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

Yann LeCun raises $1B to build AI that understands the physical world

wired.com

Post reply on HN
Print view View original

501–510 of 529 posts

Re: Yann LeCun raises $1B to build AI that understands the physical world

#501

Post by ordu »

famouswaffles wrote: post #496
ordu wrote: post #231

Earlier quoted context omitted.

> Of course you can still improve the models, but you get much more upside from data, or even better - from interactive environments. I'm on the contrary believe that the hunt for better data is an attempt to climb the local hill and be stuck there without reaching the global maximum. Interactive environments are good, they can help, but it is just one of possible ways to learn about causality. Is it the best way? I…

>With only instructional materials (a 500-page reference grammar, a dictionary, and ≈400 extra parallel sentences) all provided in context, Gemini 1.5 Pro and Gemini 1.5 Flash are capable of learning to translate from English to Kalamang— a Papuan language with fewer than 200 speakers and therefore almost no online presence—with quality similar to a person who learned from the same materials https://arxiv.org/abs/240…

I'm not entirely sure, that I totally convinced, but yeah, it is better than me. I mean, I could do the same, but it would take me ages to go through 500 pages and to use them for the actual translation.

I'm not sure, because Gemini knows a lot of languages. The third language is easier to learn than the second one, I suppose 100th language is even easier? But still Gemini do better, than I believed.

Re: Yann LeCun raises $1B to build AI that understands the physical world

#502

Post by sylware »

ismailmaj wrote: post #411
sylware wrote: post #391

Earlier quoted context omitted.

What??? microsoft openai is Big Tech. Are you ok?

Ah yes, OpenAI the puppet of Microsoft that is currently declaring war against GitHub, sounds logical.

Internal battles or scheme to dodge anti-trust regulations. Don't let them fool you, it is Big Tech.

Re: Yann LeCun raises $1B to build AI that understands the physical world

#503

Post by sothatsit »

HarHarVeryFunny wrote: post #449
sothatsit wrote: post #344

Earlier quoted context omitted.

You can’t really say it is just predicting continuations when it is learning to write proofs for Erdos problems, formalise significant math results, or perform automated AI research. Those are far beyond what you get by just being a copying and re-forming machine, a lot of these problems require sophisticated application of logic. I don’t know if this can reach AGI, or if that term makes any sense to begin with. But…

> What do you think training to predict when to use different continuations is other than learning? Sure, training = learning, but the problem with LLMs is that is where it stops, other than a limited amount of ephemeral in-context learning/extrapolation. With an LLM, learning stops post-training when it is "born" and deployed, while with an animal that's when it starts! The intelligence of an animal is a direct resu…

Memory systems built on top of LLMs could provide continual learning. I do not agree that it is some fundamental limitation.

Claude Code already writes its own memory files. And people already finetune models. There is clear potential to use the former as a form of short-term memory and the latter for long-term “learning”.

The main blockers to this are that models aren’t good enough at managing their own memory, and finetuning is expensive and difficult. But both of these seem like solvable engineering problems.

Re: Yann LeCun raises $1B to build AI that understands the physical world

#504

Post by qsera »

HarHarVeryFunny wrote: post #500
qsera wrote: post #498

Earlier quoted context omitted.

> How do you see emotion as being necessary for creativity? As its reward. Why should anyone be creative if it does not "feel" good? >releasing chemicals like adrenaline, dopamine.. That does not explain how it results in a certain kind of "feeling".. >FWIW I think consciousness is also very mechanical I think I didn't really do an OK job of why I think otherwise in my comment above...

> As its reward. Why should anyone be creative if it does not "feel" good? We've evolved to do things that are helpful for survival. Things like curiosity and exploration don't need to feel good - it's just our nature. They only need to attract and keep our attention.

>it's just our nature

Yea, its our nature to feel good about it, which is what evolution does. If you are curious, and if exploration makes you feel good, you have a better chance to survive, and you pass that trait along.

I can't believe we are arguing about this.

Re: Yann LeCun raises $1B to build AI that understands the physical world

#505

Post by rhubarbtree »

pixl97 wrote: post #424
rhubarbtree wrote: post #376

Earlier quoted context omitted.

