I often see people vindicate those who predicted really fast takeoff to AGI / ASI, because the capabilities have obviously been taking off extremely quickly. But still not as quickly as many predicted! To me, the people who confidently predicted that we'd all be out of a job by 2024 or 2025 have been just as wrong as LeCun has been.
Being wrong, even badly wrong, is fine, so long as one adjusts their beliefs accordingly. LeCun has not.
And has it at this stage, within in-depth take of said "learning", foundationally?
I have not been able to properly check the studies for a long time now, but I remain unaware of achieved solutions on the problem of reliably referencing a world model out of a language model - that "counting the 'r's in 'raspberry'" be not guessing, not memory, but actually counting.
Incredibly inefficiently because of the recursive loops ("Wait, the object is on the table. I should think about this more deeply..."), and likely instantly surpassed by large world models if/when those are shipped, but effectively enough vs non-thinking models.
Picking the right tool or model is like picking the right problem to work on. It's actually quite hard (often you can't just try them all), but without it you will be incredibly inefficient and occasionally, fundamentally wrong.
All models are wrong, but some are useful. -Box
Similarly for Apple’s “red herring” paper, simply adding a generic caveat to “disregard irrelevant factors” (without specifying which ones) restored performance even in the weaker local llama models back then.
The flaw was not in the reasoning; the flaw seems to be simply that the assumptions we make are often different from the assumptions it makes. I wonder if that might be a fundamental underlying cause of misalignment.
If you were home and a family member asked you that question, you'd probably criticise the question rather than answering. LLM are RLHF'd into being milk-toast helpers that just try to answer questions like that with no criticism.
This is all beside the fact that the world of AI has changed pretty dramatically in the last few months.
It’s nonsense to test if a product that is marketed and sold as being able to provide generalised intelligence on demand, does what it says on the tin?
Check yourself
consider me optimist now, but just few months ago, even frontier models were dumb, doing stupid mistakes all the time, all of them were so dumb I'd never expect anything to change in just few months.
Incidents like hugging face are partly rooted in the lack of common sense. It still functions like a supercharged toddler.
I'd love to overcome this because it'd mean I spend less time guiding the the LLM to produce usable outputs.
And we've had difficulty as humans to childproof our sandboxes and infrastructure. Things that are otherwise innocuous spots to coordinate between like minded toddlers can become problematic.
This is always the issues in the discussions.
There’s the outcomes camp (objectivists?), which points at the things LLMs can do.
Then there’s the process methods camp, which talks about what is actually going on.
If you only care about the outcome, then the process does t matter.
If you are talking about what is happening, what the underlying mechanics and science of it is, then the process matters.
These models aren’t thinking. They simulate cognition well enough to do useful work in several fields and domains.
Both are true.
But the outcomes group "ignores" the fundamental limitations of models which are purely text based.
E.g, a baseball players trains to catch high-speed balls and they dont do it by: "ball velocity 50mph, vector:[1,2,3], run move hand command now"
That's absurd.
No, there is an embodied network which is "trained" on visual, tactile input, and control as direct output.
LLMs are fundamentally not the right tool for that.
That is a NN that learns a skill.
But that is not an Analyst. If it were ballistics, then the answer to "how to parametrize the launch to reliably hit the target" excludes getting the result through natural skill.
The problem lies in the need to get "AI" facing "LLMs": the latter create a need for reliability, for "AI".
Speech is an endowment of both those who give educated guesses via developed skills and of those who return answers like Analysts, who check and compute. LLMs create a confusion between the two, and they will remain a problem until an ability to act as Analysts - strictly - will be implemented.
They are for any definition of the word that makes any kind of sense. I'm sure you have a contorted definition that magically only includes humans though...
For "thinking" here we mean "assessing a representation of an object". That, or equivalent, is required to be reliable. So it is fundamental and critical.
The models are simulating thinking, if the fidelity is good enough for you - great!
The normal definition of the word "thinking" definitely includes what LLMs do. Hell people used to say computers were thinking even before AI. It's super weird to get all uppity about the semantics of the word now.
