#258 – Yann LeCun: Dark Matter of Intelligence and Self-Supervised Learning

Yann LeCun is the Chief AI Scientist at Meta, professor at NYU, Turing Award winner, and one of the seminal researchers in the history of machine learning. Please support this podcast by checking out our sponsors: – Public Goods: https://publicgoods.com/lex and use code LEX to get $15 off – Indeed: https://indeed.com/lex to get $75 credit – ROKA: https://roka.com/ and use code LEX to get 20% off your first order – NetSuite: http://netsuite.com/lex to get free product tour – Magic Spoon: https://magicspoon.com/lex and use code LEX to get $5 off EPISODE LINKS: Yann’s Twitter: https://twitter.com/ylecun Yann’s Facebook: https://www.facebook.com/yann.lecun Yann’s Website: http://yann.lecun.com/ Books and resources mentioned: Self-supervised learning (article): https://bit.ly/3Aau1DQ PODCAST INFO: Podcast website: https://lexfridman.com/podcast Apple Podcasts: https://apple.co/2lwqZIr Spotify: https://spoti.fi/2nEwCF8 RSS: https://lexfridman.com/feed/podcast/ YouTube Full Episodes: https://youtube.com/lexfridman YouTube Clips: https://youtube.com/lexclips SUPPORT & CONNECT: – Check out the sponsors above, it’s the best way to support this podcast – Support on Patreon: https://www.patreon.com/lexfridman – Twitter: https://twitter.com/lexfridman – Instagram: https://www.instagram.com/lexfridman – LinkedIn: https://www.linkedin.com/in/lexfridman – Facebook: https://www.facebook.com/lexfridman – Medium: https://medium.com/@lexfridman OUTLINE: Here’s the timestamps for the episode. On some podcast players you should be able to click the timestamp to jump to that time. (00:00) – Introduction (06:58) – Self-supervised learning (17:17) – Vision vs language (23:08) – Statistics (28:55) – Three challenges of machine learning (34:45) – Chess (42:47) – Animals and intelligence (52:31) – Data augmentation (1:13:51) – Multimodal learning (1:25:40) – Consciousness (1:30:25) – Intrinsic vs learned ideas (1:34:37) – Fear of death (1:42:29) – Artificial Intelligence (1:56:18) – Facebook AI Research (2:12:56) – NeurIPS (2:29:08) – Complexity (2:37:33) – Music (2:42:28) – Advice for young people

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00Lex Fridman?

  The following is a conversation with Jan LeCun, his second time on the podcast. He is the chief AI scientist at Meta, formerly Facebook, professor at NYU, touring award winner, one of the seminal figures in the history of machine learning and artificial intelligence, and someone who is brilliant and opinionated in the best kind of way. And so it's always fun to talk to. And now, a quick few second mention of each sponsor. Check them out in the description. It's the best way to support this podcast. First is Public Goods, an online shop I use for household products. Second is Indeed, a hiring website. Third is Roka, my favorite sunglasses and prescription glasses.

• 0:45Lex Fridman?

  Fourth is NetSuite, business software for managing HR, financials, and other details. And fifth is Magic Spoon, low-carb keto-friendly cereal. So the choice is business, health, or style. Choose wisely, my friends. And now, onto the full ad reads. As always, no ads in the middle. I try to make this interesting, but if you skip them, please still check out our sponsors. I enjoy their stuff, maybe you will too. This show is brought to you by Public Goods, the one-stop shop for affordable, sustainable, healthy household products. I use their hand soap, toothpaste, and toothbrush. I think I use a bunch of other stuff too, but that's what comes to mind.

• 1:30Lex Fridman?

