YouSaid · the spoken record

Andrej Karpathy

lines on the record: 266
first: 2022-10-29
most recent: 2022-10-29
sittings or episodes: 1
sources: podcast

Every line below is reproduced as it was said and linked to the record it came from. Nothing here is summarised or generated. Directory · Search · Corrections

“I do think many people will go with that, which I think is great. I love the idea that people can just choose their own adventure. Like you are born as a conscious free entity by default, I'd like to think. And you have your unalienable rights for life.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“Could be. I don't really buy all these ideas that, oh, without death, there's no meaning, there's nothing as... I don't intuitively buy all those arguments. I think there's plenty of meaning, plenty of things to learn. They're interesting, exciting. I want to know. I want to calculate. I want to improve the condition of all the humans and organisms that are alive.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“Of whoever wants to. Not many people might not want that. But I think people who do want that, I think it's probably possible. And I don't know that people fully realize this. I kind of feel like people think of death as an inevitability. But at the end of the day, this is a physical system. Some things go wrong. It makes sense why things like this happen evolutionarily speaking. And there's most certainly interventions that mitigate it.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“I think that's the question number one, practically speaking, because you can't, you're not going to calculate the answer to.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“Know that's the meaning of life. That's like meaning of life for some people. The first level answer, I would say, is anyone can choose their own meaning of life because we are conscious entity and it's beautiful. Number one. But I do think that like a deeper meaning of life if someone is interested along the lines of like what the hell is all this and like why. And if you look at fundamental physics and the quantum field theory and the standard model, they're very complicated. There's this 19 free parameters of our universe and like what's going on with all this stuff and why is it here and can I hack it? Can I work with it? Is there a message for me? Am I supposed to create a message? And so I think there's some fundamental answers there. But I think there's actually even like you can't actually really make dent in those without more time. And so to me also there's a big question around just getting more time, honestly. Yeah, that's kind of like what I think about quite a bit as well.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“Yeah, I think it's fascinating to me how these predictions of AI and what it's going to look like and what it's going to be capable of are completely inverted and wrong. And sci-fi of 50s and 60s was just like totally not right. They imagined AI as like super calculating theorem provers and we're getting things that can talk to you about emotions. They can do art. It's just like weird.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“And so that's going to be pretty crazy. That's going to almost certainly work. And it's going to be really interesting when the cost of content creation is going to fall to zero. You used to need a painter for a few months to paint a thing and now it's going to be speak to your phone to get your video.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“But yeah, there's always fun projects, basically. And stable diffusion also is opening up a huge amount of experimentation, I would say, in the visual realm and generating images and videos and movies, ultimately.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“Yeah, well, I definitely was surprised that transcription with OpenAI's Whisper was working so well compared to what I'm familiar with from Siri and a few other systems, I guess. It worked so well. That's what gave me some energy to try it out, and I thought it could be fun to run on podcasts. It's kind of not obvious to me why Whisper is so much better compared to anything else, because I feel like there should be a lot of incentive for a lot of companies to produce transcription systems and that they've done so over a long time. Whisper is not a super exotic model. It's a transformer. Takes mal spectrograms and just outputs tokens of text. It's not crazy. The model and everything has been around for a long time. Not actually 100% sure why this came out”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“Always like some fun side projects. Archive Sanity is one basically like there's way too many archive papers. How can I organize it? And recommend papers and so on. I transcribed all of your podcasts. What did you learn from?”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“For me, it is the ultimate meta problem. I don't want to work on any one specific problem. There's too many problems. So, how can you work on all problems simultaneously? You solve the meta problem, which to me is just intelligence. And how do you automate it?”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“Consistent thing I've been always interested in for me at least is AI That's probably what I'm spending the rest of my life on because I just care about it a lot. And I actually care about many other problems as well, like say aging, which I basically view as disease. And I care about that as well. But I don't think it's a good idea to go after it specifically. I don't actually think that humans will be able to come up with the answer. I think the correct thing to do is to ignore those problems and you solve AI and then use that to solve everything else. And I think there's a chance that this will work. I think it's a very high chance. And that's kind of like the way I'm betting at least.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“Yeah, being low level learning and then also communicating so that others can go through the same realizations and shortening that gap. Because I usually have to do way too much work to understand a thing. And then I'm like, okay, this is actually like, okay, I think I get it. And like, why was it so much work? Should have been much less work. And that gives me a lot of frustration, and that's why I sometimes go teach.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“Week doesn't converge, it's a problem. Yeah, you can like low pass filter yourself in terms of what has consistently been true for you. But yeah, I definitely see how it can be hard. But I would say you're going to work the hardest on a thing that you care about the most. So low pass filter yourself and really introspect in your past were the things that gave you energy and what are the things that took energy away from you, concrete examples. And usually from those concrete examples, sometimes parents can emerge. I like it when things look like this when I'm these positions.