It was embarrassing a year+ ago when much of this technology barely worked at all.
If you did not perceive a step change in model capabilities in late 2025, it's possible you are working on problems that are just too hard or not well aligned (e.g., arts and crafts). If you are working on software and need to move the occasional mountain, it's really hard to argue with the capabilities of contemporary models and harnesses.
This space is evolving rapidly. Faster in some areas than others, but you should still keep an open mind and adapt to the new realities as they emerge. Wishing for a thing to fail can blind you to more opportunities than you would probably be willing to reject if they were presented directly.
We have access to and use the same tech, the difference is elsewhere.
> It was embarrassing a year+ ago when much of this technology barely worked at all.
The things some people say, how they talk about "outraged AI critics" what they would prefer those to be, would be fucking stupid even in first grade.
These, and things like "AI-generated news sites to attack critics" really speak for themselves. There is something worth really calling out here, and it fell out naturally.
Nov'25-Dec-25 something changed. I believe that was the time when Opus 4.5 was launched. (we went from function auto-complete to module auto-complete) People came from the vacations, and were like, make AI do your work. One could actually start offloading work to LLMs.
A year+ ago, LLMs worked nearly as good as they work today.
I say this as someone that uses LLMs on a daily basis. It is a moderately useful tool for software development, but it was already useful in January 2025 (to pick a random date). Check what models were available by then. There were improvements since them, but they were mostly marginal.
The thing is, ever since ChatGPT 4 days there are people claiming on a daily basis exactly what you are claiming now, and it feels pretty silly. Sometimes it's hard to not imagine the accounts breathlessly claiming 100x productivity as bots.
In other words, I presume that if you are 100x more productive now, you were a 1/100x developer before LLMs. I actually question the capabilities of anyone making claims like that.
The "100x productive" thing is some kind of bogeyman. It's pretty clear these things are useful and helpful and the degree to which they are such continually improves, but putting some specific large number on it is pointless.
If you didn't notice a difference between AI assisted coding in Jan 2025 and now beyond "marginal" improvements... well to be frank I don't think you've really been paying very close attention. It's a combination of the way we interact with the models and the models themselves, and this really did get much more effective in late 2025.
You can say the model improvements in that year were incremental, but the degree to which the overall capabilities have changed and what that means for software development is more than the sum of those parts.
> The "100x productive" thing is some kind of bogeyman.
> If you didn't notice a difference between AI assisted coding in Jan 2025 and now beyond "marginal" improvements... well to be frank I don't think you've really been paying very close attention.
Those are the same claims, just phrased differently. If the first is a bogeyman, so is the second.
And of course, if someone disagrees that the improvements since 2025 were not dramatic, they can only possibly not be paying attention. Except we have been seeing the same claims on a daily basis for years.
LLMs are clearly world changing technology. I'm no fan of the big corporations but I'm very excited to see the models themselves develop. I'm looking forward to the day I can self host these things.
You can? It’s pretty remarkable what you can do with Qwen 3.6 on decent hardware, and Gemma 12b on nearly anything. If you can’t own it, take something for a spin on vast.ai where you can get all kinds of machines on tap for ~$0.20/hour+
llama.cpp has gotten so good that you can really get optimized performance with a few flags or none at all, and serve any model on-demand. Spend $45 to have your own GLM 5.2 machine for an hour. If you can’t manage full utilization you’d get 33.2M tokens, retailing for $128.
It takes some smarts to use dumb models, but renting time for frontier tokens is easy math. Vast also lets you run on spot which is great for training or agents with harnesses that can accommodate uncertain model availability.
It's the metrics though. I've noticed that AI topics get hundreds of votes, that means either someone is trying to beat the algorithm to stay in the top news, or suddenly this place is filled with thousands of AI enthusiasts who ignore all the rest, non-AI topics. (Thousands because not everyone will upvote, so if only a fraction upvotes hundreds of upvotes should translate to thousands.)
This kind of conspiratorial thinking is ridiculous.
My life is affected for better or worse because of AI - I’m forced to use it at my day job as an R&D. It is perfectly rational and expected and preferable for AI to be the front news in a technology forum.
You probably don’t like AI but that is completely different from being affected by it.
I was talking about the metrics on HN you are talking about how AI improved you life (or didn't). Two different things in my opinion. I'm using AI myself btw. This isn't the point.
Your life is changed by it. You expect to not be shown posts on a technology forum on a technology that actually changed your life. How is this coherent?
I honestly think even the hypebros are less embarrassing than the full-on Luddites that think LLMs are an ontological evil.
It should really not surprise you that Hacker News, a community with an extremely technical audience, is interested in a bleeding edge technical tool that is literally completely reshaping the industry.
Hypebros and full-on Luddites that think LLMs are an ontological evil believe exactly the same things about LLMs.
