Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

A new AI winter is coming?

taranis.ie

Post reply on HN
Print view View original

21–30 of 301 posts

Re: A new AI winter is coming?

#21

Post by gishh »

stanfordkid wrote: post #16

This article uses the computational complexity hammer way too hard, discounts huge progress in every field of AI outside of the hot trend of transformers and LLMs. Nobody is saying the future of AI is autoregressive and this article pretty much ignores any of the research that has been posted here around diffusion based text generation or how it can be combined with autoregressive methods… discounts multi-modal model…

> something fundamental has changed that enables a computer to pretty effectively understand natural language.

You understand how the tech works right? It's statistics and tokens. The computer understands nothing. Creating "understanding" would be a breakthrough.

Edit: I wasn't trying to be a jerk. I sincerely wasn't. I don't "understand" how LLMs "understand" anything. I'd be super pumped to learn that bit. I don't have an agenda.

Re: A new AI winter is coming?

#22

Post by api »

aroman wrote: post #11

When the hype is infinite (technological singularity and utopia), any reality will be a let down. But there is so much real economic value being created - not speculation, but actual business processes - billions of dollars - it’s hard to seriously defend the claim that LLMs are “failures” in any practical sense. Doesn’t mean we aren’t headed for a winter of sobering reality… but it doesn’t invalidate the disruption…

This is why I hate hype.

"We just cured cancer! All cancer! With a simple pill!"

"But you promised it would rejuvenate everyone to the metabolism of a 20 year old and make us biologically immortal!"

New headline: "After spending billions, project to achieve immortality has little to show..."

Re: A new AI winter is coming?

#23

Post by keybored »

aroman wrote: post #11

When the hype is infinite (technological singularity and utopia), any reality will be a let down. But there is so much real economic value being created - not speculation, but actual business processes - billions of dollars - it’s hard to seriously defend the claim that LLMs are “failures” in any practical sense. Doesn’t mean we aren’t headed for a winter of sobering reality… but it doesn’t invalidate the disruption…

Hype Infinity is a form of apologia that I haven’t seen before.

Re: A new AI winter is coming?

#24

Post by neom »

Blog posts like this make me think model adoption and appropriate use case for the model is...lumpy at best. Every time I read something like it I wonder what tools they are using and how? Modern systems are not raw transformers. A raw transformer will “always output something,” they're right, but nobody deploys naked transformers. This is like claiming CPUs can’t do long division because the ALU doesn’t natively understand decimals. Also, a model is stat aprox trained on the empirical distribution of human knowledge work. It is not trying to compute the exact solution to NP complete problems? Nature does not require worst case complexity, real world cognitive tasks are not worst case NP hardness instances...

Re: A new AI winter is coming?

#25

Post by cardanome »

deadbabe wrote: post #2

Modern LLMs could be like the equivalent of those steam powered toys the Romans had in their time. Steam tech went through a long winter before finally being fully utilized for the Industrial Revolution. We’ll probably never see the true AI revolution in our lifetime, only glimpses of what could be, through toys like LLMs. What we should underscore though, is that even if there is a new AI winter, the world isn’t goi…

The development of LLM's required access to huge amounts of decent training data.

It could very well that the current generation of AI has poisoned the well for any future endeavors of creating AI. You can't trivially filter out the AI slop and humans are less likely to make their handcrafted content freely available for training. In fact violating GPL code to train models on it might be ruled to be illegal as well generally stricter rules on which data you are allowed to use for training.

We might have reached a local optimum that is very difficult to escape from. There might be a long, long AI winter ahead of us, for better or worse.

> the world isn’t going back to what it was before AI. This is it, forever.

I feel this so much. I though my longing for the pre-smartphone days was bad but damn we have lost so much.

Re: A new AI winter is coming?

#26

Post by hombre_fatal »

agubelu wrote: post #9
tarr11 wrote: post #7

This has convinced many non-programmers that they can program, but the results are consistently disastrous, because it still requires genuine expertise to spot the hallucinations. I've been programming for 30+ years and now a people manager. Claude Code has enabled me to code again and I'm several times more productive than I ever was as an IC in the 2000s and 2010s. I suspect this person hasn't really tried the most…

Isn't that what the author means? "it still requires genuine expertise to spot the hallucinations" "works very well if you do know what you are doing"

But it can work well even if you don't know what you are doing (or don't look at the impl).

For example, build a TUI or GUI with Claude Code while only giving it feedback on the UX/QA side. I've done it many times despite 20 years of software experience. -- Some stuff just doesn't justify me spending my time credentializing in the impl.

Hallucinations that lead to code that doesn't work just get fixed. Most code I write isn't like "now write an accurate technical essay about hamsters" where hallucinations can sneak through lest I scrutinize it; rather the code would just fail to work and trigger the LLM's feedback loop to fix it when it tries to run/lint/compile/typecheck it.

But the idea that you can only build with LLMs if you have a software engineer copilot isn't true and inches further away from true every month, so it kinda sounds like a convenient lie we tell ourselves as engineers (and understandably so: it's scary).

Re: A new AI winter is coming?

#27

Post by sukhdeepprashit »

gishh wrote: post #21
stanfordkid wrote: post #16

This article uses the computational complexity hammer way too hard, discounts huge progress in every field of AI outside of the hot trend of transformers and LLMs. Nobody is saying the future of AI is autoregressive and this article pretty much ignores any of the research that has been posted here around diffusion based text generation or how it can be combined with autoregressive methods… discounts multi-modal model…

> something fundamental has changed that enables a computer to pretty effectively understand natural language. You understand how the tech works right? It's statistics and tokens. The computer understands nothing. Creating "understanding" would be a breakthrough. Edit: I wasn't trying to be a jerk. I sincerely wasn't. I don't "understand" how LLMs "understand" anything. I'd be super pumped to learn that bit. I don't…

[dead]

Re: A new AI winter is coming?

#28

Post by numbers_guy »

AlexNet was only released in 2011. The progress made in just 14 years has been insane. So while I do agree that we are approaching a "back to the drawing board" era, calling the past 14 years a "failure" is just not right.

Re: A new AI winter is coming?

#29

Post by pzo »

agubelu wrote: post #9
tarr11 wrote: post #7

This has convinced many non-programmers that they can program, but the results are consistently disastrous, because it still requires genuine expertise to spot the hallucinations. I've been programming for 30+ years and now a people manager. Claude Code has enabled me to code again and I'm several times more productive than I ever was as an IC in the 2000s and 2010s. I suspect this person hasn't really tried the most…

Isn't that what the author means? "it still requires genuine expertise to spot the hallucinations" "works very well if you do know what you are doing"

The author headline starts with "LLMs are a failure", hard to take author seriously with such a hyperbole even if second part of headline ("A new AI winter is coming") might be right.

Re: A new AI winter is coming?

#30

Post by weare138 »

tarr11 wrote: post #7

This has convinced many non-programmers that they can program, but the results are consistently disastrous, because it still requires genuine expertise to spot the hallucinations. I've been programming for 30+ years and now a people manager. Claude Code has enabled me to code again and I'm several times more productive than I ever was as an IC in the 2000s and 2010s. I suspect this person hasn't really tried the most…

..and works very well if you do know what you are doing

That's the issue. AI coding agents are only as good as the dev behind the prompt. It works for you because you have an actual background in software engineering of which coding is just one part of the process. AI coding agents can't save the inexperienced from themselves. It just helps amateurs shoot themselves in the foot faster while convincing them they're a marksman.

Post reply on HN