And the purpose there was research right. And the implication is that once they figured out it was causing problems it was turned off for forensic analysis.
Meanwhile in the real world hundreds of billions are being invested into humanoid robot development. In 2-3 years my Optimus 4 will do all the gardening and plumbing I need.
Look at the whole robot vacuum market. These aren't exactly great devices. They have low suction small bins and dont do a great job. People love them and think that they work so well. Why? Because they keep their house clean and avoid making the sorts of messes that would be easy to address with a larger vacuums.
We have nuclear reactors, helicopters, and self driving cars already. And we finally have powerful AI that we could only dream about just 5 years ago.
We now have everything we need to scale up humanoid robot training: algorithms, hardware, and money to buy a lot of training data/compute. Fierce competition and strong economic motivation will force rapid progress in this field.
It is rivaling Astra, on their own benchmark that they made (FrontierCode), that they ran themselves in their own closed-source ecosystem that isn’t reproducible by anyone.
Could you tell the difference between a grandmaster and stockfish if playing them online? If not, why would you care which one you are playing against?
I’ve never played against a grandmaster, but I have a feeling that he/she would play very different compared to me and would never make mistakes I could notice. Though admittedly I’m not very good at chess.
I've played both. The GM plays tremendously differently than Stockfish.
Engines - specifically heuristically-driven ones like Stockfish - don't play like a strong GM. They play engine-perfect chess, which isn't how a GM plays with any consistency.
I'm only a decent amateur (1550 USCF) but when I lose to a titled player it's largely explainable in human terms how it happened.
I think I'd care because the entity on the other side cares about the game in a similar way to me. It's not just the technical details of how the pieces move, it's a human interaction.
I'm currently running two 24/7 semi-autonomous AI research projects using Fable 5.1. It's on track to burn through my weekly quota in about 3 days. I check progress in the morning and in the evening, and provide some light steering.
See my sibling comment, you can probably robo-code yourself some tooling to alleviate a lot of that in a few hours but if you want help shoot me an e-mail. I'm interested in seeing other people's workflows.
What has all this token burn done for them, actually?
They have been consistently pushing AI frontier. What other impact do you want to see? A year ago they said that in a year they will have a level of capabilities of an AI research intern - I believe they have achieved it, even before Astra.
Personally I’d like to see them actually start benefiting humanity by doing all the things Sam has claimed they will like curing disease, cancer, global warming, etc.
But I guess a computer intern so we can avoid paying / training the next generation is better.
> Personally I’d like to see them actually start benefiting humanity by doing all the things Sam has claimed they will like curing disease, cancer, global warming, etc.
It makes more sense to leave curing disease & cancer to the experts, with tools (like AI) being developed by AI experts.
Call me crazy, but I want separate organizations and experts for medical vs finance vs space vs climate vs AI research.
What the op was pointing out is that guys like Altman and Dario are repeatedly saying they’re going to cure xyz diseases and solve xyz huge global problems. Maybe their companies will eventually do these things, but haven’t yet.
I don’t have an opinion either way, I think it’s too soon to tell if llms will be able to cure cancer or whatever. But at the very least it will be a good tool to help researchers do their jobs.
The thing is... AI is not going to solve any problems. People needs to solve their problems. AI can give us clever solutions, but its up to us to do it!
> Maybe their companies will eventually do these things, but haven’t yet.
I think they are working with customers to improve the LLMs and tools for these use-cases. They almost certainly also hire experts to help filter out nonsense, pseudo-science and help curate trusted knowledge bases for training, but it will almost certainly be the customers who deliver the major results, and the AI companies will claim some of the credit. That said, patents for important medicine might help with the bottom line, so I could imagine partnerships and JVs.
> at the very least it will be a good tool to help researchers do their jobs.
When that happens, OpenAI will own 100% of your life. I’d rather they keep spinning their wheels long enough for these problems to be solved elsewhere.
I think people very much should care about who ends up owning these solutions. The person or entity that controls things like that just has more power, which isn’t necessarily a good thing
I don't see how it's different from what we have now? Expensive cancer treatments already exist and are rationed by the companies who create them, then over time become more widely available.
If AI companies create better treatments they'll offer them to people who pay for them. So wealthier people will have better cancer treatments. If they can make them inexpensively, then a broader audience also gets treatment.
