episode stringlengths 45 100 | text stringlengths 1 528 | timestamp_link stringlengths 56 56 |
|---|---|---|
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | reinforcement learning and how you think about it from the fifties to now? | https://karpathy.ai/lexicap/0015-large.html#00:09:18.960 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | One thing is that it's oscillates, right? So things become fashionable and then they go out | https://karpathy.ai/lexicap/0015-large.html#00:09:23.600 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | and then something else becomes cool and that goes out and so on. And I think there's, so there's | https://karpathy.ai/lexicap/0015-large.html#00:09:29.360 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | some interesting sociological process that actually drives a lot of what's going on. | https://karpathy.ai/lexicap/0015-large.html#00:09:33.680 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | Early days was kind of cybernetics and control, right? And the idea that of homeostasis, | https://karpathy.ai/lexicap/0015-large.html#00:09:38.880 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | right? People have made these robots that could, I don't know, try to plug into the wall when they | https://karpathy.ai/lexicap/0015-large.html#00:09:46.240 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | needed power and then come loose and roll around and do stuff. And then I think over time, the | https://karpathy.ai/lexicap/0015-large.html#00:09:51.680 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | thought, well, that was inspiring, but people said, no, no, no, we want to get maybe closer to what | https://karpathy.ai/lexicap/0015-large.html#00:09:59.200 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | feels like real intelligence or human intelligence. And then maybe the expert systems people tried | https://karpathy.ai/lexicap/0015-large.html#00:10:03.200 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | to do that, but maybe a little too superficially, right? So, oh, we get the surface understanding of | https://karpathy.ai/lexicap/0015-large.html#00:10:10.160 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | what intelligence is like, because I understand how a steel mill works and I can try to explain | https://karpathy.ai/lexicap/0015-large.html#00:10:20.240 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | it to you and you can write it down in logic and then we can make a computer and for that. | https://karpathy.ai/lexicap/0015-large.html#00:10:24.960 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | And then that didn't work out. But what's interesting, I think, is when a thing starts to not | https://karpathy.ai/lexicap/0015-large.html#00:10:29.200 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | be working very well, it's not only do we change methods, we change problems, right? So it's not | https://karpathy.ai/lexicap/0015-large.html#00:10:36.400 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | like we have better ways of doing the problem of the expert systems people were trying to do. We | https://karpathy.ai/lexicap/0015-large.html#00:10:43.200 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | have no ways of trying to do that problem. Oh, yeah, no, I think maybe a few, but we kind of | https://karpathy.ai/lexicap/0015-large.html#00:10:47.040 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | give up on that problem and we switched to a different problem and we worked that for a while | https://karpathy.ai/lexicap/0015-large.html#00:10:54.880 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | and we make progress. As a broad community. As a community, yeah. And there's a lot of people who | https://karpathy.ai/lexicap/0015-large.html#00:11:00.720 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | would argue, you don't give up on the problem, it's just you decrease the number of people working | https://karpathy.ai/lexicap/0015-large.html#00:11:04.320 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | on it. You almost kind of like put it on the shelf, say, we'll come back to this 20 years later. | https://karpathy.ai/lexicap/0015-large.html#00:11:09.520 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | Yeah, I think that's right. Or you might decide that it's malformed. Like you might say, | https://karpathy.ai/lexicap/0015-large.html#00:11:13.920 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | it's wrong to just try to make something that does superficial symbolic reasoning | https://karpathy.ai/lexicap/0015-large.html#00:11:21.600 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | behave like a doctor. You can't do that until you've had the sensory motor experience of being | https://karpathy.ai/lexicap/0015-large.html#00:11:26.160 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | a doctor or something. So there's arguments that say that that problem was not well formed. Or it | https://karpathy.ai/lexicap/0015-large.html#00:11:33.200 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | could be that it is well formed, but we just weren't approaching it well. So you mentioned | https://karpathy.ai/lexicap/0015-large.html#00:11:38.320 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | that your favorite part of logic and symbolic systems is that they give short names for large | https://karpathy.ai/lexicap/0015-large.html#00:11:43.760 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | sets. So there is some use to this. They use symbolic reasoning. So looking at expert systems | https://karpathy.ai/lexicap/0015-large.html#00:11:48.800 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | and symbolic computing, what do you think are the roadblocks that were hit in the 80s and 90s? | https://karpathy.ai/lexicap/0015-large.html#00:11:56.960 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | Ah, okay. So right. So the fact that I'm not a fan of expert systems doesn't mean that I'm not a | https://karpathy.ai/lexicap/0015-large.html#00:12:01.680 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | fan of some kinds of symbolic reasoning, right? So let's see, roadblocks. Well, the main road | https://karpathy.ai/lexicap/0015-large.html#00:12:07.920 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | block, I think, was that the idea that humans could articulate their knowledge effectively | https://karpathy.ai/lexicap/0015-large.html#00:12:15.520 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | into some kind of logical statements. | https://karpathy.ai/lexicap/0015-large.html#00:12:22.080 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | So it's not just the cost, the effort, but really just the capability of doing it. | https://karpathy.ai/lexicap/0015-large.html#00:12:26.240 