Brilliant insight. The success of LLM reasoning, ie “telling yourself a story”, has greatly increased my belief that humans are actually much less impressive than they seem. I do think it’s mostly pattern matching and a bunch of interacting streams analogous to LLM tokens. Obviously the implementations are different, because nature has to be robust and learn online, but I do not think we are as different from these m…

This is why I also think humans being logical inference machines is mostly not true. We are seemingly capable of it, but there must be some cost that keeps it from being commonly used. While humans did seemingly evolve socially very fast, with the tools we seem to have had for a few hundred thousand years it could have been far faster if there were not some other limitations that are being applied.

Agreed. This also explains why maths is so difficult for humans. It doesn't come "naturally" to use, we have to force ourselves to use it and it "makes our head hurt".

Re: Yann LeCun raises $1B to build AI that understands the physical world

#506

Post by _giorgio_ »

snek_case wrote: post #436
_giorgio_ wrote: post #408

LeCun has had every advantage imaginable — and the scoreboard remains empty. He joined Facebook (now Meta) in December 2013. That's over 12 years of access to one of the largest AI labs in the world, near-unlimited compute, and some of the best researchers money can buy. He introduced I-JEPA in 2023, nearly 3 years ago . It was supposed to represent a fundamental shift in how machines learn — moving beyond generative…

First, believe it or not, 3 years is not that long. It's also not a given that LeCun was given the resources he needed to work on this tech at Meta. Zuck wanted another llama. Second, AMI Labs just secured a billion in funding, and while that's a lot of money, it's literally just a fraction of the yearly salary they are paying to Wang. Big tech companies are literally throwing tens of billions to keep doing the same…

He has coauthored tens of papers on the same subject.

This means sponsorships, millions spent on training etc.

Re: Yann LeCun raises $1B to build AI that understands the physical world

#507

Post by _giorgio_ »

samrus wrote: post #416
_giorgio_ wrote: post #171

Earlier quoted context omitted.

If his ideas had real substance, we would have seen substantial results by now. He introduced I-JEPA in 2023, so almost three years ago at this point. If he still hasn’t produced anything truly meaningful after all these years at Meta, when is that supposed to happen? Yann LeCun has been at Facebook/Meta since December 2013. Your chronological sequence is interesting, but it refers to a time when the number of resear…

> If his ideas had real substance, we would have seen substantial results by now This is naive. Like saying if backprop had any real substance, it would have had results within 10 years of its publication in 1989 > Your chronological sequence is interesting, but it refers to a time when the number of researchers and the amount of compute available were a tiny fraction of what they are today. Again. Those resources ar…

Backprop kept producing wins. That bought it time.

“Wait longer” is not a blank check. In 2026, with Meta-scale talent, data, and compute, serious ideas should show strong intermediate results, not just theory.

Time is necessary, but it is not evidence. More compute does not replace insight, but it does speed up falsification.

So no, skepticism is not naive. If a research program still cannot point to a clear empirical advantage after years, “it just needs more time” stops sounding like science and starts sounding like insulation from the scoreboard.

Re: Yann LeCun raises $1B to build AI that understands the physical world

#508

Post by HarHarVeryFunny »

sothatsit wrote: post #503
HarHarVeryFunny wrote: post #449

Earlier quoted context omitted.

> What do you think training to predict when to use different continuations is other than learning? Sure, training = learning, but the problem with LLMs is that is where it stops, other than a limited amount of ephemeral in-context learning/extrapolation. With an LLM, learning stops post-training when it is "born" and deployed, while with an animal that's when it starts! The intelligence of an animal is a direct resu…

Memory systems built on top of LLMs could provide continual learning. I do not agree that it is some fundamental limitation. Claude Code already writes its own memory files. And people already finetune models. There is clear potential to use the former as a form of short-term memory and the latter for long-term “learning”. The main blockers to this are that models aren’t good enough at managing their own memory, and…

Continual learning isn't a "fundamental limitation" or unsolvable problem. Animal brains are an existence proof that it's possible, but it's tough to do, and quite likely SGD is not the way to do it, so any attempt to retrofit continual learning to LLMs as they exist today is going to be a hack...

Memory and learning are two different things. Memorization is a small subset of learning. Memorizing declarative knowledge and personal/episodic history (cf. LLM context) are certainly needed, but an animal (or AI intern) also needs to be able to learn procedural skills which need to become baked into the weights that are generating behavior.

