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