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | Right. Because we're all experts in vision, right? But totally don't have introspective | https://karpathy.ai/lexicap/0015-large.html#00:12:31.280 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | access into how we do that. Right. And it's true that, I mean, I think the idea was, well, | https://karpathy.ai/lexicap/0015-large.html#00:12:36.720 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | of course, even people then would know, of course, I wouldn't ask you to please write | https://karpathy.ai/lexicap/0015-large.html#00:12:44.960 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | down the rules that you use for recognizing a water bottle. That's crazy. And everyone | https://karpathy.ai/lexicap/0015-large.html#00:12:48.240 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | understood that. But we might ask you to please write down the rules you use for deciding, | https://karpathy.ai/lexicap/0015-large.html#00:12:52.800 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | I don't know, what tie to put on or how to set up a microphone or something like that. | https://karpathy.ai/lexicap/0015-large.html#00:12:58.640 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | But even those things, I think people maybe, I think what they found, I'm not sure about | https://karpathy.ai/lexicap/0015-large.html#00:13:04.640 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | this, but I think what they found was that the so called experts could give explanations | https://karpathy.ai/lexicap/0015-large.html#00:13:10.880 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | that sort of post hoc explanations for how and why they did things, but they weren't | https://karpathy.ai/lexicap/0015-large.html#00:13:16.000 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | necessarily very good. And then they depended on maybe some kinds of perceptual things, | https://karpathy.ai/lexicap/0015-large.html#00:13:20.080 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | which again, they couldn't really define very well. So I think fundamentally, I think the | https://karpathy.ai/lexicap/0015-large.html#00:13:28.960 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | underlying problem with that was the assumption that people could articulate how and why they | https://karpathy.ai/lexicap/0015-large.html#00:13:35.840 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | make their decisions. Right. So it's almost encoding the knowledge | https://karpathy.ai/lexicap/0015-large.html#00:13:40.960 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | from converting from expert to something that a machine could understand and reason with. | https://karpathy.ai/lexicap/0015-large.html#00:13:45.680 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | No, no, no, no, not even just encoding, but getting it out of you. | https://karpathy.ai/lexicap/0015-large.html#00:13:51.440 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | Right. Not, not, not writing it. I mean, yes, hard also to write it down for the computer, | https://karpathy.ai/lexicap/0015-large.html#00:13:56.320 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | but I don't think that people can produce it. You can tell me a story about why you do stuff, | https://karpathy.ai/lexicap/0015-large.html#00:14:02.240 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | but I'm not so sure that's the why. Great. So there are still on the | https://karpathy.ai/lexicap/0015-large.html#00:14:08.320 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | hierarchical planning side, places where symbolic reasoning is very useful. So as you've talked | https://karpathy.ai/lexicap/0015-large.html#00:14:14.400 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | about, so where's the gap? Yeah. Okay, good. So saying that humans can't provide a description | https://karpathy.ai/lexicap/0015-large.html#00:14:24.560 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | of their reasoning processes. That's okay. Fine. But that doesn't mean that it's not good to do | https://karpathy.ai/lexicap/0015-large.html#00:14:34.960 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | reasoning of various styles inside a computer. Those are just two orthogonal points. So then | https://karpathy.ai/lexicap/0015-large.html#00:14:41.040 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | the question is what kind of reasoning should you do inside a computer? Right. | https://karpathy.ai/lexicap/0015-large.html#00:14:47.120 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | And the answer is, I think you need to do all different kinds of reasoning inside a computer, | https://karpathy.ai/lexicap/0015-large.html#00:14:52.240 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | depending on what kinds of problems you face. I guess the question is what kind of things can you | https://karpathy.ai/lexicap/0015-large.html#00:14:56.240 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | encode symbolically so you can reason about? I think the idea about, and even symbolic, | https://karpathy.ai/lexicap/0015-large.html#00:15:02.400 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | I don't even like that terminology because I don't know what it means technically and formally. | https://karpathy.ai/lexicap/0015-large.html#00:15:12.480 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | I do believe in abstractions. So abstractions are critical, right? You cannot reason at completely | https://karpathy.ai/lexicap/0015-large.html#00:15:17.760 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | fine grain about everything in your life, right? You can't make a plan at the level of images and | https://karpathy.ai/lexicap/0015-large.html#00:15:24.160 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | torques for getting a PhD. So you have to reduce the size of the state space and you have to reduce | https://karpathy.ai/lexicap/0015-large.html#00:15:29.520 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | the horizon if you're going to reason about getting a PhD or even buying the ingredients to | https://karpathy.ai/lexicap/0015-large.html#00:15:36.960 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | make dinner. And so how can you reduce the spaces and the horizon of the reasoning you have to do? | https://karpathy.ai/lexicap/0015-large.html#00:15:42.160 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | And the answer is abstraction, spatial abstraction, temporal abstraction. I think abstraction along | https://karpathy.ai/lexicap/0015-large.html#00:15:49.280 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | the lines of goals is also interesting, like you might, well, abstraction and decomposition. Goals | https://karpathy.ai/lexicap/0015-large.html#00:15:53.760 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | is maybe more of a decomposition thing. So I think that's where these kinds of, if you want to call | https://karpathy.ai/lexicap/0015-large.html#00:16:00.480 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | it symbolic or discrete models come in. You talk about a room of your house instead of your pose. | https://karpathy.ai/lexicap/0015-large.html#00:16:05.600 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | You talk about doing something during the afternoon instead of at 2.54. And you do that because it | https://karpathy.ai/lexicap/0015-large.html#00:16:12.800 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | and you do that because it makes your reasoning problem easier. And also because | https://karpathy.ai/lexicap/0015-large.html#00:16:20.800 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | you have, you don't have enough information to reason in high fidelity about your pose of your | https://karpathy.ai/lexicap/0015-large.html#00:16:27.120 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | elbow at 2.35 this afternoon anyway. Right. When you're trying to get a PhD. | https://karpathy.ai/lexicap/0015-large.html#00:16:34.800 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | Or when you're doing anything really. Yeah. Okay. | https://karpathy.ai/lexicap/0015-large.html#00:16:39.520 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | Except for at that moment, at that moment, you do have to reason about the pose of your elbow, | https://karpathy.ai/lexicap/0015-large.html#00:16:42.720 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | maybe, but then you, maybe you do that in some continuous joint space kind of model. | https://karpathy.ai/lexicap/0015-large.html#00:16:46.080 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | And so again, I, my biggest point about all of this is that there should be the dogma is not | https://karpathy.ai/lexicap/0015-large.html#00:16:50.000 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | the thing, right? We shouldn't, it shouldn't be that I'm in favor against symbolic reasoning | https://karpathy.ai/lexicap/0015-large.html#00:16:58.640 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | and you're in favor against neural networks. It should be that just, just computer science | https://karpathy.ai/lexicap/0015-large.html#00:17:02.880 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | tells us what the right answer to all these questions is. If we were smart enough to figure | https://karpathy.ai/lexicap/0015-large.html#00:17:08.800 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | it out. Well, yeah. When you try to actually solve the problem with computers, the right answer comes | https://karpathy.ai/lexicap/0015-large.html#00:17:12.560 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | out. But you mentioned abstractions. I mean, neural networks form abstractions or rather | https://karpathy.ai/lexicap/0015-large.html#00:17:17.200 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | there's, there's automated ways to form abstractions and there's expert driven ways to | https://karpathy.ai/lexicap/0015-large.html#00:17:24.480 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | form abstractions and expert human driven ways. And humans just seem to be way better at forming | https://karpathy.ai/lexicap/0015-large.html#00:17:28.960 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | abstractions currently and certain problems. So when you're referring to 2.45 PM versus afternoon, | https://karpathy.ai/lexicap/0015-large.html#00:17:35.280 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | how do we construct that taxonomy? Is there any room for automated construction of such | https://karpathy.ai/lexicap/0015-large.html#00:17:44.000 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | abstractions? Oh, I think eventually, yeah. I mean, I think when we get to be better | https://karpathy.ai/lexicap/0015-large.html#00:17:50.080 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | and machine learning engineers, we'll build algorithms that build awesome abstractions. | https://karpathy.ai/lexicap/0015-large.html#00:17:55.280 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | That are useful in this kind of way that you're describing. Yeah. So let's then step from | https://karpathy.ai/lexicap/0015-large.html#00:18:01.120 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | the, the abstraction discussion and let's talk about POMM MDPs. Partially observable | https://karpathy.ai/lexicap/0015-large.html#00:18:05.760 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | Markov decision processes. So uncertainty. So first, what are Markov decision processes? | https://karpathy.ai/lexicap/0015-large.html#00:18:14.800 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | What are Markov decision processes? And maybe how much of our world can be models and MDPs? How | https://karpathy.ai/lexicap/0015-large.html#00:18:20.080 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | much, when you wake up in the morning and you're making breakfast, how do you, do you think of | https://karpathy.ai/lexicap/0015-large.html#00:18:26.320 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | yourself as an MDP? So how do you think about MDPs and how they relate to our world? Well, so | https://karpathy.ai/lexicap/0015-large.html#00:18:30.160 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | there's a stance question, right? So a stance is a position that I take with respect to a problem. | https://karpathy.ai/lexicap/0015-large.html#00:18:36.640 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | So I, as a researcher or a person who designs systems, can decide to make a model of the world | https://karpathy.ai/lexicap/0015-large.html#00:18:42.240 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | around me in some terms. So I take this messy world and I say, I'm going to treat it as if it | https://karpathy.ai/lexicap/0015-large.html#00:18:50.720 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | were a problem of this formal kind, and then I can apply solution concepts or algorithms or whatever | https://karpathy.ai/lexicap/0015-large.html#00:18:57.600 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | to solve that formal thing, right? So of course the world is not anything. It's not an MDP or a | https://karpathy.ai/lexicap/0015-large.html#00:19:02.960 |
Leslie Kaelbling: Reinforcement Learning, Planning, and Robotics | Lex Fridman Podcast #15 | POMM DP. I don't know what it is, but I can model aspects of it in some way or some other way. | https://karpathy.ai/lexicap/0015-large.html#00:19:07.920 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.