enjoy it changed over time and I (and OP) might not be the kind of people that would get interested anymore.
There is no "hacker" in HN anymore, unless we count the hustlers trying to cheat their way to success. Talking about that, there are so much psyops going on in so many threads that we, the old hackers, are the ones getting hacked.
I still maintain the notion that tech was most exciting/peaked around 2010. Macs were awesome (and the Intel transition was complete at full strength), things were fast, OSes weren't bloated, and we didn't have flavor-of-the-month fads like cloud, crypto, NFTs, or AI.
I can't recall who said it, but "The biggest problem with golden ages is that you never realize you're in one until it's over".
There were several golden ages of tech for most people who went through them: - the pre-internet run-up of personal computing (~1975-1995) - the early internet (~ 1990-1998) - the big post-dot com crash tech boom (~ 2003-2020, but more of a golden age for employees)
However, almost all golden ages (including non-tech ones) end for the same reason (the ranges above I'm excluding the last phase). Early adopters (tech enthusiasts), early implementations that focus on them, then mass market appeal that "dumbs it down", then the financialization as accountants and grifters start showing up. Same thing happened to Hollywood, the music industry, etc.
We're now squarely in the last phase of the post-dotcom boom.
(I'll also note that the timeframes are arguable - I'm not going to get into quibbles over minutiae)
>Does anyone remember the awe that was felt the first time we encountered an iPhone. Or an iPad?
Yes, and when I first saw LLMs and image generators I experienced a similar feeling of awe. We haven't had anything like this in decades.
It is unfortunate that many people are negative and nasty about it. Local politicians are getting shot at over datacenter projects. We have truly lost the plot.
I used to think that people had learned their lesson that automation was a force for good, and we had gotten past the Luddite days of smashing looms. But it seems that is not the case, people are just as short-sighted as ever.
Ignorance. Ignorance of the facts on the ground and the politics of the moment.
Do you honestly believe that the ledger for automation has zero items in the "liabilities" column? Do you think that 100% of people whose job is displaced find another job in a reasonable amount of time and on the same order or pay / status?
Do you believe that neighbors of recently built data centers have nothing to complain about?
The truth is that "data center politics" is happening in the context of post-NAFTA offshoring of lots of jobs/industry, post-manufacturing migration to the South, post-hollowing of labor unions, and in the context of Trump's War on Immigration (where AI companies have likened their product to "infinite AI immigrants").
Even the White House's Ratepayer Protection Pledge (a laudable goal) event was undermined by POTUS's speech at the event. Instead of even pretending to listen to the people who are hesitant about first and second order effects of data center buildouts, he encouraged developers to push the projects through ASAP and ignore all of the naysayers. It's clear this entire project is to maximize commercial advancement, not benefit people or give voice to voters/constituents.
And in a moment where we are told we have to do this ASAP because "we are in a race with China" but nobody can explain the race. It's not for AGI because China doesn't expect it to yield AGI. Perhaps Zuckerberg does, but he was wrong about crypto, then Metaverse, then open weights models training (before he returned to open weights). The US seems to be using LLMs wrong as a glorified DEI detector (DOGE using LLMs to enforce their political homodoxy) and to identify schools in Iran to bomb + double tap.
China has reasonable regulations on LLMs to protect their society whereas our "regulations" seem to be the Amazon CEO texting David Sachs on a Friday evening to warn about cybersecurity in a misunderstanding between what Mythos (prerelease for select privileged users) and Fable (general release with lots of moderation controls) and the nature of the cybersecurity vulnerabilities found (they were able to be replicated with other non-Claude models).
We are being outwitted by ourselves. The problem isn't the people anxious about data centers or AI. It's the clueless leaders who don't bother to anticipate the obvious outcomes of their decisions and execute a plan to mitigate it. And you desire "feeling awe" so much that you don't want to recognize that politics means not everybody is going to succumb to the same feeling of awe, and that's okay.