Fine tuning is also no substitute for incremental learning. You might think of it as addressing somewhat the same goal, but really fine tuning is about specializing a model for a particular use, and if you repeatedly fine tune a model for different specializations (e.g. what I learnt yesterday, vs what I learnt the day before) then you will run into the catastrophic forgetting problem.

I agree that incremental learning seems more like an engineering problem rather than a research one, or at least it should succumb to enough brain power and compute put into solving it, but we're now almost 10 years into the LLM revolution (attention paper in 2017) and it hasn't been solved yet - it's not easy.

Re: Yann LeCun raises $1B to build AI that understands the physical world

#509

Post by HarHarVeryFunny »

qsera wrote: post #504
HarHarVeryFunny wrote: post #500

Earlier quoted context omitted.

> As its reward. Why should anyone be creative if it does not "feel" good? We've evolved to do things that are helpful for survival. Things like curiosity and exploration don't need to feel good - it's just our nature. They only need to attract and keep our attention.

>it's just our nature Yea, its our nature to feel good about it, which is what evolution does. If you are curious, and if exploration makes you feel good, you have a better chance to survive, and you pass that trait along. I can't believe we are arguing about this.

It seems we're basically in agreement, not arguing!

The only quibble I have is whether "feeling good" is the right way to describe how evolution has made us choose to engage in exploration/etc. I don't think it's quite as simple as evolution making things that are good for us feel good, and making things that are bad for us feel bad.

There are a bunch of neurotransmitters and hormones that control how we behave. Evolution discourages us from doing things that are bad for us via a range of emotions including things like fear and disgust (not just "feeling bad"). Evolution also encourages us to do, or keep on doing, things that are good for us via a range of emotions such as enjoyment (this is a tasty fruit), contentment (this feels nice - I'll keep doing it) to curiosity, again not just "feeling good". I think curiosity and exploration (which may lead to learning and discovery, which are good for us) are based around attention and focus ... rather than feeling good, it feels interesting.

Re: Yann LeCun raises $1B to build AI that understands the physical world

#510

Post by HarHarVeryFunny »

sothatsit wrote: post #499
HarHarVeryFunny wrote: post #444

Earlier quoted context omitted.

I think you're conflating mechanism with function/capability. I'm not sure what I wrote that made you conclude that I thought these models are not learning anything from their RL training?! Let me say it again: they are learning to steer towards reasoning steps that during training led to rewards. The capabilities of LLMs, both with and without RL, are a bit counter-intuitive, and I think that, at least in part, come…

This is what you said: > they are still predicting training set continuations But this is underselling what they do. Probably a large part of what they predict is learnt from their training set, but RL has added a layer on top that does not just come from just mimicry. Again, I doubt this is enough for “AGI” but I think that term is not very well-defined to begin with. These models have now shown they are capable of…

In general there's a difference between novel and discovering something new.

Pretraining has given the LLM a huge set of lego blocks that it can assemble in a huge variety of ways (although still limited by the "assembly patterns" is has learnt). If the LLM assembles some of these legos into something that wasn't directly in the training set, then we can call that "novel", even though everything needed to do it was present in the training set. I think maybe a more accurate way to think of this is that these "novel" lego assemblies are all part of the "generative closure" of the training set.

Things like generating math proofs are an example of this - the proof itself, as an assembled whole, may not be in the training set, but all the piece parts and thought patterns necessary to construct the proof were there.

I'm not much impressed with Karpathy's LLM autoresearch! I guess this sort of thing is part of the day to day activities of an AI researcher, so might be called "research" in that regard, but all he's done so far is just hyperparameter tuning and bug fixing. No doubt this can be extended to things that actually improve model capability, such as designing post-training datasets and training curriculums, but the bottleneck there (as any AI researcher will tell you) isn't the ideas - it's the compute needed to carry out the experiments. This isn't going to lead to the recursive self-improvement singularity that some are fantasizing about!

I would say these types of "autoresearch" model improvements, and pretty much anything current LLMs/agents are capable of, all fall under the category of "generative closure", which includes things like tool use that they have been trained to do.

It may well be possible to retrofit some type of curiosity onto LLMs, to support discovery and go beyond "generative closure" of things it already knows, and I expect that's the sort of thing we may see from Google DeepMind in next 5 years or so in their first "AGI" systems - hybrids of LLMs and hacks that add functionality but don't yet have the elegance of an animal cognitive architecture.

Post reply on HN