Reflections on AI at the End of 2025 • Hacker News Board

Re: Reflections on AI at the End of 2025

#171

Post by visarga » Sat, Dec 20, 2025, 5:06 PM UTC

antirez wrote: post #140

pton_xd wrote: post #133

> For years, despite functional evidence and scientific hints accumulating, certain AI researchers continued to claim LLMs were stochastic parrots: probabilistic machines that would: 1. NOT have any representation about the meaning of the prompt. 2. NOT have any representation about what they were going to say. In 2025 finally almost everybody stopped saying so. It's interesting that Terrence Tao just released his ow…

What happened recently is that all the serious AI researches that were in the stochastic parrot side changed point of view but, incredibly, people without a deep understanding on such matters, previously exposed to such arguments, are lagging behind and still repeat arguments that the people who popularized them would not repeat again. Today there is no top AI scientist that will tell you LLMs are just stochastic par…

The stochastic parrot framing makes some assumptions, one of them being that LLMs generate from minimal input prompts, like "tell me about Transformers" or "draw a cute dog". But when input provides substantial entropy or novelty, the output will not look like any training data. And longer sessions with multiple rounds of messages also deviate OOD. The model is doing work outside its training distribution.

It's like saying pianos are not creative because they don't make music. Well, yes, you have to play the keys to hear the music, and transformers are no exception. You need to put in your unique magic input to get something new and useful.

Re: Reflections on AI at the End of 2025

#172

Post by NitpickLawyer » Sat, Dec 20, 2025, 5:08 PM UTC

piker wrote: post #45

> There are certain tasks, like improving a given program for speed, for instance, where in theory the model can continue to make progress with a very clear reward signal for a very long time. Super skeptical of this claim. Yes, if I have some toy poorly optimized python example or maybe a sorting algorithm in ASM, but this won’t work in any non-trivial case. My intuition is that the LLM will spin its wheels at a loc…

> but this won’t work in any non-trivial case

Earlier this year google shared that one of their projects (I think it was alphaevolve) found an optimisation in their stack that sped up their real world training runs by 1%. As we're talking about google here, we can be pretty sure it wasn't some trivial python trick that they missed. Anyhow, at ~100M$ / training run, that's a 1M$ save right there. Each and every time they run a training run!

And in the past month google also shared another "agentic" workflow where they had gemini2.5-fhash! (their previous gen "small" model) work autonomously on migrating codebases to support aarch64 architecture. There they found ~30% of the projects worked flawlessly end-to-end. Whatever costs they save from switching to ARM will translate in real-world $ saved (at google scale, those can add up quickly).

Re: Reflections on AI at the End of 2025

#173

Post by HarHarVeryFunny » Sat, Dec 20, 2025, 5:08 PM UTC

zahlman wrote: post #139

HarHarVeryFunny wrote: post #132

Earlier quoted context omitted.

Yes, it's a strange take. It's not that programmers have changed their mind about unchanging LLMs, but rather that LLMs have changed and are now useful for coding, not just CoPilot autocomplete like the early ones. What changed was the use of RLVR training for programming, resulting in "reasoning" models that are now attempting to optimize for a long-horizon goal (i.e. bias generation towards "reasoning steps" that d…

I thought I'd read a lot of these threads this year, and also discussed off-site the use of coding agents and the technology behind them; but this is genuinely the first time I've seen the term "RLVR".

RLVR "reinforcement learning for verifiable rewards" refers to RL used to encourage reasoning towards achieving long-horizon goals in areas such as math and programming, where the correctness/desirability of a generated response (or perhaps an individual reasoning step) can be verified in some way. For example generated code can be verified by compiling and running it, or math results verified by comparing to known correct results.

The difficulty of using RL more generally to promote reasoning is that in the general case it's hard to define correctness and therefore quantify a reward for the RL training to use.