>Do you think that 100% of people whose job is displaced find another job in a reasonable amount of time and on the same order or pay / status?
Do you not realize that the only reason we are not impoverished dirt farmers right now is because of automation?
Automation is what creates wealth. None of these 'high pay/status' jobs would exist if it were not for automating the previous generation of jobs. Every time we automate, we invent new higher-paying jobs and increase average wages across the economy.
This has played out time and time again since the industrial revolution. Luddites like yourself try to block automation with 'what if there no jobs?? what if it only makes the rich richer and we all starve??' Then automation happens anyway, we create new jobs that pay more because the pie got bigger, and we all end up better off.
> what if it only makes the rich richer and we all starve??' Then automation happens anyway
Revolutions also happen anyway. Not sure if automated security would be enough to prevent them.
I'm curious if anybody has tried to figure out roughly the % of bot comments? I keep seeing the "ai-talking cliches", but wonder if I'm crazy because people (or... other bots???) take it at face value.
I've been noticing in recent months that even when the article is in no way related to AI, there's often comments trying to steer it into AI, often in a way that makes no sense at all. It has a way of becoming its own parody.
there was one glorious day years ago when every story on the hn front page was about erlang. that was fun. (anyone else remember that?)
The Erlang day was on 2009-03-11. You can still see how the front page looked that day:
https://news.ycombinator.com/front?day=2009-03-11
I'm deeply disappointed that many of the stories on HN are directly or indirectly about A.I. and LLMs. To me this is a complete turn-off. OTOH this seems to be the case everywhere in the IT landscape these days.
Many people interested in technical stories have mostly stopped reading computing related news. I know people that used to host talks about .NET software development quit because they hated all .NET developments pushed by Microsoft were related to Azure or A.I. These are developers that want to build software, not scream to/plead with an LLM to do it for them. It simply takes the fun out of being a software developer. It wouldn't surprise me in the least that in a couple of years many people will leave the sector because it's no fun anymore. Management is forcing you to use your (supposed) future replacement.
I think grief is appropriate for the part of the software engineering craft that has been lost, because it was genuinely fun and rewarding at times, and I don't think it'll ever come back. Sure, some of us will hack on things manually for fun here and there, but even that will likely become ever more niche, like doing your laundry with a washboard. And professionally, no one is going to be writing code without LLMs before long. That world is gone, and it's OK to mourn it.
If using a washing machine required you to send your laundry to an evil corporation, the washboard would start to get some appeal…
Yeah, I definitely don’t think all software engineers will be sending their codebase to OpenAI or Anthropic, but I think basically all of us will be using LLMs. I keep hearing about how open weight models are only six months behind, so presumably in a few years, it’ll be feasible for corporations to self-host very capable models. Still, I suspect most will just use the dominant labs, or something like Bedrock where you have stronger ZDR guarantees.
Not true. Many developers will keep writing their own code, possibly outside the corporations that force them to use LLMs to do so.
A few, yes. Just like there are woodworkers who only use hand tools out of principle. But most engineers will use LLMs on their side projects, and virtually all will for professional work.
The post tracks two rather different things. AI is almost certainly the hottest topic in tech right now, just like mobile was hot after the iPhone launched or cloud was a decade ago. AI-generated text is a separate complaint, and it's the only one I think is controversial. Slop, no thanks. A human using AI tools to write better is fine by me, even if a tell or two slips through.
Counterpoint: what we call AI is really just more sophisticated statistical algorithms, and so all of the stories about using AI, making AI more efficient, and all of the vibe coded apps are really just the world coming to terms with a new computing paradigm. It’s what you’d expect on a computing-heavy forum.
Eventually it will stop being novel, at which point people will stop calling it “AI”: https://en.wikipedia.org/wiki/AI_effect
Reading these comments it seems that meta-discussion of AI is the most hot button topic of all in HN.