Eventually they lose IP rights and they become generics and the world gets better cancer treatments.
In all scenarios having them find better treatments benefits humans with cancer more than if they didn't.
I don't like trusting any organization, corporate or governmental, with control over people's lives, but it's obvious that we're not going to get cutting edge medical treatments without these organizations.
I've hired many AI research interns (and was one many years ago), and I agree with them - frontier models are currently at the level of an average AI research intern.
Am I the only one who's a bit disappointed that we're spending trillions, destroying the ecosystem, drowning democracies and learning in slop, preparing a big financial crash, all of this to achieve an "average AI research intern"?
A long time ago, I used to be a (AI-adjacent) research intern, and frankly, I wouldn't trust any non-trivial task to that younger me. Fortunately, by opposition to an already trained LLM or agent, I have the ability to learn, so I eventually got better.
And if you don't find "average AI research intern" impressive, I'm not sure what to tell you. Have the goalposts moved so far that open ended problem solving at "average CS student fresh out of the uni" levels is suddenly trivial?
Think of what AI was capable of in 2016. Or even 2022. Compare that to now. We had more AI progress in the last five years than I expected to happen in five decades.
The ecosystem absolutely is being destroyed. We're looking at anywhere from 3-5C warming by 2060, which is going to be devastating if not flat out apocalyptic. And wherever we're at in 2060 it's not like it's going to stop there, nor is it going to be comfortable until then. Things may start to crumble much sooner.
I don't believe AI and data centers have played that much of a role in this though, we could have powered those without burning billions of tons of coal and gas etc, and im sure there's already a significant fraction of green energy powering then depending on location. Anyway we would have been roughly in the same spot right now with or without AI and some new data centers. The media just loves spinning the narrative to make the hordes of sheep scream about anything other than the real issues.
We're not even looking at 5C of warming by 2100 realistically. Like, that was considered to be an unlikely extreme scenario in 2014 AR5, and also in the tightened down 2021 AR6, and things have happened since! Renewables are cheaper than ever, and Ukrainian war and Iranian war both curbed the appetite for long term fossil fuel power investment.
The median is what, a bit under 3C by 2100? Not even by 2060 - by 2100. And we're in 2026, so that's more than twice as slow as your expectation.
Agreed on AI not being a meaningful factor in climate change though. We'd have to go full "humankind is obsolete" technological singularity to have AI dominate energy use to this extent, and current numbers are nowhere near that. It's a FUD distraction from the real culprits: the fossil fuel energy complex. That's currently lobbying to slow the inevitable energy transition.
> that was considered to be an unlikely extreme scenario
By the same people who just realized we're missing 1.5C as we're blazing past it at mach 12, still accelerating not slowing down? You really believe those guys?
You have to understand that there are several camps of climate science. The mainstream ones like IPCC and UN etc are heavily politicized, they can't publish anything that isn't sugarcoated beyond recognition. At least I assume that's why they're so obviously wrong.
Here's a judgement I think is more realistic
> There is a strong probability that the ambition gap will lead to a temperature rise of 2 to 5 degrees Centigrade compared to pre-industrial temperatures by 2100, the realisation gap to a further rise of several degrees Centigrade.[1,2,VI] There is a danger that the mean temperature will already have risen by 3 degrees Centigrade by 2050.
Yes, I do. IPCC's reports are sensible. They're not unreliable just because they don't support the "doom and burning land" narratives.
By the way, there is no "just realized we're missing 1.5C". That projection was always the very low end of possibilities - the "assume rapid, radical climate action on global level" scenario.
Yes, that's a dumb thing to assume. We've never been on track for it. But the "assume extremely high emissions and no green transition ever, 5C+ by 2100" scenario on the other end is about as unlikely to materialize. Those are the boundaries of the expectation range - not median expectations.
I'm pessimistic. We're still producing more CO2 each year than the last, and several feedback loops are kicking in that accelerate the warming further such as permafrost thawing, arctic and antarctic sea ice disappearing, glaciers are melting, the amazon is being demolished, etc. I don't know how much of this is included in the projections.
I hope the optimists are right, it just doesn't look like it to me at all. It looks to me like we're speeding along right into the worst predictions and beyond. We're building lots of green energy production but it seems to just come in on top of existing and new fossil production not replace it.