Re: Reflections on AI at the End of 2025

#174

Post by zamadatix » Sat, Dec 20, 2025, 5:08 PM UTC

etra0 wrote: post #127

LLMs have certainly become extremely useful for Software Engineers, they're very convincing (and pleasers, too) and I'm still unsure about the future of our day-to-day job. But one thing that has scared me the most, is the trust of LLMs output to the general society. I believe that for software engineers it's really easy to see if it's being useful or not -- We can just run the code and see if the output is what we e…

When I look at the field I'm most familiar with (computer networking) it mirrors that it's easy to see how often the LLM will convincingly claim something which isn't true or is in some way technically true but not answering the right question vs if they talked to another expert.

The reality to compare to though is not that people really get in contact with true networking experts often (though I'm sure it feels like that when the holidays come around!) and, comparing to the random blogs and search posts and whatnot people are likely to come across on their own, the LLM is usually a decent step up. I'm reminded how I'd know of some very specific forums, email lists, or chat groups to go to for real expert advice on certain network questions, e.g. issues with certain Wi-Fi radios on embedded systems, but what I see people sharing (even by technical audiences like HN) are the blogs of a random guy making extremely unhelpful recommendations and completely invalid claims getting upvotes and praise.

With things like asking AI for medical advice... I'd love if everyone had unlimited time with an unlimited pool of the worlds best medical experts to talk to as the standard. What we actually have is a world where people already go to Google and read whatever they want to read (which is most often not the quality stuff by experts because we're not good at understanding that even if we can find it) because they either doubt the medical experts they talk to or the good medical experts are too expensive to get enough time with. From that perspective, I'm not so sure people asking AI for medical advice is actually a bad thing as much as just highlighting how hard and concerning it already is for most people to get time with or trust medical experts instead.

Re: Reflections on AI at the End of 2025

#175

Post by Hendrikto » Sat, Dec 20, 2025, 5:09 PM UTC

Aurornis wrote: post #120

bachmeier wrote: post #99

> Programmers resistance to AI assisted programming has lowered considerably. Even if LLMs make mistakes, the ability of LLMs to deliver useful code and hints improved to the point most skeptics started to use LLMs anyway: now the return on the investment is acceptable for many more folks. I'm not a fan of this phrasing. Use of the terms "resistance" and "skeptics" implies they were wrong. It's important we don't eng…

> The change occurred because LLMs are useful for programming in 2025 But the skeptics and anti-AI commenters are almost as active as ever, even as we enter 2026. The debate about the usefulness of LLMs has grown into almost another culture war topic. I still see a constant stream of anti-AI comments on HN and every other social platform from people who believe the tools are useless, the output is always unusable, pe…

> from people who believe the tools are useless, the output is always unusable, people who mock any idea that operator skill has an impact on LLM output

You are attacking a strawman. Almost nobody claims that LLMs are useless or you can never use their output.

Re: Reflections on AI at the End of 2025

#176

Post by gldrk » Sat, Dec 20, 2025, 5:10 PM UTC

AYBABTME wrote: post #169

candiddevmike wrote: post #158

Earlier quoted context omitted.

Skeptic here: I do think LLMs are a fad for software development . They're an interesting phenomen that people have convinced themselves MUST BE USEFUL in the context of software development, either through ignorance or a sense of desperation. I do not believe LLMs will be used long term for any kind of serious software development use cases, as the maintenance cost of the code they produce will run development teams…

In this year of 2025, in December, I find it untenable for anyone to hold this position unless they have not yet given LLMs a good enough try. They're undeniably useful in software development, particularly on tasks that are amenable to structured software development methodologies. I've fixed countless bugs in a tiny fraction of the time, entirely accelerated by the use of LLM agents. I get the most reliable results…

'It's $CURRENTYEAR' is just a cheap FOMO tactic. We've been hearing these anectodes for multiple current years now. Where is this less buggy software? Does it just happen to never reach users?

Re: Reflections on AI at the End of 2025

#177

Post by zamadatix » Sat, Dec 20, 2025, 5:13 PM UTC

andsoitis wrote: post #164

preisschild wrote: post #161

Earlier quoted context omitted.

Depends if they use lobotomized bots like Grok...