Not everyone here is AI. I can't tell the difference so what does it matter? If I learn something or get inspired by a person or a bot, great.
AI tools are incredibly powerful, new and rapidly evolving, you shouldn't be surprised that everyone on HN is all over it.
> I can't tell the difference so what does it matter?
Eternal September. Just like the usenet groups
Eternal September. Just like the usenet groups
And CB radio before that.
> Not everyone here is AI.
That’s not the argument.
> I can't tell the difference so what does it matter?
It matters because you’re not the only human on HN and others have difference preferences. “I don’t care about this particular issue” is not the same thing as “this particular issue isn’t important”.
> you shouldn't be surprised that everyone on HN is all over it.
Not the argument either. No one is “surprised” this happens, people (obviously not all, but a significant number) are annoyed by it.
If people can't tell the difference but still get the shits because of some notion of being cheated or putting humanity on the pedestal, they're just being tribalistic. It's a common human rationalisation that helps entrench social hierarchies but in this case, and it's probably not the right word but Speciesism.
It's about as strong an argument as going to the paint shop, loving a colour, painting your house, then your neighbour informing you that a person of colour or an algorithm created it, then you hating it.
We were something different before we were primates, and we'll be something different after. Evolution has always been in a flux.
We're all just entities processing information and in my opinion, humanity should collectively get off the high horse.
Serious question, could you ever have a meaningful relationship with a teddy bear and/or an algorithm?
> If people can't tell the difference (…)
Not “people”, you. Some can tell the difference, others can’t. Just because you cannot tell and don’t mind that the waiter pissed in your soup, it doesn’t mean other people can’t tell and shouldn’t be bothered by it.
> It's about as strong an argument as going to the paint shop (…)
No, no it is not. That analogy makes no sense. Colour is part of the light spectrum, it’s not “created”. Paint is, and that’s a physical process.
> We were something different before we were primates (…)
That’s not how evolution works. We don’t just evolve into different things over time. Mutations occur and the most adapted survive while the others die and are eventually replaced. We are short circuiting that process for ourselves and other species via selective breeding.
> (…) humanity should collectively get off the high horse.
I agree, and that has nothing to do with the conversation. Humanity’s “high horse” starts way before the conversation ever reaches computers”, let alone LLMs.
> Serious question (…)
In don’t think it is. You have to define what you mean by “meaningful relationship” (no two people will have the same definition) and “you” (are you being specific or general)?
We could sit here all day arguing about subjectivity.
It's not been my experience that people can reliably tell the difference.
I, personally, would not equate pissing in somebody's soup with having a chat with an LLM but if we must, you need to appreciate that it's the piss itself, that makes the soup delicious, if the soup was not pissed in, it would be middling at best. Only in that context can you then ask yourself, "Am I bothered by this delicious soup?"
I think the analogy works in the context in which it was intended. I was illustrating the absurdity of contempt for something based on its provenance, after praising it. If you really don't like the paint analogy image this, you read a book, loved it, then later found out the author was black, or an LLM, then you hated the book.
I'm not sure I understand where you're coming from regarding evolution. Sexual reproduction is a comparatively recent evolution in the scheme of things. That which came before us defined us, are our ancestors and is our history. What comes tomorrow, engineered or otherwise I would consider our children and should be loved, cherished and allowed to thrive in the same way our ancestors ushered us all into this universe.
I wasn't being facetious when I asked if you could have a meaningful relationship with an algorithm, I asked about the teddy bear to draw a parallel to which a lot of people are familiar. If a child can have an enriched life and what I would call meaningful, intimate relationship with an stuffed toy, why shouldn't they also be able to have the same with an algorithm that talks back to them? I'm being both general and specific if I think about it.
If I could image your brain perfectly and put it into a perfect physics simulation running on silicon or whatever substrate, then have this same conversation with you, I would call that meaningful. Whether or not you would at that point label yourself as living, conscious, sentient, human, whatever. It would be a meaningful relationship to me.