The primary difference is that hypebros brag about incoming promised apocalypse, 40% unemployment, economy ruled by unaccountable monopolies and permanent underclass. Luddites think that such development would be a bad thing.
I dont think apocalypse will happen, but I do think AI companies are massive assholes who deserves all the pushback they can get for the way they doom hype themselves up.
> Could just be a surge in enthusiasm after 10 years when all we had to be excited about was a new Javascript framework.
No. I distinctly remember HN being fairly skeptical of LLMs the way they were sold back in 2023. Then, the astroturfing began happening gradually. First in the comments and then in the posts.
Our jobs are basically sifting through PR garbage and making use of what it’s actually capable of. Obviously there’s a lot of fluff, but there is real use cases for myself as well.
Yeah, I was very skeptical and now I've drank the kool-aid. Got multiple parallel cloud agents with independent loops making PRs and me reviewing them. AI is also reviewing PRs and sometimes even catches useful things.
LLMs in 2023, from what I remember, were slightly better auto complete. It's pretty different now.
I'm using AI in many ways but I'm not ignoring that it has some real costs. The code AI writes is too voluminous to fully review. I'm quite disconnected from it, and probably won't be able to maintain it myself. The more we use AI to code, the more our real skills atrophy (already feeling that). And when AI pricing finally catches up with the astronomical spending going into it, people are going to have a rude awakening. Nothing good is coming from any of this, unless you're able to cash in before the economy collapses when this bubble finally bursts.
AI pricing finally catching up? What do you mean? Every few months prices are going down by 5x. Your price correction will not happen and it’s time to accept it.
Yes I agree with this - models today are better than they were a year ago. But not sure if I can "do for $1 what used to cost $25" across the board (meaning two times occurred 5x price reduction). But it's maybe not that simple to compare as there are some simple things also that just were not feasibly possible at all some time ago.
I guess it depends, but what are you doing with all your time if you aren't reviewing the code and trying to make it maintainable?
The time I previously spent writing code is now spent reviewing code and giving feedback so that it gets written differently if don't like it.
The skills required have changed completely imo. You need to be able to context switch efficiently and review "better". Also identify which tasks require lots of supervision of what the AI generates and which don't.
I think AI is here to stay. Open models keep the costs at a certain level. If hosted solutions get too expensive, people will selfhost. Even if models never improve beyond today, the open ones are good enough at this point to provide immense value.
Multiple comments made by this account in a timeframe I at least would personally have a hard time reading the comment I’m responding to and then and replying to this quickly, and from an account made in 2025.
Kind of making the point for the people you disagree with, assuming you are an actual human and not a bot tasked with pushing a narrative.
I understand why forums don't expose user IDs for scraping, but man it would be nice to just be able to tag users online with accounts created after the introduction of ChatGPT.
I'm not gonna say where we're at now is better, but as someone who has never had any interest in webdev, the perpetual JavaScript frameworks baffled me. How could they still be interesting and how was there not one that was good enough yet?
Yeah but that's just how things have gone. My enthusiasm about a new JavaScript framework is practically zero these days compared to several years ago, and I do genuinely use AI for all of my coding especially since I developed a health issue that prevents me from typing
Or, you know, many people genuinely find it exciting to see AI pumping their long overdue side-projects, just like many people once found bikeshedding their frontend frameworks exciting?
I don’t suppose that could have anything to do with the fact that AI has become a huge part of day to day business in the tech industry and elsewhere? Totally mystifying that anyone would say anything about that!
People are easily divided and suddenly in a binary position.
You're either totally pro AI and can't stop talking about it all the time, it's like your new god.
Or you're a hater. And old-school person, who deceived themselves and got a good portion of their self-worth from typing strange invocations in an even stranger electronic device.
Both are bullshit, but if you're not in either camp you get flack from both.
My (!) reality is that it's useful technology, but far from the claims that were made by the people trying to sell (!) it.
I've seen quite some improvements over the years, but I haven't seen the claims materialized in any release in the last 18+ months.
Yes, it's impressive, yes, it's probably here to stay in some form or another.
> would have to be state law so the president can't pardon them
Sure. Based on cursory analysis, California (B&P § 17200 + § 17941), New York (§ 63(12)) and New Hampshire (RSA 664:14 / 664:16 + RSA 358-A) seem to be good places to start. Worst case, we get a public record of who paid for what under what pretences.
Unfortunately you're not going to see any sort of legal redress against American oligarchs in any sort of American legal venue.
A country like France has shown that they're willing to take no an oligarch[0], and they have the nuclear means to protect themselves from American oligarchs leveraging the American military to punish France for doing so.
This kind of stuff will continue until a country like France forces a jet carrying someone like Musk, Altman, Thiel, or Zuckerberg to land so that they can be arrested, tried, convicted, and forced to spend a few years to decades in a drab little room.