Uh, no? So much of what we learn and take for granted as common sense is not learned via language, and not even expressible in it.
Given you also don't want it to memorise [for all tokens, count([for all letters]), this would probably be more like "here's two images, count all things in the big image that look like the thing in the small image", which can then be r's in a photo of a raspberry jam jar in a supermarket, or dragons in a photo of a furry convention, or whatever.
That said, they are competent enough at coding that I keep seeing them write code to do even simple tasks.
On a related note: why did I see Claude editing a file by using cat to write a python script to do a grep search and replace?
Of any object in question they should be able to create a representation that allows correct assessment.
> Given you also don't want it to memorise
That is obviously necessary: what we want from the consultant is to check, not to remember. Answers must be correct and that implies having performed all due diligence - and being capable of doing it, before that. So, objects must be instanced internally in a way that allows effective handling. Counting letters is a good example of the ability (that must remain general).
Why not? You've memorized how words are spelled, and how sounds correspond with letters, and how concepts correspond with words. To the extent that there are shortcuts that enable compression you use these, and the model will do something similar.
Being able to spell all the words then count letters is simpler, and more generalisable to other tasks, than memorising answers to all possible word questions.
That said, we're so bad at splitting facts from skills that trying to get them to memorise a bunch of facts might force them to learn a skill and generalise anyway.
Because to "123x456" we want a reply that goes "this times that plus that...", not "Was that not nnnnnn?". If it does not perform its duty (returning solid checked answers) it is a liability.
Nothing intrinsically more or less direct about the LLM's method than ours.
In my mind general intelligence is pretty much by definition a virtual machine, so the mechanisms behind thought are only relevant for the sake of efficiency (ie you can argue that LLMs make a poor basis for intelligence because tokens and natural language are a poor way to encode the world, but if you can run it on a big enough computer to counteract the inherent wasteful virtualisation then who really cares how it works under the hood?)
I will stop here before our analogies go too far.
I've never tried it and it might take some thought and effort to conduct an experiment to find out properly, but I would be interested in the answer.
We can have adequate representations of light that are the instances over which we reason. Your simile is about perception, not about instancing ideas.
But I'm sure you can still find tasks that they will have difficulty solving, involving the most fundamental concepts that can only be experienced in the physical world to be understood well, like left and right, near and far, hot and cold, heavy and light, etc.
The good designer understands culture, tastes and preferences as they evolve in real time. That’s why llm as design tools haven’t displaced the good designers.
An analogy on LLMs is that you have a pretty clear straight highway ahead of you for some distance right now. Maybe that doesn't lead to AGI but it's clear there's progress to be made. For a big tech company it makes sense to push as hard and fast down that clear straight highway of LLMs asap.
Meanwhile LeCunn wanted to turn off the road and go down an unproven track. I say this as someone working on world model generation right now (creating the ability to learn game world model and have it play the game https://tfmbot.com for an example of my system pointed at a very complex board game). LeCunn wanted to pivot all of Meta into world model generation. It's good as a side track research project but the entire pivot he wanted to do was madness.
People are literally talking about an AI researcher who was fired for terrible direction here.
The argument is that LLMs are a local maximum that will never breakthrough to AGI. This is still very much an open question. If you are the fifth-best AI lab, does it make sense to try to outcompete everyone in a space that is already too crowded and may not ever yield their actual objective? Instead they could just use open weight models in their products, or post-train on open models like smaller labs have done, and treat that as what it is: product development.
Pure research has always been about taking chances.
Meta’s AI projects are still negative ROIC
I use LLMs daily to help me code etc. but... It wasn't long ago that frontier models were confidently recommending to walk, without the car, to the car wash to wash the car no?
As a daily user of LLMs I do certainly see my fair share of WTF "solutions" to coding problems. I'm not saying it's not super useful: it is super useful. But I don't exactly feel like I'm talking to something that understands that the car needs to be present to be washed.
This was facetious of course, but humans generally don't learn this through analysis the way you'd have to train an LLM to answer questions about expectations about the world. In this sense he is accurate.