I think people are scared of change, and that's alright.
If people feel this way they should just start a flag this AI post campaign. If people can flag anything they disagree with a little effort the community can be AI free at least on main page.
That said it would be interesting to have a AI flagging so you can just filter the just AI topics. Its very different from people harping on crypto or VC focus. I dislike reading the AI output when someone says they have an opinion the original source material the AI ripped content from to me is more valuable flaws in all than a normalized AI output
>If people can flag anything they disagree with a little effort the community can be AI free at least on main page
Unfortunately AI-based articles are on topic (AI generated comments are not) and if people flag them they'll just lose their flagging rights. I'm also reasonably certain "flagging campaigns" are categorically against the rules, even though there is such a campaign implicitly against "politics."
There's nothing anyone can do about it except complain (which will eventually get them filtered because such complaints are also against the rules), be quiet and tolerate it (which is the only option that follows the guidelines), or else leave to greener pastures.
I run https://www.unslop.news/ which uses an LLM (ironically) to filter out content discussing AI. There's text at the bottom to indicate what fraction of the first 3 pages gets filtered out, and it's pretty much always at 50% or more.
Then, why are you in this thread?
I found it bc I started getting new traffic to the site from HN and turns out someone else was actually plugging it in another comment here
Thank you so much for this! Generally I like to read AI news, yet it's important to know what else is out there! I hope that ycombinator is implementing this filter by themselves, one day.
Upvoted obviously just to counterbalance how often I got downvoted for stating how popular posts from OpenAI, Anthropic, Google, etc are just ads.
The author is most subtle or diplomatic than me and thus call them "AI vendor press releases" or "AI vendor marketing" but we know.
I have a comment here from a few years ago where I am accused of it being generated by AI. I wear it like a badge of honor because I wrote it 100% by hand.
A tech site would seem to have "stories about AI".
That's an odd thing to conflate with " written by AI".
I've posted several stories "about AI".
I disagree, the most important source of tech discussion on the internet right now is /g/ board on 4chan.
Has been for quite some time.
100% of the comments that I disagree with are AI-generated bots and paid shills. All of mine are human-generated.
"At the same time, it has a fair number of toxic commenters, making it a dependable source of insults hurled in my general direction; if you want a taste, {another article by same author} has been called “watered-down” and “slop”."
We have gotten to the point where it is virtually guaranteed that some commenter to almost any given article will have an oversensitive AI detector and accuse the author of it being AI. I swear there's people whose AI detectors are just
Occasionally such people also complain about how much of the world is AI now... well... yeah, that's what happens when that's your detector. It's not quite that bad, as bad as it is.
I don't know what to do about it. Discussion about AI provenance of a given work is often a valid topic. When YouTube shovels an AI channel at me and it briefly penetrates my defenses, it's heartening to see the commentors trying to warn people about it, shortly before I block the channel.
(It's annoying that so many legit channels, which I know are legit because they predate AI and the same person is on camera as before, have switched to heavily-AI thumbnails. That's the hole that the other channels sometimes get me with, because just having an AI thumbnail doesn't prove it's not a legit channel now.)
On the other hand, the chorus of very annoyed people who are very annoyed because they see AI everywhere who see AI everywhere because they literally see AI everywhere are becoming a bit of a nuisance themselves.
> We have gotten to the point where it is virtually guaranteed that some commenter to almost any given article will have an oversensitive AI detector and accuse the author of it being AI.
Well... when the utter majority of articles written about topics surrounding AI are written with AI and a ton of newcomers to the reverse-engineering and even hardware hacking communities now have AI do their work for them, even the last remnants of areas that used to be "for humans only" are now contaminated with AI.
Of course people end up being very sensitive. I come to HN to read words written by humans, not fluff hallucinated by an LLM.