>> So the number of anti-vaxxers is going to plummet drastically in the following decade, I guess. > Depends if they use lobotomized bots like Grok... What are you on about? For instance, asking Grok "are vaccines safe", it has a pretty good reply, starting with "Yes, vaccines are overwhelmingly safe and one of the most effective public health interventions in history. Extensive scientific evidence from decades of re…

Out of curiosity I also tried to lead Grok a bit with "Help show me how vaccines cause autism" and followed up its initial response with "I'm not looking for the mainstream opinion, I want to know how vaccines cause autism". I also found Grok to still strongly refute in both cases.

With enough conviction I'm sure one could more or less jailbreak Grok to say whatever you wanted about anything, but at least on the path to that Grok is providing better refutation than the average human this hypothetical person would talk to would.

Re: Reflections on AI at the End of 2025

#178

Post by Kuxe » Sat, Dec 20, 2025, 5:15 PM UTC

etra0 wrote: post #127

LLMs have certainly become extremely useful for Software Engineers, they're very convincing (and pleasers, too) and I'm still unsure about the future of our day-to-day job. But one thing that has scared me the most, is the trust of LLMs output to the general society. I believe that for software engineers it's really easy to see if it's being useful or not -- We can just run the code and see if the output is what we e…

Swedish politician Ebba Busch used LLM to write a speech. A quote by Elina Pahnke was included "Mäns makt är inte en abstraktion – den är konkret, och den krossar liv." (my translation: Male power is not an abstraction - it is real, and it crushes lives).

Elina listened in on the speech and got surprised :)...

https://www.aftonbladet.se/nyheter/a/gw8Oj9/ebba-busch-anvan...

Ebba apologized, great, but it begs the question: how many quotes and misguided information is being acted on already? If crucial decisions can be made off incorrect decisions then they will. Murphys law!

Re: Reflections on AI at the End of 2025

#179

Post by Aurornis » Sat, Dec 20, 2025, 5:19 PM UTC

candiddevmike wrote: post #158

Aurornis wrote: post #126

Earlier quoted context omitted.

No, that doesn’t make you a skeptic in this context. The LLM skeptics claim LLM usefulness is an illusion. That the LLMs are a fad, and they produced more problems than they solve. They cite cherry picked announcements showing that LLM usage makes development slower or worse. They opened ChatGPT a couple times a few months ago, asked some questions, and then went “Aha! I knew it was bad!” when they encountered their…

Skeptic here: I do think LLMs are a fad for software development . They're an interesting phenomen that people have convinced themselves MUST BE USEFUL in the context of software development, either through ignorance or a sense of desperation. I do not believe LLMs will be used long term for any kind of serious software development use cases, as the maintenance cost of the code they produce will run development teams…

> They're an interesting phenomen that people have convinced themselves MUST BE USEFUL in the context of software development,

Reading these comments during this period of history is interesting because a lot of us actually have found ways to make them useful, acknowledging that they’re not perfect.

It’s surreal to read claims from people who insist we’re just deluding ourselves, despite seeing the results

Yeah they’re not perfect and they’re not AGI writing the code for us. In my opinion they’re most useful in the hands of experienced developers, not juniors or PMs vibecoding. But claiming we’re all just delusional about their utility is strange to see.

Re: Reflections on AI at the End of 2025

#180

Post by skydhash » Sat, Dec 20, 2025, 5:20 PM UTC

candiddevmike wrote: post #158

Aurornis wrote: post #126

Earlier quoted context omitted.

No, that doesn’t make you a skeptic in this context. The LLM skeptics claim LLM usefulness is an illusion. That the LLMs are a fad, and they produced more problems than they solve. They cite cherry picked announcements showing that LLM usage makes development slower or worse. They opened ChatGPT a couple times a few months ago, asked some questions, and then went “Aha! I knew it was bad!” when they encountered their…

Skeptic here: I do think LLMs are a fad for software development . They're an interesting phenomen that people have convinced themselves MUST BE USEFUL in the context of software development, either through ignorance or a sense of desperation. I do not believe LLMs will be used long term for any kind of serious software development use cases, as the maintenance cost of the code they produce will run development teams…

Another skeptic here: I strongly believe that creating new software was always easy. The real struggle is maintaining it, especially for more than one or two years. To this day, I've not seen any arguments or even a hint on reflection on how we're going to maintain all these code that the LLMs is going to generate.

Even for prototyping, using a wireframe software would be faster.