  Their products often have this minimalist, black-and-white design that I just absolutely find beautiful. I love it. I love minimalism in design. It doesn't go over the top. It doesn't have all these extra things and features that you don't need, just the essentials. I think it's hard to explain, but there's something about the absence of things that can take up your attention that allows you to truly be attentive to what matters in life. Anyway, go to publicgoods.com/lex or use code LEX at checkout to get $15 off your first order. Plus, you will receive your choice of either free Pack of bamboo straws or reusable food

• 2:13Lex Fridman?

  storage wraps. Visit publicgoods.com/lex or use code Lex at checkout. This show is also brought to you by Indeed, a hiring website. I've used them as part of many hiring efforts I've done for the teams I've led in the past. Most of those for was for engineering for research efforts. They have tools like Indeed Instant Match, giving you quality candidates whose resumes on Indeed fit your job description immediately. For the past few months, I've been going through this process of building up a team of folks to help me. I've been doing quite a bit of hiring. It's a treacherous and an exciting process because you get to meet some friends. So it's

• 2:56Lex Fridman?

  a beautiful process, but I think it's one of the most important processes in life is selecting the group of people with whom you spend your days with. And so you gotta use the best tools for the job. Indeed, I think is an excellent tool. Right now, you can get a free $75 sponsored job credit to upgrade your job post at indeed.com/lex. Terms and conditions apply. Go to indeed.com/lex. This show is also brought to you by Roka, the makers of glasses and sunglasses that I love wearing for their design, feel, and innovation on material optics and grip. Roka was started by two All-American swimmers from Stanford, and it was born out of an obsession with performance.

• 3:42Lex Fridman?

  I like the way they feel. I like the way they look, whether I'm running like a fast-paced run, if we're talking about eight-minute mile or faster, or if we're doing a slow-paced run, nine, 10-minute mile along the river, in the heat or in the cold, or if I'm just wearing my suit out on a town, however that expression goes, I'm not sure, but they look classy with a suit. They look badass in running gear. It's just my go-to sunglasses. Check them out for both prescription glasses and sunglasses at roka.com and enter code Lex to save 20% on your first order. That's roka.com and enter code Lex.

• 4:26Lex Fridman?

  This show is also brought to you by NetSuite. NetSuite allows you to manage financials, human resources, inventory, e-commerce, and many more business-related details all in one place. I'm not sure why I was doing up speak on that sentence. Maybe because I'm very excited about NetSuite. Anyway, there's a lot of messy things you have to get right when running a company. If you're a business owner, if you're an entrepreneur, if you're a founder of a startup, this is something you have to think about. You have to use the best tools for the job to make sure all the messy things required to run a business are taken care of for you so you can focus on the things that you're best at, where your brilliance shines.

• 5:10Lex Fridman?

  If you are starting a business, I wish you the best of luck. It's a difficult journey, but it's worth it. Anyway, right now, special financing is back. Head to netsuite.com/lex to get their one-of-a-kind financing program. That's netsuite.com/lex. Netsuite.com/lex. This episode is also brought to you by Magic Spoon, the OG, not quite OG, but really old-school sponsor of this podcast that I love. It's a low-carb, keto-friendly cereal. It has zero grams of sugar. It's delicious. I don't say that enough. It really is delicious. Given that it's zero grams of sugar, it's very surprising how delicious it is.

• 5:53Lex Fridman?

  13 to 14 grams of protein, only four net grams of carbs, and 140 calories in each serving. You can build your own box or get a variety pack with available flavors of cocoa, fruity, frosted, peanut butter, blueberry, and cinnamon. Cocoa is my favorite. It's the flavor of champions. I don't know why I keep saying that, but it seems to be true. Anyway, Magic Spoon has a 100% happiness guarantee, so if you don't like it, they will refund it. Who else will give you a 100% happiness guarantee? Go to magic spoon.com/lex and use code Lex at checkout to save $5 off your order. That's magic spoon.com/lex and use code Lex.

• 6:35Lex Fridman?

  This is the Lex Fridman Podcast, and here's my conversation with Jan Laakun.

• 6:58Lex Fridman?

  You co-wrote the article Self-supervised Learning: The Dark Matter of Intelligence. Great title, by the way, with Ishaan Misra. So let me ask, what is self-supervised learning? And why is it the dark matter of intelligence?

• 7:11Yann LeCun?

  I'll start by the dark matter part.

• 7:15Yann LeCun?

  There is obviously a kind of learning that humans and animals are

• 7:21Yann LeCun?

  are doing that we currently are not reproducing properly with machines or with AI. Right? So the most popular approaches to machine learning today are or paradigms, I should say, are supervised learning and reinforcement learning. And they are extremely inefficient. Supervised learning requires many samples for learning anything. And reinforcement learning requires a ridiculously large number of trial and errors to for, you know, a system to learn anything.