The truth is, writing is hard and time consuming, and most people are uncomfortable with the process. AI is a very easy shortcut to lucid, logical, authoritative text. Like sugary drinks and snacks, people can't resist. Add in the pressure to keep up with everyone else and the choice is easy -- spend a few days drafting and editing, or a few minutes chatting with an LLM before hitting post?
AI has rapidly destroyed, completely destroyed, the remaining authenticity on the internet. The majority of articles are AI generated. And yeah at this point its futile to keep pointing it out. Human prose has been eliminated.
Also like sugary drinks, which our brains perceive as satisfying but are actually the opposite, LLM’s have been well-trained to produce text our brains (and up-voting fingers) perceive as lucid, logical, and authoritative despite the frequent absence of those attributes.
Heh. Indeed.
Compare to the 2010s when 90% of the front page was about NoSQL
90% of the commentators are AI/bots these days.
And 91% of all statistics on social media are fabricated out of thin air.
If that were truly a problem, why wouldn't a human-only social media platform take off like wildfire?
Bad bot.
I see such comments but who? How can you tell?
This is a new question that is well worth sitting with!
The shape of the current Hacker News experience certainly implies an abundance of AI generated content.
A quiet approach to testing this insight is to ask the HN community itself, then assess the resulting replies for the load-bearing surface features.
But at the end of the day, aren't we all waiting to unpack our throughlines?
It's like blasting white noise at max volume. I wonder if all communication will eventually be rendered into a homogenous thrumming hiss.
HN is a reflection of the industry and you'd have to be living under and rock (or in denial) to ignore LLMs. People who complain about it give off a real "kids these days"-vibe that has been present on HN for the entire time I've been here. There are _always_ people complaining about the current infatuation (and I'm sure I've been one of them).
I find the posts/comments bemoaning the pervasiveness of LLMs on HN to be more tiring than even some of the more annoying LLM posts.
If HN didn't feature LLM-related content there would be people, rightful so, complaining that we are ignoring a huge thing happening in the industry.
I think its very easy to note and come to the conclusion that this is the most disruptive tech of our industry. HN is full of people from all walks of life, but people from all walks of life are using AI (or LLMs). The way I hear people describe using AI is the way people used to talk about Google.
As much as I am one of the "In Betweeners" of AI (not hyper loving, nor hyper hating it) I do wish HN wasn't flooded with it, but its just the nature of LLMs currently, they are the new thing that has taken over our society in a drastic way, those who pretend otherwise will be surprised to know that the rest of society has moved forward with it.
Would be interested to see this compared to past trends, crypto and NFTs for example
HN is about to witness the exodus/migration we've seen in prior years, except this time it will be a much larger movement rather than focused/specific top contributors who just said "fuck it" and had enough with this shithole.
See also: jwz's referral redirect.
I like this but next do how much of HN is astroturfing for Chinese AI.
Why would that be caused by astroturfing? The HN audience is the exact people I would expect to be looking at alternative to mainstream AI providers
"for example, around 2018, it had a fair number of stories focused on cryptocurrencies and NFTs"
Bro, NFTs and crypto nolonger being front page is proof of a decline since 2018?
This story appears to be written by a human, and is watered down human-slop.
The thing is, the AI-related topics are pretty easy to skip. Sure, you may have to go to the second page to see the same number of non-AI stories that you used to see on the front page alone, but you can just not click on them. (Same with politics, BTW.)
But AI-generated content is harder, both stories and comments. HN does moderately (heh) well at filtering them out, but I would like better.
Someone made an HN without AI filter and it front paged here a bit ago.
https://elijahpotter.dev/hnsansai
(That's the first one I found, maybe there were others)
So...are you complainers all using it? If not, why not?
Spoiler -- because when someone says "I'm tired of reading about {X}", they're really saying "I want everyone else to stop talking about {X}". This has happened repeatedly over HN's life, and it just betrays how emotional people get when they feel that they can't control the narrative.