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“I think I'm very hesitant to give general advice. I think it's really hard. I've mentioned some of the stuff I've mentioned is fairly general, I think, like focus on just the amount of work you're spending on a thing. Compare yourself only to yourself, not to others.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“But certainly I've come across papers, I would say, in, say, like synthetic biology or something that I thought were quite readable for the abstract and the introduction. And then you're reading the rest of it and you don't fully understand, but you kind of are getting a gist, and I think it's cool.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“I think, yeah. Many times papers can be actually quite readable, not always, but sometimes the introduction in the abstract is readable, even for someone outside of the field. This is not always true. And sometimes I think, unfortunately, scientists use complex terms even when it's not necessary. I think that's harmful. I think there's no reason for that.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“Honestly, I find papers are quite good. I especially like the appendix of any paper as well. It's like the most detail you can have.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“What is the actual process of working with these cells and growing them and incubating them? And it's kind of like a massive cooking recipe. So making sure your cell slows and proliferate and then you're sequencing them, running experiments. Just how that works, I think, is kind of like the source of truth at the end of the day, what's really useful in terms of creating therapies and so on.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“And then I'm also suspicious of textbooks, honestly, because, as an example, in deep learning, there's no amazing textbooks and the field is changing very quickly. I imagine the same is true in, say, synthetic biology and so on. These books like The Cell are kind of outdated. They're still high level. Like what is the actual real source of truth? It's people in wet labs working with cells, sequencing genomes and... Actually, working with it. And I don't have that much exposure to that or what that looks like. So I still don't fully, I'm reading through the cell and it's kind of interesting and I'm learning, but it's still not sufficient, I would say, in terms of understanding.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“Okay, so selfish gene and the lane, I would say sometimes books are not sufficient. I like to reach for textbooks sometimes. I kind of feel like books are for too much of a general consumption sometime. And they just kind of like, they're too high up in the level of abstraction and it's not good enough. So I like textbooks. I like the cell. I think the cell was pretty cool. That's why I like the writing of Nick Lane is because he's pretty willing to step one level down and he doesn't, yeah, he's willing to go there. But he's also willing to sort of be throughout the stack, so he'll go down to a lot of detail, but then he will come back up. And I think basically I really appreciate that.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“Well, from the alternating point of view, like what makes me special is more like the gang of jeans that are writing in my chromosomes, I suppose, right? Like they're the replicating unit, I suppose.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“Many, of course. I think in biology, as an example, the vital question is a good one. Anything by Niclane, really life ascending, I would say, is like a bit more potentially representative as a summary of a lot of the things he's been talking about. I was very impacted by the selfish gene. I thought that was a really good book that helped me understand altruism as an example and where it comes from and just realizing that the selection isn't a level of genes was a huge insight for me at the time. And it sort of like cleared up a lot of things for me.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“I don't love when it sort of gets in the way of humanity in many ways. I like just people being humans in the way we sort of like slightly evolved and prefer, I think, just by default.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“Yeah, I think currently the virtual reality is not that compelling. I do think it can improve a lot, but I don't really know to what extent. Maybe there's actually even more exotic things you can think about with Neuralinks or stuff like that. Currently, I kind of see myself as mostly a team human person. I love nature. I love harmony. I love people. I love humanity. I love emotions of humanity. And I just want to be like in this like solar punk little utopia that's my happy place. Happy place is like people I love thinking about cool problems surrounded by a lush, beautiful, dynamic nature. And secretly high tech in places that count.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“Yeah, I think so. That seems to be how it's going right now. And I don't see that trend really reversing. I think people are diverse and they're able to choose their own path in existence. And I sort of celebrate that.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“I mean, to be clear, I think what's interesting about the future is it's not that I kind of feel like the variance in a human condition grows. That's the primary thing that's changing. It's not as much the mean. Distribution is like the variance of it, so there will probably be people on Mars, and there will be people in VR, and there will be people here on Earth. It's just like there will be so many more ways of being. And so I kind of feel like I see it as like a spreading out of a human experience.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“Eventually, I would once it's safe enough, but I don't actually know if it's on my lifetime scale unless I can extend it by a lot. Think that, for example, a lot of people might disappear into virtual realities and stuff like that. And I think that could be the major thrust of sort of the cultural development of humanity if it survives. So it might not be, it's just really hard to work in physical realm and go out there. And I think ultimately all your experiences are in your brain. And so it's much easier to disappear into digital realm. And I think people will find that more compelling, easier, safer. Interesting”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“So, by the way, I love the backup on Mars. I think that's amazing. You should absolutely do that. Yes. And I'm so thankful.