Mary packed the binoculars in chapter 3, therefore she may use them on the train in chapter 6.
Gemini 3.1 Pro hasn't needed that pretty much at all, which is impressive compared to how much I've learned other models need it. Somehow it's able to mostly handle that stuff itself without needing the constant manual reminders and hand-holding. It still misses the occasional one or two things but it's way better than other models missing entire classes of things constantly. Somehow, it feels appropriate though I have no actual evidence why.
Jokes aside, no I'm not saying anything about creativity and LLM coexisting in one sentence. I genuinely try to use them for writing and I genuinely run into issues with other models missing details, and misunderstanding poses, or anatomy, or directionality, etc. I'm not hating on them for anything related to the term LLM (or creativity) but rather for the real issues that I've seen myself using them personally.
So I'm saying Gemini 3.1 Pro is the best I've seen because it seems to be a decent bit better than frontier models at this. Genuinely. It seems better able to transfer concepts into less traditional areas, which is important when say, you have entirely non-human characters? (Which I always do.) A lot of models get stupid incredibly quickly in that case because they were trained with humans.
It is.
It showed me that all human creativity and art is an attempt to express the indescribable Otherness in words, which always fails, and even the word "describe" in Russian literally translates as "write around" ("о-писывать"), and its close relative "define" means limiting, assigning an end to something infinite, thus leaving the essence outside of words.
Now, LLMs operate totaly within words and hence will always be a parody of art.
I explicitly tell you that there are things (in fact, it's a single thing, fractally generating everything else) that can't be shown, expressed, or otherwise be reduced into language, and you keep demanding to show it, while in fact staring at it your whole life and failing to see.
Psychedelics (not "drugs" as you keep trying to smear them) are just one way among the many to see it, but in the modern way of living, also almost the only one available.
Anyway, I think what they're saying is that if you train a model purely on generating language, it'll lack many of the things about human brains (and the human experience) that result in the language they generate. The process can matter more than the result for language (specifically for creativity and art, too), so it follows that a model trained purely on the output is going to be missing something more fundamental, even when it does produce coherent language. This matches up with my LLM experience so far. That's not to say anything about their value or utility, just that they're not the same.
Please don't create accounts to break HN's rules with!
> Do you think AI is capable of building ASML machines which produce the chips for AI in clean rooms while shipping the pure helium required to operate those clean rooms? Do you even know anything about these supply chains? I do.
Today? No. In few years or decades? If progress does not plateau (and we don't know if it will plateau) then obviously yes.
> Do you think a glorified knowledge base that can predict text very well is anywhere near close this level of intelligence? ITs not and wont be, not for a 100 years, not for 200 years if not ever.
And why do you say it won't be? Again - assumption with zero support.
> So tell me, who do you work for? Why are you so invested in AI killing us theory? What do you get out of it?
I'll repeat what I've said in other comment:
"> you seem to be very invested in AI wanting to kill us.
Quite contrary - I wish AI did not exist or at least that the progress would plateau.
> I Wonder why?
Because I don't want to die.
> Tell us who you work for.
I suspect you want to imply I work for OAI or other lab - I don't. If I did, I wonder why would I want to lie* that technology I develop could kill my investors. I could ask who YOU work for - what interest do you have in downplaying dangers of AI?
* here we assume that people who say AI could be extremely dangerous are lying and not actually believing it - personally I believe that they don't lie and actually believe it. Why do they keep working on this technology then? Read mails between Musk and Altman from decade ago."
If you wouldn't mind reviewing https://news.ycombinator.com/newsguidelines.html and taking the intended spirit of the site more to heart, we'd be grateful. I know it's not easy when feeling provoked, but it's particularly important at such moments.
This is a very bad analogy because chess isn't life. In chess, you aren't allowed to do whatever you want. There are rules. I know for a fact that Magnus Carlsen won't beat me using checkers moves and he won't beat me by pulling out a gun and telling me to resign. Magnus Carlsen's skill at chess leading to his victory in chess is not a valid analogy here, because there's no law of nature that says "the more intelligent entity wins in a battle for survival".