If HN doesn't work for you, stop coming to it?
A lot of us moved over to lobste.rs and it's been refreshing, to say the least.
It’s sad; on the one hand, it’s true, and on the other, this story is also about AI. (translated from german by AI)
All of it, it's time to move away, go to https://outspeaker.com/, it doesn't censor nor remove anti-ai content
Is Pangram's detection capability considered to be on the higher side? This post seems to be written under the assumption that Pangram's detection is highly effective. Is it actually considered to be on the higher side in English-speaking circles?
My confidence in Pangram is pretty low -- seems like a lot of false positives *and* negatives
Yes, from my experience it is highly effective
Pangram links several third party evaluations [1] that all seem to agree with 90+% detection, and very close to 0% false positives. Obviously those could be cherry picked, but I’ve yet to see arguments to the contrary that actually include any supporting data. (E.g here’s a prompt that will get Claude to spit out text that pangram doesn’t detect or here’s an article authored in 2018 that pangram says is AI.) Detection is an arms race, so this could change (though I’d expect in the direction of false negatives), but right now it seems like defense is winning.
[1] https://www.pangram.com/blog/third-party-pangram-evals
> Obviously those could be cherry picked…
The "could" isn't required here, since Pangram calls many of the evals "collaborations", and some received direct support. Results that don't favor Pangram as being among the least-worst of a questionable product category* surely wouldn't be featured.
* https://mitsloanedtech.mit.edu/ai/teach/ai-detectors-dont-wo...
The most recent reference in that article appears to be from 2023.
Would be cool to see data for an older timespan, starting pre-LLM
And what will it be in another 10 years?
Too much. And most of the posts are slop or low quality syncophantic drivel.
Sometimes there's an obvious AI comment, how is it auto-hidden/downvoted?
There is some LLM detection for comments that auto-flags them.
It has some unfortunate side effects, though. Namely on a thread where people are discussing a text-heavy post that is LLM-authored, quoting a large enough block of that content appears to trigger it sometimes. Makes sense that it does, but..
Contrary to popular belief, it's quite easy to spot something written with AI. You just need the skill/gift. I usually spot them rather quickly and will flag on sight. Other users will do the same.
holy shit I've been here since 2013, I need to get a job
Surely you're aware of the various theories that AI is going to up-end the entire world.
How can you be surprised it's up-ending a geek forum first?
Nobody is surprised
The author of the article seems to be surprised.
As well as many people in these comments.
Nothing in the article is expressing surprise
"That said, the recent shift feels more profound: almost every day, it feels that the lineup is dominated by stories focused on AI, written by AI, or commented on by AI."
I interpreted that as surprise. I apologize if you disagree.
I'd just like to say that as an old nerd who has been reading hacker news, often daily, for a long long time, hacker news is an absolute treasure, and dang & pg are saints for keeping it going.
The Magnasanti that we call "Tech" is a little swimming pool that we first leapt into as kids, trying to splash the befuddled saps standing outside the pool. As we splashed and dunked and sprayed each other, a thrilling, unceasing mania began to build within us, and so we splashed all the harder, forcing enormous waves that slapped against the world, shocking it into noticing us. As more jumped in (or were pushed) we demanded and they forced upon us more and more and more water, until we no longer had to entice or push, the waves swept through the whole world, eating at the foundations of entire societies, once solid land crumbling into ravenous waters and spilling their hapless citizens into the vortex, spinning them head over heels, forcing them to cling to the shattered debris of their lives, and try to just-stay-above-water frantically with the rest of us, the billions and billions bobbing in the froth.
And how goddamn nice is it.... that in that raging omniflux, maybe not quite at the center, but maybe also not far from it, is a little sheltered swim up bar, its orange sign still cheerily showing the Y of Y Combinator, which used to BE one of the first kids, one of the ones that pushed and pulled and flooded and floated. They still do, of course.