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“Probably yes, but I don't know if it's dominant feature of future humanity. There might be some people on some planets and so on, but I'm not sure if it's like, yeah, if it's like a major player in our culture and so on.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“I'm going to have to, I do feel like I have to try to be optimistic and so on. I think even in this case, I still am predominantly optimistic, but there's definitely.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“Yeah, I think in general it was really weird about the dynamics of humanity and this explosion we talked about is just like the insane coupling afforded by technology And just the instability of the whole dynamical system. I think it just doesn't look good, honestly”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“Just it's a huge amount of power and then also with the proliferation. I basically don't actually really see what the good outcomes are here. So, I'm definitely worried about it a lot. And then AGI is not currently there, but I think at some point we'll more and more become something like it. The danger with AGI even is that I think it's even slightly worse in the sense that There are good outcomes of AGI. And then the bad outcomes are like an epsilon away, like a tiny one away. And so I think capitalism and humanity and so on will drive for the positive ways of using that technology. But then if bad outcomes are just like a tiny, like flip and minus sign away, that's a really bad position to be in it.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“I think so. And it's not even about the full destruction. To me, it's bad enough if we reset society. That would be like terrible. That would be really bad. And I can't believe we're so close to it. It's like so crazy to me.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“I mean, I'm definitely concerned about nuclear weapons and so on, not just as a result of the recent conflict even before that. That's probably my number one concern for humanity.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“Yes, that's my hope that it happens slowly enough in an open enough way where a lot of people can see and participate in it. Just figure out how to deal with this transition, I think, which is going to be interesting.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“I do worry I'm being 100% worry about it. And so the, I mean, the some of these fears of AGS and how this will plan out. I mean, these will be like very powerful entities probably at some point. And so for a long time, they're going to be tools in the hands of humans. People talk about alignment of AGIs and how to make the problem is like even humans are not aligned. How this will be used and what this is going to look like is troubling.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“Yeah, I just really like this tortured genius sort of character who's like grappling with whether or not he has like any responsibility or like what to do with this gift that he was given or like how to think about the whole thing.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“So many properties that make it like beautiful, interesting. So there's all these philosophical questions, but then there was also AGIs and there's simulation and it's cool and there's the black, you know.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“They laugh, it's funny, and you're showing you need a lot of knowledge to create humor about the, like you mentioned, human condition and so on. And then you need to be clever with it.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“To make you laugh, yeah. If it's able to be like a very effective stand up comedian that is doing something very interesting computationally, I think being funny is extremely hard.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“Yeah, I could imagine just having a conversation with this kind of oracle entity as you'd imagine them. And yeah, maybe it can tell you about, you know, based on my analysis of human condition, you might not want this. And here are some of the things that might...”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“As an example, it has a fixed amount of compute on its physical body, and it might just be that even though you can have a super amazing mega brain, super intelligent AI, you also can have less intelligent AIs that you can deploy in a power-efficient way. And then they're not perfect, they might make mistakes.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“Yeah, it's limited. And I think that's totally fine, by the way. I don't think it's fine and plausible to have a limited and imperfect AGIs.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“I love AI interstellar, by the way. I think it's such a sidekick to the entire story. At the same time, it's like really interesting”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“I would expect it to give solutions. I would expect it to be like, well, I've read all of these textbooks and I know all these things that you've produced. And it seems to me like here are the experiments that I think it would be useful to run next and here's some gene therapies that I think would be helpful and here are the kinds of experiments that you should run.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“Yeah, it's very fascinating. And we can also, of course, also shape it. It's not just a pure average. We can mess with the training data. We can mess with the objective. We can fine-tune them in various ways. So we have some impact on what those systems look like.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“And I think there's some canary in the coal mines, even with what we have today. And, you know, like, for example, there's these like waifus that you can work with and some people are trying to this company is going to shut down, but this person really like. Loved their waifu and is trying to port it somewhere else, and it's not possible. I think definitely people will have feelings towards these systems. Because in some sense, they are like a mirror of humanity because they are a big average of humanity. In a way that it's trained”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“Well, that's what the Lambda chatbot already told this Google engineer, right? It was talking about not wanting to die or so on.”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source
“I don't think it will emerge. I think it's going to be something very boring. We'll be talking to these digital AIs. They will claim their conscious. They will appear conscious. They will do all the things that you would expect of other humans. And it's going to just be a stalemate”
2022-10-29 · Lex Fridman Podcast · #333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI · IDENTIFIED FROM THE TRANSCRIPT · source