Andrej Karpathy — AGI is still a decade away
10/17/20252 hr 25 min
The Andrej Karpathy episode.During this interview, Andrej explains why reinforcement learning is terrible (but everything else is much worse), why AGI will just blend into the previous ~2.5 centuries of 2% GDP growth, why self driving took so long to crack, and what he sees as the future of education.It was a pleasure chatting with him.Watch on YouTube; listen on Apple Podcasts or Spotify.Sponsors* Labelbox helps you get data that is more detailed, more accurate, and higher signal than you could get by default, no matter your...
Clips
Transcript preview
First 90 secondsDwarkesh Patel· Host0:00
Today, I'm speaking with Andrej Karpathy. Andrej, why do you say that this will be the Decade of Agents and not the Year of Agents?
Andrej Karpathy· Guest0:06
Mm-hmm. Uh, well, first of all, uh, thank you for, uh, having me here. I'm, uh, excited to be here. So, the quote that you just mentioned, it's the Decade of Agents, that's actually a reaction to an existing- pre-existing quote, I should say, where I think a lot of th- some of the labs... I'm not actually sure who said this, but, uh, they were alluding to this being the Year of Agents- Hmm. ... uh, with respect to LLMs and, uh, how they were gonna evolve. And I think, um, I was triggered by that- (laughs) ... because I feel like there's some over-predictions going on in the industry.
Dwarkesh Patel· Host0:32
Yeah.
Andrej Karpathy· Guest0:32
And, uh, in my mind, this is really a lot more accurately described as the Decade of Agents.
Dwarkesh Patel· Host0:36
Yeah.
Andrej Karpathy· Guest0:36
And we have some very early agents that are actually, like, extremely impressive and that I use daily. Uh, you know, Claude and Codex and so on. But I still feel like there's, uh, so much work to be done. And so I think my, like, my reaction is, like, we'll be working with these things for a decade. They're gonna get better, uh, and, uh, it's gonna be wonderful. But I think I was just reacting to the timelines, I suppose, of the- of the, uh, implication.
Dwarkesh Patel· Host0:58
And w- what do you think will take a decade to accomplish?
Andrej Karpathy· Guest1:00
Yeah.
Dwarkesh Patel· Host1:00
What are the bottlenecks?
Andrej Karpathy· Guest1:02
Well, um, actually make it work.
Dwarkesh Patel· Host1:04
Hmm.
Andrej Karpathy· Guest1:04
So in my mind, I mean, when you're talking about an agent, I guess, or what the labs have in mind, and what maybe I have in mind as well, is it's, uh, you should think of it almost like an employee or like an intern that you would hire- Yeah. ... to work with you. Uh, so for example, you work with some employees here.
Dwarkesh Patel· Host1:15
Yeah.
Andrej Karpathy· Guest1:15
Um, when would you prefer to have an agent, like Claude or Codex, uh, do that work?
Dwarkesh Patel· Host1:19
Yeah.
Andrej Karpathy· Guest1:19
Like currently, of course, they can't. Uh, what would it take for them to be able to do that?
Dwarkesh Patel· Host1:23
Right.
Andrej Karpathy· Guest1:23
Why don't you do it today?
Dwarkesh Patel· Host1:24
Yeah.
Andrej Karpathy· Guest1:24
And the reason you don't do it today is because they just don't work, so- Right. ... uh, like, they don't have enough intelligence, they're not multimodal enough, they