It does seem like the CO2 output is plateauing which is good, but we really need it to start declining drastically very soon and I don't really see that happening with the current political climate. Also remember CO2 is far from the only greenhouse gas - methane, nitrous oxide and fluorinated gas emissions all seem to be rising rapidly still.
They are, in fact, included in the projections - we'd be on track to ~2C by 2100 instead of ~3C by 2100 if they weren't. They just aren't that big.
There is no "Make Earth Into Venus Feedback Loop Of Doom" that a lot of people seem to imagine when they hear "feedback loop". There is, however, a dozen of things that add about +5% each.
Let's say it is. What about the rest of my paragraph?
> And if you don't find "average AI research intern" impressive, I'm not sure what to tell you. Have the goalposts moved so far that open ended problem solving at "average CS student fresh out of the uni" levels is suddenly trivial?
At this stage, I'm the one who doesn't know what to tell you. It took me years to grow from "research intern" into a competent researcher (and parallel years to turn into a competent developer). The research interns I've worked with were... vaguely useful, at best?
Is this not what you wanted? You created a culture that dissolves responsibility by making it the worst thing to strife for - so everyone dissolves it, in processes, mass decisions and AI. Its the system, society, god, the great spirit. This is what you strove for, how can you be unhappy with things you demanded yourself?
Intelligence is a capability, not a level of knowledge, and in discussing AI vs human intelligence, or human vs dog, it is perhaps better to regard it as a species-specific capability, not an individual-specific one.
Was the genius or dumb person really born with different levels of learning ability, or were they just raised differently: nature vs nurture ?
The "intelligence" of one species vs another comes down to differences in cognitive architecture, ultimately reflected in ability to learn and predict/infer. Intelligence, as a capability, not IQ test score, is best regarded as ability to learn from experience and use that learning to accurately predict future outcomes, ranging from passive observation, to the outcome of one's own actions, to the ability to reason.
Comparing the intelligence (not knowledge) of different AI system to humans should therefore be assessed by comparing their ability to learn and use what they have learnt.
An LLM is what it is - a language model, not a learning system. It's really an expert system of sorts, highly capable in terms of what it can infer based on the knowledge it encodes, but with very limited ability to learn.
When you talk of comparing AI to a genius/professional/graduate/etc, you are really talking about comparing level of knowledge, comparing one expert system to another, which is fine and perhaps useful in some contexts, but it is not the same as comparing actual intelligence - learning ability, and especially so if you want to discuss general intelligence which is all about the ability to successfully take on any task (perhaps needing to learn it first), not just do well on some limited set of tasks you are already familiar with.
If ability to learn is limited to one modality, such as language, then that indicates a lack of generality. A good test for whether an AI system is in the same ballpark as a human in learning ability, aka intelligence, would be whether it can (at run-time) learn language itself, from a blank slate start, when running in a suitable environment.
>If ability to learn is limited to one modality, such as language, then that indicates a lack of generality. A good test for whether an AI system is in the same ballpark as a human in learning ability, aka intelligence, would be whether it can (at run-time) learn language itself, from a blank slate start, when running in a suitable environment.
So a transformer ? Run-time? Seems like an arbitrary distinction to me. Because humans function in a certain way, every learning system no matter how capable must function in the same way to be 'truly' intelligent ?
The mechanism is irrelevant if the results are similar. No, it doesn't have to be implemented the same way as a human to qualify as a human-level intelligence. That would be silly.
How is text-based memorization going to let you learn non-linguistic skills, whether animal/human level or beyond (based on new data senses)?
How is text-based memorization going to substitute for learning? The two are not the same. Perhaps this is more applicable to robotics than a text generator, but I also doubt an LLM could learn text-based skills like programming or math if it had not been pre-trained on them via SGD & RL, and had to instead rely on some poor-man's-learning context-based recall instead. What else can't it learn? How is the LLM intern, the "drop-in replacement remote worker" going to do on day #2?
Instead of pretending that an LLM can be human-level, or super-human, or become generalist, why not just admit that this is not the final form of AI. An LLM is not an animal/human-like intelligence, it is something different - a language model, with it's own strengths and weaknesses.
