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

Andrej Karpathy – It will take a decade to work through the issues with agents

dwarkesh.com

Post reply on HN
Print view View original

571–580 of 1001 posts

Re: Andrej Karpathy – It will take a decade to work through the issues with agents

#571

Post by pastel8739 »

ajross wrote: post #293
SR2Z wrote: post #246

Earlier quoted context omitted.

Right, but modeling the structure of language is a question of modeling word order and binding affinities. It's the Chinese Room thought experiment - can you get away with a form of "understanding" which is fundamentally incomplete but still produces reasonable outputs? Language in itself attempts to model the world and the processes by which it changes. Knowing which parts-of-speech about sunrises appear together an…

> Knowing which parts-of-speech about sunrises appear together and where is not the same as understanding a sunrise What does "understanding a sunrise" mean though? Arguments like this end up resting on semantics or tautology, 100% of the time. Arguments of the form "what AI is really doing" likewise fail because we don't know what real brains are "really" doing either . I mean, if we knew how to model human language…

Is it really so rare? I feel like I know of tons of fields where we have methods that work empirically but don’t understand all the theory. I’d actually argue that we don’t know what’s “actually” happening _ever_, but only have built enough understanding to do useful things.

Re: Andrej Karpathy – It will take a decade to work through the issues with agents

#572

Post by judahmeek »

__MatrixMan__ wrote: post #523
joshellington wrote: post #473

To throw two pennies in the ocean of this comment section - I’d argue we still lack schematic-level understanding of what “intelligence” even is or how it works. Not to mention how it interfaces with “consciousness”, and their likely relation to each other. Which kinda invalidates a lot of predictions/discussions of “AGI” or even in general “AI”. How can one identify Artificial Intelligence/AGI without a modicum of u…

I don't think we can ever know that we are generally intelligent. We can be unsure, or we can meet something else which possesses a type of intelligence that we don't, and then we'll know that our intelligence is specific and not general. So to make predictions about general intelligence is just crazy. And yeah yeah I know that OpenAI defines it as the ability to do all economically relevant tasks, but that's an awfu…

All intelligence is specific, as evidenced by the fact that a universal definition regarding the specifics of "common sense" doesn't exist.

Re: Andrej Karpathy – It will take a decade to work through the issues with agents

#573

Post by BoorishBears »

echoangle wrote: post #499
BoorishBears wrote: post #438

Earlier quoted context omitted.

what about the fact frontier labs are spending more compute on viral AI video slop and soon-to-be-obsoleted workplace usecases than research? Even if you don't understand the technicals, surely you understand if any party was on the verge of AGI they wouldn't behave as these companies behave?

What does that tell you about AI in 100 years though? We could have another AI winter and then a breakthrough and maybe the same cycle a few times more and could still somehow get AGI at the end. I’m not saying it’s likely but you can’t predict the far future from current companies.

You're making the mistake of assuming the failure of the current companies would be seperated from the failures of AI as a technology.

If we continue the regime where OpenAI gets paid to buy GPUs and they fail, we'll have a funding winter regardless of AI's progress.

I think there is a strong bull case for consumer AI but it looks nothing like AGI, and we're increasingly pricing in AGI-like advancements.

Re: Andrej Karpathy – It will take a decade to work through the issues with agents

#575

Post by kllrnohj »

aik wrote: post #540
eddiewithzato wrote: post #532

Earlier quoted context omitted.

Because that's the definition that is leading to all these investments, the promise that very soon they will reach it. If Altman said plainly that LLMs will never reach that stage, there would be a lot less investment into the industry.

Hard disagree. You don’t need AGI to transform countless workflows within companies, current LLMs can do it. A lot of the current investments are to help with the demand with current generation LLMs (and use cases we know will keep opening up with incremental improvements). Are you aware of how intensely all the main companies that host leading models (azure, aws, etc) are throttling usage due to not enough data cent…

> Eg. At my company we have 100x more demand than we can get capacity for, and we’re barely getting started. We have a roadmap with 1000x+ the current demand and we’re a relatively small company.

OpenAI's revenue is $13bn with 70% of that coming from people just spending $20/mo to talk to ChatGPT. Anthropic is projecting $9bn in revenue in 2025. For nice cold splash of reality, fucking Arizona Iced Tea has $3bn in revenue (also that's actual revenue not ARR)

You might have 100x more demand than you can get capacity for, but if that 100x still puts you at a number that in absolute terms is small, it's not very impressive. Similarly if you're already not profitable and achieving 100x growth requires 1,000x in spend, that's also not a recipe for success. In fact it's a recipe for going bankrupt in a hurry.