You know whats crazy? We’re in bizarre times were this can actually be true. Things we joked about before which was obviously too nonsensical to be true is no longer the case in my world.
Funny, I was told by HN commenters that worries about AI risk are just a clever marketing strategy on the part of OpenAI and friends. I wonder why they would be working so hard to push back?
Recent events have shown that it's a double-edged sword. Sure the Mythos hype probably drove a lot of sales and investment, but it also led to a crackdown from the government.
Regulatory capture. They are the hyperscalers who invested and still invest heavily in hard tech, like data centers. They need regulation of "uncontrolled" AI models, so that Joe Smoe doesn't ditch his ChatGPT subscription and uses his local LLMs and agent harness setup in his home lab. Or some open weights model, or the condensed Chinese ones... it's "regulate everything else to death that is better than our enshittified OG models behind a paywall".
The trajectory is clear and the hyperscalers are in an disadvantage: they need to become bigger and bigger while models, frameworks and harnesses that are "good enough" for 95% of all people become smaller and specialized hardware becomes cheaper and less energy hungry. In the end the only viable customers that are left for the big ones is the military and... the military. But even the Pentagon realizes that a decentralized less fragile approach is actually safer instead of having big data centers that could be Wildberried by an adversary, that is if the alcoholic in charge isn't firing some generals who act too smart again. Machine vision which isn't even real AI is capable of destroying targets in Ukraine for instance. Dirt cheap. And Palantir's Maven Smart System was incapable of predicting the most obvious flaws in the Iran war, like how easy it is to to strike US bases or close the Strait of Hormuz. I'm not saying condensing the killchain is useless, I'm saying that it's not everything.
People aren't dumb. So they must be strong-manned and cheated into submission.
If AI industry critics are valuable as a means of OpenAI achieving regulatory capture, why is OpenAI attacking critics using underhanded means, as in the OP?
>But there don’t seem to be any contributors, or at least not human ones. I ran every article on the site through Pangram (the AI detector that credibly claims a near-zero false-positive rate). Of the 94 articles, 69% came back flagged as fully AI-generated, with another 28% flagged as partially AI-generated. Only three articles were classified as human-authored.
If anything this shows how pangram is unreliable. Chances are, if someone is running an article generation thing, it'll be 100% generated. Writing 3 / 100 articles "by hand" doesn't make any sense.
They do seem to have an internal review process, some articles might have been heavily edited, particularly early in the development process.
But indeed, if Pangram is marketing itself as "near-zero false-positives" now, that entails that the are aggressively biasing it towards being conservative, likely leading to lots of false-negatives and low match scores on AI content.
Regardless of that claim, here’s a good test for how real this site is - it’s called “The Wire by Acutus”, but try to find a site for “Acutus”.
Is it Acutus Medical, a “cardiac arrhythmia management company”, or Acutus Partners, “Expert healthcare consulting for transformative growth strategies that empower healthcare organizations to thrive and succeed”?
Neither seem obviously correct, and there are no other good candidates, not to mention that the lack of a link from “The Wire” to its supposed parent is suspicious.
> When I ran the message through Pangram, an AI content detector that credibly claims a near-zero false-positive rate, the email came back as “fully AI-generated.”
Uh huh.. AI content detectors don't work.
This article plays right into my biases but I don't think it's actually very credible? They're just guessing that some unknown blog is funded by OpenAI because it has maybe-AI-generated articles that use similar talking points
"These are the talking points of Leading The Future, the $125 million super PAC funded primarily by OpenAI president Greg Brockman"
I'm sure OpenAI is doing a tonne of astroturfing but my god this is weak
I agree that the connection to OpenAI is very speculative here. It's true that Pangram is also very unreliable, and the match is pretty weak anyways: "69% came back flagged as fully AI-generated, with another 28% flagged as partially AI-generated".
But the analysis of the internal admin pages for generating articles, the API and the interview structure are pretty damming evidence that this is largely driven by AI.
Oh with that I wasn't arguing the blog wasn't AI. I'm sure it is. My point there is that stating that any Pangram settles anything is a huge mark against the author's credibility
The only reason OpenAI wants to give 5% of the company to the US government is so they become 'to big to fail'. They dont give a s** about the US, its just a way to get funding when the bubble bursts
“Using your own firm’s president as a supposedly independent expert, without acknowledging the relationship, is a striking departure from basic journalistic norms”
Well that’s about the nicest possible way to phrase that…
Why do people insist on giving shady characters the benefit of the doubt? It’s pretty clear what’s happening here and yet the level of effort put into avoiding judgement seems excessive.
This place average post went from “look at this new javascript framework” to “I just became 100x more productive with AI” in the last 3 years