Despite all the AGI hype, an LLM seems to have more in common with a pre-trained single-purpose system like AlphaGo than a brain, but with the rubric/reward-based policy function baked into the weights.
In another 10-20 years some new idea, hopefully more brain-like, will have superseded LLMs and they will indeed be labelled as "LLMs" as the AI/AGI label becomes attached to the new more brain-like creative intelligence. Perhaps it'll be sooner than 10-20 years, but I doubt it given the current 10-year fixation with LLMs which doesn't appear to be slowing down anytime soon. Perhaps Sutskever is working on something a bit different?
>How is text-based memorization going to let you learn non-linguistic skills
I don't know. How is Astra a step change in computer use and spatial reasoning to the extent it can play games, paint good looking stuff with e.g canva and a whole number of other things ?
>How is text-based memorization going to substitute for learning? The two are not the same.
Of course if you call it something else then you can say it's not the same.
>How is the LLM intern, the "drop-in replacement remote worker" going to do on day #2?
Just fine I imagine ? ICL and the memory tools around a harness are pretty good. I'm not sure what sort of magic you're expecting from the human, but they're not getting any improvement in that time frame a frozen transformer can't match.
>Instead of pretending that an LLM can be human-level, or super-human, or become generalist, why not just admit that this is not the final form of AI.
It doesn't seem like I'm the one pretending here.
>Despite all the AGI hype, an LLM seems to have more in common with a pre-trained single-purpose system like AlphaGo than a brain, but with the rubric/reward-based policy function baked into the weights.
If you say so.
>Perhaps it'll be sooner than 10-20 years, but I doubt it given the current 10-year fixation with LLMs which doesn't appear to be slowing down anytime soon.
The architecture that keeps delivering results isn't slowing down ? You don't say.
There's no shortage of people, even researchers, who for one reason or the other are convinced we are in need of some paradigm shift.
But guess what? Talk is cheap. You beat the current paradigm or you don't.
> How is Astra a step change in computer use and spatial reasoning to the extent it can play games and paint good looking stuff with e.g canva ?
Presumably because of pre-release training, because some alien outside of the model, armed with the reinforcement learning algorithm, came in and programmed its weights.
> The architecture that keeps delivering results isn't slowing down ? You don't say
Sure, nothing wrong with that, as long as you don't misrepresent the limitations of the approach.
> There's no shortage of people, even researchers, who for one reason or the other are convinced we are in need of some paradigm shift.
> But guess what? Talk is cheap. You beat the current paradigm or you don't. Do you seriously think that Meta never scaled JEPA ?
The idea has not been taken very far, so what is there to scale? It's not a complete cognitive architecture. So far it's also been using a pre-trained transformer as the learning component, which makes it of limited interest.
The animal intelligence approach, even in it's most fledgling form (that you would apparently dismiss), requires a complete agentic architecture, including new learning algorithms and generative behavior, before it can be compared to LLMs. We know that, done right, our brain architecture is more capable than an LLM, so even if any hypothetical attempts to reproduce it were not highly performant, we know that the idea itself is sound. You might compare with Uszkoreit's initial poor-performing implementation of his new language model architecture - should he have given up?
>Presumably because of pre-release training, because some alien outside of the model, armed with the reinforcement learning algorithm, came in and programmed its weights.
They didn't program anything. They gave it data at best.
>The idea has not been taken very far, so what is there to scale?
LeCun was the head of Meta AI for over a decade and his baby that he keeps harping on about wasn't taken very far ? Come on. You're smarter than that. It went the way all the alternate architectures have gone since the transformer, a sidegrade at best, probably not even that.
>You might compare with Uszkoreit's initial poor-performing implementation of his new language model architecture - should he have given up?
What are you talking about? There was no poor performing implementation of transformers that was published that he needed to push through. Are you talking about self attention experiments before the finished transformer? That's literally just research. And if he languished on that for a decade then yeah I'd tell him to probably look at something else, but of course he didn't.
You regard the LLM post-training process as "giving it data" ?!
Have you looked at all the published JEPA research both while LeCun was at Meta, and since (up to and including the latest AdaJEPA from June)? Please enlighten us as to exactly which line(s) of research you think were "scaled" at Meta, and then tell us which of these constituted anything even remotely resembling a complete testable intelligence?