Re: Andrej Karpathy – It will take a decade to work through the issues with agents

#576

Post by pastel8739 »

robotresearcher wrote: post #459
godelski wrote: post #383

Earlier quoted context omitted.

> that does not have a technical meaning I don't think the definition is very refined, but I think we should be careful to differentiate that from useless or meaningless. I would say most definitions are accurate, but not precise. It's a hard problem, but we are making progress on it. We will probably get there, but it's going to end up being very nuanced and already it is important to recognize that the word means d…

> It's a hard problem So people say. I’m not sidestepping the Hard Problem. I am denying it head on. It’s not a trick or a dodge! It’s a considered stance. I'm denying that an idea that has historically resisted crisp definition, and that the Stanford Encyclopedia of Philosophy introduces as 'protean', needs to be taken seriously as an essential missing part of AI systems, until someone can explain why. In my view, t…

Regardless of whether you think understanding is important, it’s clear from this thread that a lot of people find understanding valuable. In order to trust an AI with decisions that affect people, people will want to believe that the AI “understands” the implications of its decisions, for whatever meaning of “understand” those people have in their head. So indeed I think it is important that AI researchers try to get their AIs to understand things, because it is important to the consumers that they do.

Re: Andrej Karpathy – It will take a decade to work through the issues with agents

#577

Post by danielmarkbruce »

DanHulton wrote: post #570
Imnimo wrote: post #68

>What takes the long amount of time and the way to think about it is that it’s a march of nines. Every single nine is a constant amount of work. Every single nine is the same amount of work. When you get a demo and something works 90% of the time, that’s just the first nine. Then you need the second nine, a third nine, a fourth nine, a fifth nine. While I was at Tesla for five years or so, we went through maybe three…

The thing about this, though - cars have been built before. We understand what's necessary to get those 9s. I'm sure there were some new problems that had to be solved along the way, but fundamentally, "build good car" is known to be achievable, so the process of "adding 9s" there makes sense. But this method of AI is still pretty new, and we don't know it's upper limits. It may be that there are no more 9s to add, o…

While you are right about the broader (and sort of ill defined) chase toward 'AGI' - another way to look at it is the self driving car - they got there eventually.And, if you work on applications using LLMs you can pretty easily see that Karpathy's sentiment is likely correct. You see it because you do it. Even simple applications are shaped like this, albeit each 9 takes less time than self driving cars for a simple app.. it still feels about right.

Re: Andrej Karpathy – It will take a decade to work through the issues with agents

#578

Post by hashiyakshmi »

klipt wrote: post #563
ziofill wrote: post #545

Earlier quoted context omitted.

I agree with this. A metaphor I like is that the reason why humans say the night sky is beautiful is because they see that it is, whereas an LLM says it because it’s been said enough times in its training data.

What about a blind human? Are they just like an LLM? What about a multimodal model trained on video? Is that like a human?

This is actually a great point but for the opposite reason - if you ask a blind person if the night sky is beautiful, they would say they don't know because they've never seen it (they might add that they've heard other people describe it as such). Meanwhile, I just asked ChatGPT "Do you think the night sky is beautiful?" And it responded "Yes, I do..." and went on to explain why while describing senses its incapable of experiencing.

Re: Andrej Karpathy – It will take a decade to work through the issues with agents

#579

Post by hyperadvanced »

kllrnohj wrote: post #575
aik wrote: post #540

Earlier quoted context omitted.

Hard disagree. You don’t need AGI to transform countless workflows within companies, current LLMs can do it. A lot of the current investments are to help with the demand with current generation LLMs (and use cases we know will keep opening up with incremental improvements). Are you aware of how intensely all the main companies that host leading models (azure, aws, etc) are throttling usage due to not enough data cent…

> Eg. At my company we have 100x more demand than we can get capacity for, and we’re barely getting started. We have a roadmap with 1000x+ the current demand and we’re a relatively small company. OpenAI's revenue is $13bn with 70% of that coming from people just spending $20/mo to talk to ChatGPT. Anthropic is projecting $9bn in revenue in 2025. For nice cold splash of reality, fucking Arizona Iced Tea has $3bn in re…

This is correct, it should burn the retinas of anyone thinking that OAI or Anthropic are in any way worth their multi-billion dollar valuations. I liked AK’s analysis of AI for coding here (it’s overly defensive, lacks style and functionality awareness, is a cargo cultist, and/or just does it wrong a lot) but autocomplete itself is super valuable, as is the ability to generate simple frontend code and let you solve the problem of making a user interface without needing a team of people with those in-house skills.
Post reply on HN