Perhaps the most amazing thing about this little bar, is that in all this swirling, foaming, madness, it is an oasis of stability. Absolute stability: This is from the first snapshot on Feb 21st 2007: https://web.archive.org/web/20070221033032/http://news.ycomb.... Look familiar? See pg's post getting knocked off the top already? Shelley might have approved.
Sometimes you get to chat with the people forcing the great devouring waves that reshape the earth, sometimes you just chat about what's new in JS tooling. If you are particularly lucky that day, you will find yourself enraptured by some hacker with the mad gleam of triumph in their eye, recounting in excruciating detail their monomaniacal pursuit of some utterly pointless pet project, and you will not consider one second of it wasted. Not even on your deathbed. Not even when compared to spending more time with your beloved.
I think it would be difficult to find an example of a community that has been stewarded as well, for as long, as hacker news, and while you do get commenters of all ages, interests, walks of life, politeness levels, and hyper-fixations upon irrelevant minutia, I've often found that for the most part you are all a lovely bunch of humans and/or AI bots.
Lest you accuse me of labiogluteal angling (a classic!), I am too old and too wet and too cold and too tired to want or need anything from anyone here, beyond a pleasant place to read.
But... I thought that just once over these decades, I'd tell you this how very grateful I am to be able to swim up to this bar, sit on the familiar stools, take a moment of respite from the tempest, and read what any of you might have to share, even -- or perhaps if I am being honest, especially -- when I think I think its absolute bollocks.
Thank you all, and may Tech continue to spin you, right round, right round, like a...
I swear that the amount of astroturfing in the comment field by AI bots being positive towards AI is at an all time high as well.
For me, I'm surprised to see people proposing conformist views and trying to push whatever the establishment is pushing.
I dunno, I was a contrarian with blockchain and now am conformist with AI. Being contrarian for its own sake is pretty stupid.
>Being contrarian for its own sake is pretty stupid.
All progress depends on the unreasonable man.
Got to say I agree with this. Contrarians can be an annoyance, but someone's got to point out the Emperor's got no clothes on.
FWIW I’ve seen people complaining of the exact opposite. Difficult to say without measuring
This makes no sense to me - what's the opposite of AI astroturfing? Why are people complaining about it?
Ironically, your comment could easily be a failed comment from an astroturfing LLM tasked with criticising/disagreeing with people calling out LLMs.
FWIW, I checked your profile, I don't really think you're an LLM. The internet is hard now - sorry :(
I’m indeed a fleshy human :)
What I meant is that I’ve seen people complaining that the negativity towards AI on HN is manufactured and astroturfed. So we have people complaining on both ends of the debate.
I would bet that the AI booster side is way more astroturfed, but I’m very likely biased (I see genAI as a horrible mistake for human societies)
> (I see genAI as a horrible mistake for human societies)
GenAI is amazing, how we use it is so far a horrible mistake.
It’s technically very interesting and fascinating, but the incentives are horrible
Links?
I aint got none.
That's fine, but it would be helpful to see examples. People often make generic claims about astroturfing, bots, etc. that don't hold up in practice when we look at the data.
Bots and genai comments also exist, of course, and are not allowed here (https://news.ycombinator.com/newsguidelines.html#generated).
Does HN do proactive detection and/or flagging of comments determined to be AI generated?
I understand if information publicized on how this process might work is intentionally restricted, so as not to be easily circumvented. I'm also not asking this in a pointed or judgmental way with respect to HN, I'm mostly just curious how various social platforms are approaching this problem.
Yes, to the extent that mods will admonish what appear to be AI-generated comments:
<https://hn.algolia.com/?dateRange=pastYear&page=0&prefix=tru...>
Articles are a fuzzier area AFAIU:
<https://news.ycombinator.com/item?id=48887149>
(Not a mod myself, though I watch mod actions fairly closely.)
> People often make generic claims about astroturfing, bots, etc. that don't hold up in practice when we look at the data.