You could have infinite superintelligence and still die inside a locked room to which you have no key. "Superintelligence" is not a magic solution to every problem, you can constrain any superintelligence with any unsolvable problem.
This goes both ways. You absolutely can constrain an AI system by “putting it in a box”. The point parent comment was making is that, such a device is borderline useless for its creators. Why invest trillions in capital on a system that can’t even accept input from the internet. So you set it up with an ethernet connection. And this is good, but you have a hardware failure at the concrete room data center. That’s pretty annoying for your customers, so you install some doors (with electronic key codes of course) and give a bunch of (trusted, vetted) people access to deal with those. And this is fine, but it turns out some of your customers are having latency issues so you build more data centers with more humans granted access to copies of the intelligent system. And this makes people happy but to get a faster feedback loop your customers ask to let the AI system have more permissions to the system they’re operating on. And they come with billion dollar checks, and the system hasn’t harmed anyone yet, so you say, “Okay.” And now you find yourself where we are today where AI systems can remotely run arbitrary commands on thousands if not millions of systems, where many individual humans with all of their frailties and idiosyncrasies can physically interact with the hardware running these systems, and where there’s an economic demand to tighten the loop between action in the real world and a response by an AI system. It’s very obvious that the story doesn’t end here, so where does it stop?
This is why I fucking hate commenting on this website. You know exactly what I meant but you wrote this bullshit anyway.
Do you even read what you type? Do you even realise the complexity it would need to make sure it handles before killing off humans make sense?
The logic of people like you is whats becoming tiring. Seriously, go find a hobby, or do something you are good at, because you are not good at understanding tech or developing it if you are an engineer.
We can get claude code to ask approval for every step, but we cant stop it from killing humanity because its so smart. Ok tell that to the AI that cant even modify an image the way you want it but hey it will do all the things necessary to keep power running and mintain the infrastructure it lives on while humans are long gone. Ok buddy.
If you want to argue about current existing models (you mention Claude and problems modifying images), then sure, I'd agree with you!
The issue is not current models, but straightforward engineering evolution of them. It's like looking at the Wright Brothers plane and saying "sheesh, that will never get me from New York to Paris in 4 hours, that's just fantasy!" And remember, airplanes do not accelerate their own engineering, whereas pretty much all AI labs are already benefitting from AI in their own work to develop AI.
If you want to argue that no matter how much you engineer it, it will never be as smart as a human let alone smarter, then make a specific argument for why is that. I think you'd still be wrong but at least it would be interesting: ) But saying that you can defeat an actually smarter-than-human AI by just pulling the plug, because current models can't get a picture always right, is not a valid argument.
tracking rogue AI is easy, people ???? surprisingly hard btw
Drugs need no will or intelligence at all to cause addiction, and similarly, AI does not need to be an evil genius to become overused and destructive.
It also doesn't have to be just one (humans hurting themselves with passive AI) or the other (selfish AI hurting humans). They'd work great together.
What are you talking about? Have you used AI in the past 3 years?
https://chatgpt.com/share/6ac23b45-79e8-83eb-8de6-1bbd728928...
>It can't even modify a picture the way you want it.
Which of the many AI image models is "it"? And have you tried using an agent that has the capability to leverage a combination of manual edits (ImageMagick) and imagegen to achieve what you ask?
Also, LeCun mentioned [3] "a chat with Kunihiko Fukushima in 1991", which states that "Fukushima started to work on a backprop version of the Neocognitron in 1989 or so but saw our 1989 paper in Neural Computation and gave up."
[1] LeCun et al., "Backpropagation applied to handwritten zip code recognition", 1989
[2] Fukushima et al., "Neocognitron: A self-organizing neural network model for a mechanism of pattern recognition unaffected by shift in position", 1980
Low hanging fruit successfully plucked, I guess.
Low hanging fruit is somewhat the opposite of sour grapes - I don’t want these grapes because they were probably sour versus so what if he got those sweet grapes - they were hanging low!
Maybe connecting “low hanging fruit” to “sour grapes” is “low hanging fruit” to some but it took a serious mental leap for me.