FYI, it's been a long time since FaceBook/Meta even had a single head of AI. Since 2018 it has been split into two groups, FAIR and Generative AI, with LeCun being in the FAIR group, not as head, but as Chief AI scientist. LeCun only invented JEPA in 2022 (shortly after FaceBook became Meta), first writing about it in his "A Path Towards Autonomous Machine Intelligence" paper, perhaps unhappy with the work of the GenAI group, which he had no control over, that presumably was getting all the compute.
I was referring to Uszkoreit's personal telling (on YouTube) of the origin story of the Transformer, his motivations with the design, his initial personal failure to implement his idea in a performant enough manner to beat the current LSTM SOTA, and Noam Shazeer then throwing the kitchen sink at it and eventually coming up with the Transformer design.
>You regard the LLM post-training process as "giving it data" ?!
Yeah. Presumably, lots of synthetic data is being generated, experiments being run, but post-training is still a largely automated process.
>Have you looked at all the published JEPA research both while LeCun was at Meta, and since (up to and including the latest AdaJEPA from June)? Please enlighten us as to exactly which line(s) of research you think were "scaled" at Meta, and then tell us which of these constituted anything even remotely resembling a complete testable intelligence?
I have. My point isn't that his ideas are trash or that he should stop working on them. My point is it's not "gone very far" because he's taking it as far as he can, which isn't very far. He's not had a lack of influence, resources or will, either from his time at Meta or now with his billion dollar startup. He's had far more of it than most, if anything. That there's not much to show for it so far is not for a lack of trying.
>I was referring to Uszkoreit's personal telling (on YouTube) of the origin story of the Transformer, his motivations with the design, his initial personal failure to implement his idea in a performant enough manner to beat the current LSTM SOTA, and Noam Shazeer then throwing the kitchen sink at it and eventually coming up with the Transformer design.
So it's what I thought. This is just regular research unless an inordinate amount of time was spent on it and that's not the case.
> My point is it's not "gone very far" because he's taking it as far as he can, which isn't very far.
I wouldn't really agree - I'm no fan of LeCun, but the problem with JEPA isn't that it's a bad idea, or can't go very far, but just that it's not much of an idea in the first place!
It's no secret that our brain basically works by prediction, and what we're predicting is necessarily the external world as we perceive it though our own senses, aka latent representations, aka JEPA.
So, you COULD take this unoriginal smidgen of an idea and built it out to a full model of a human/animal brain, whether or not it's LeCun's intention to do so (he seems more interested in just the representational / world model aspect to it), but he certainly hasn't done so yet, nor created any research manifesto indicating that as his intent.
The fact that JEPA implementations to date are using pre-trained Transformers doesn't seem inherent to the approach - one could, with more effort, still predict latent representations (i.e. sensory feedback) but do so using a new real-time learning algorithm based on prediction failure.
LeCun seems more of an academic / research director than a builder, and I would never have put much stock in him being the one to build an animal brain.
Well I also don't think JEPA is a bad idea or anything. I don't think most of the alternative architectures or tweaks i've seen are bad ideas. On the contrary, some of them seem very cool and i'm all for interesting new ideas. I guess my opinion is more on the supposed necessity of it all.
I think the transformer is a powerful general learner. I think it's enough, and on the matter of intelligence, sometimes i think we make the mistake of zoning in too much on potentially spurious details. We still don't know much about Intelligence at the end of the day, what is and isn't really important, what is and isn't simply a detail of the environment and what kind of seemingly bizarre but surprisingly equivalent mechanisms can arise in the face of vastly different environments.
I think the distinction between these options probably doesn’t matter alll that much if the threshold you’re using is either “random person” or higher? Maybe bump it up to “random educated person”?
They are still different concepts of course, but I imagine that once one is achieved, the others aren’t far off.
Ok, but a random educated person will not perform well on vast majority of specialized tasks where professionals operate. A model like Astra probably will beat random educated person performance on majority of specialized tasks. It’s getting close to the level of professionals in many domains, and to genius level on some (e.g. math).
I’m just trying to understand the implications of the current frontier model capabilities.
We literally just saw how OpenAI’s model got out and hacked HuggingFace
reply