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

511–520 of 1001 posts

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

#511

Post by godelski »

naasking wrote: post #330
godelski wrote: post #224

Earlier quoted context omitted.

> that to understand knowledge you have to have a model of the world. You have a small but important mistake. It's to recite (or even apply ) knowledge. To understand does actually require a world model. Think of it this way: can you pass a test without understanding the test material? Certainly we all saw people we thought were idiots do well in class while we've also seen people we thought were geniuses fail. The t…

> It's to recite (or even apply) knowledge. To understand does actually require a world model. This is a shell game, or a god of the gaps. All you're saying is that the models "understand" how to recite or apply knowledge or language, but somehow don't understand knowledge or language. Well what else is there really?

  > Well what else is there really?
Differentiate from memorization.

I'd say there's a difference between a database and understanding. If they're the same, well I think Google created AGI a long time ago.

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

#512

Post by makotech221 »

scuppernong wrote: post #389
jb1991 wrote: post #58

I would bet all of my assets of my life that AGI will not be seen in the lifetime of anyone reading this message right now. That includes anyone reading this message long after the lives of those reading it on its post date have ended. Which of course raises the interesting question of how I can make good on this bet.

genuinely curious to hear your reasoning for why this is the case. i'm always somewhere between bemused and annoyed opening the daily HN thread about AGI and seeing everyone's totally unfounded confidence in their predictions. my position is I have no idea what is going to happen.

its incredibly stupid to believe general intelligence is just a series of computations that can be done by a computer. The stemlords on the west coast need to take philosophy classes.

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

#513

Post by tim333 »

pier25 wrote: post #362
oytis wrote: post #272

Earlier quoted context omitted.

Fusion is 30 years away, AGI is much closer

Fusion has been 30 years away since the 50s or 60s. I don't think I will see either AGI or commercial fusion in my lifetime.

Helion says it's 3 years away which is progress from saying it's 30 years away.

I remain a bit skeptical though. (https://www.innovationnewsnetwork.com/helion-breaks-ground-a...)

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

#514

Post by karmakaze »

rootusrootus wrote: post #268
jjulius wrote: post #247

Earlier quoted context omitted.

>Why is there a presumption that we (as people who have only studied CS) know enough about biology/neuroscience/evolution to make these comparisons? Hubris.

Exactly. Someone way back when decided to call them neural networks, and now a lot of people think that they are a good representation of the real thing. If we make them fast enough, powerful enough, we'll end up with a brain! Or not.

There was an actual simulation of a brain that could respond appropriately to stimuli. It ran many orders of magnitude slower than real-time but demonstrated the correlation. Probably not using the DNNs that we use now, but still a machine.

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

#518

Post by dang »

hax0ron3 wrote: post #355

If the transcript is accurate, Karpathy does not actually ever, in this interview, say that AGI is a decade away, or make any concrete claims about how far away AGI is. Patel's title is misleading.

Hmm good point. I skimmed the transcript looking for an accurate, representative quote that we could use in the title above. I couldn't exactly find one (within HN's 80 char limit), so I cobbled together "It will take a decade to get agents to work", which is at least closer to what Karpathy actually said.

If anyone can suggest a more accurate and representative title, we can change it again.

Edit: I thought of using "For now, autocomplete is my sweet spot", which has the advantage of being an exact quote; but it's probably not clear enough.

Edit 2: I changed it to "It will take a decade to work through the issues with agents" because that's closer to the transcript.

Anybody have a better idea? Help the cause of accuracy out here!

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

#519

Post by dang »

nextworddev wrote: post #497
hax0ron3 wrote: post #355

If the transcript is accurate, Karpathy does not actually ever, in this interview, say that AGI is a decade away, or make any concrete claims about how far away AGI is. Patel's title is misleading.

There's a lot of salt here

> Hey, podcast bro needs to get clicks

Please don't cross into personal attack. It's not what this site is for, and destroys what it is for.

Edit: please don't edit comments to change their meaning once someone has replied. It's unfair to repliers whose comments no longer make sense, and it's unfair to readers who can no longer understand the thread. It's fine, of course, to add to an existing comment in such a case, e.g. by saying "Edit:" or some such and then adding what else you want to say.

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

#520

Post by jhanschoo »

tyre wrote: post #226
ekjhgkejhgk wrote: post #134

Earlier quoted context omitted.

The interview which I've watched recently with Rich Sutton left me with the impression that AGI is not just a matter of adding more 9s. The interviewer had an idea that he took for granted: that to understand language you have to have a model of the world. LLMs seem to udnerstand language therefore they've trained a model of the world. Sutton rejected the premise immediately. He might be right in being skeptical here…

There is some evidence from Anthropic that LLMs do model the world. This paper[0] tracing their "thought" is fascinating. Basically an LLM translating across languages will "light up" (to use a rough fMRI equivalent) for the same concepts (e.g. bigness) across languages. It does have clusters of parameters that correlate with concepts, not just randomly "after X word tends to have Y word." Otherwise you would expect…

Let's make this more concrete than talking about "understanding knowledge". Oftentimes I want to know something that cannot feasibly be arrived at by reasoning, only empirically. Remaining within the language domain, LLMs get so much more useful when they can search the web for news, or your codebase to know how it is organized. Similarly, you need a robot that can interact with the world and reason from newly collected empirical data in order to answer these empirical questions, if the work had not already been done previously.
Post reply on HN