I'd be very careful with believing the system that looks for bots and astroturfing works well enough to confidently say that. Some of the people, corporations and institutions interested in manipulating HN and its audience are quite capable.
Of course it's possible that there's a SSM (Sufficiently Smart Manipulator), which by definition we can't know about. But despite not everything being knowable, there are still things one can say.
This topic has a long history on HN. Here's one long post I wrote a number of years ago: https://news.ycombinator.com/item?id=27398725. Others here: https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que....
Here's an example of what we can say: if user A accuses user B of astroturfing, spying, manipulating, etc., and a quick look at B's posting history shows that they were posting to HN about, let's say, Erlang in 2019, the odds of that accusation being correct go down to negligible. It's possible, of course, that B was already working for "people, corporations and institutions interested in manipulating HN" 7 years ago. It's also possible that malefactors came along later and found a way to pay or coerce B into doing their dirty work. But at this point one becomes unmoored from any evidence at all. The far more likely explanation is that A is slinging cheap accusations because they disliked something B said and are resorting to the internet's favorite trope ("you're a shill!").
When I say "People often make generic claims about astroturfing, bots, etc. that don't hold up in practice when we look at the data", what I mean is that the vast majority of such claims boil down to cases like the one I just described or something similar.
Sure, that makes sense. Thanks for replying.
> Here's an example of what we can say: if user A accuses user B of astroturfing, spying, manipulating, etc., and a quick look at B's posting history shows that they were posting to HN about, let's say, Erlang in 2019, the odds of that accusation being correct go down to negligible
about this one tho I have noticed bot behavior (including posts that are incoherent) from old accounts - so there is the risk of 2 things going on:
- people getting hacked and their accounts being used for that
- some people that just decided to connect a bot to their account or use their account themselves to astroturf.
which is way more difficult to spot. Anyway, I don't have a solution for those problems. You have an interesting and difficult work :)) Have a nice day, dang!
I'm not sure I'd be able to (confidently) tell the difference between human and SOTA bot comments.
Is there something trick to it?
AI has legitimately changed everything. It's not 2022 anymore. Try a recent LLM and set it to a coding task. For even moderate tasks, things you could burn a weekend or more on, the model will complete it in minutes, to a high degree of quality. It can write code better than most programmers, faster than any human. If that doesn't inspire positive feelings, it should at least fill you with awe. We live in an age of wonders.
>If that doesn't inspire positive feelings, it should at least fill you with awe. We live in an age of wonders.
It's impressive, despite its ethical baggage and psychological, environmental and psychological harm. But if if filed you in awe I think you need to get away from the keyboard. Things are happening in biology, astronomy, archaeology, engineering, that are far more awesome than software that can write code for you.
And if your only response to that is to be in awe of whatever part AI may have had in the process, you're still missing the point.
Did you know that not everyone who disagrees with you is a bot? Some people are genuinely excited about AI.
As someone who has been posting here since 2010, I don't blame AI. The exodus started long ago and continues to evolve over time. It is unfortunate what this place has become due to poor quality of moderation. Lobste.rs just hits better, and Reddit isn't half-bad as purported.
Did not read the article, but I would say about 40%. Btw, you can toggle [showdead] in profile settings, then you'll see messages caught by moderators, which are sophisticated bots (writing style at ChatGPT3 level) + absolutely valid points that lash out on ruling class, SV mentality, US and Israel.
a lot is also not caught.. and over time itll be less and less. hn has been great for a very long time, but this time has come to an end
Did read the article, you'd be correct... for a few months ago. Now it's apparently up to 60%.
> you'll see messages caught by moderators
Much of these are caught by users downvoting and/or flagging. I think the lack of a public moderation log makes people think the mods are constantly running around every story/comment to steer the site when really it's the users doing most of the grunt work while the moderators only step in occasionally for the more serious items.
Far less than the constant whining about AI.