• 7:51Yann LeCun?

  And that's why we don't have self-driving cars.

• 7:54Lex Fridman?

  That's a big leap from one to the other. Okay, so that to solve difficult problems, you have to have a lot of human annotation for supervised learning to work. And to solve those difficult problems with reinforcement learning, you have to have some way to maybe simulate that problem such that you can do that large scale kind of learning that reinforcement learning requires.

• 8:16Yann LeCun?

  Right. So how is it that, you know, most teenagers can learn to drive a car in about 20 hours of practice? Whereas even with millions of hours of simulated practice, a self-driving car can't actually learn to drive itself properly.

• 8:34Yann LeCun?

  And so obviously, we're missing something, right? And it's quite obvious for a lot of people that, you know, the immediate response you get from many people is, well, you know, humans use their background knowledge to learn faster. And they're right. Now, how was that background knowledge acquired? And that's the big question. So now you have to ask, you know, how do babies in the first few months of life learn how the world works? Mostly by observation, because they can hardly act in the world. And they learn an enormous amount of background knowledge about the world. That may be the the basis of what we call common sense. This type of learning, it's not learning a task. It's not being reinforced for anything. It's just observing the world and figuring out how it works.

• 9:20Yann LeCun?

  Building world models, learning world models. How do we do this? And how do we reproduce this in machines? So self-supervised learning is, you know, one instance or one attempt at trying to reproduce this kind of learning.

• 9:35Lex Fridman?

  Okay, so you're looking at just observation. So not even the interacting part of a child. It's just sitting there watching mom and dad walk around, pick up stuff, all of that. That's what you mean by background knowledge.

• 9:47Yann LeCun?

  Perhaps not even watching mom and dad, just, you know, watching the world go by.

• 9:51Lex Fridman?

  Just having eyes open or having eyes closed, or the very act of opening and closing eyes, that the world appears and disappears, all that basic information.

• 10:01Lex Fridman?

  And you're saying, in order to learn to drive, like, the reason humans are able to learn to drive quickly, some faster than others, is because of the background knowledge. They were able to watch cars operate in the world in the many years leading up to it, the physics of basics objects, and all that kind of stuff.

• 10:17Yann LeCun?

  That's right. I mean, the basic physics of objects, you don't even know, you don't even need to know, you know, how a car works, right? Because that you can learn fairly quickly. I mean, the example I use very often is you're driving next to a cliff. And you know in advance, because of your, you know, understanding of intuitive physics, that if you turn the wheel to the right, the car will veer to the right, will run off the cliff, fall off the cliff, and nothing good will come out of this, right? But if you are a sort of, you know, tabula rasa reinforcement learning system that doesn't have a model of the world,

• 10:50Yann LeCun?

  you have to repeat falling off this cliff thousands of times before you figure out it's a bad idea. And then a few more thousand times before you figure out how to not do it. And then a few more million times before you figure out how to not do it in every situation you ever encounter.

• 11:04Lex Fridman?

  So self-supervised learning still has to have some source of truth being told to it by somebody. So you have to figure out a way without human assistance or without a significant amount of human assistance to get that truth from the world. So the mystery there is how much signal is there? How much truth is there that the world gives you? Whether it's the human world, like you watch YouTube or something like that, or it's the more natural world. So how much signal is there?

• 11:36Yann LeCun?

  So here is the trick. There is way more signal in sort of a self-supervised setting than there is in either a supervised or reinforcement setting. And this is going to my analogy of the cake.

• 11:50Lex Fridman?

  Yes.

• 11:51Yann LeCun?

  The, you know, le cake has someone that's called it. Where when you try to figure out how much information you ask the machine to predict and how much feedback you give the machine at every trial. In reinforcement learning, you give the machine a single scalar. You tell the machine you did good, you did bad. And you, and you, and you only tell this to the machine once in a while. When I say you, it could be the universe telling the machine, right?

• 12:16Yann LeCun?

  But it's just one scalar. So as a consequence, there is, you, you cannot possibly learn something very complicated without many, many, many trials where you get many, many feedbacks of this type. Supervised learning, you, you give a few bits to the machine at every, every sample. Let's say you're training a system on, you know, recognizing images on ImageNet. There is 1,000 categories, that's a little less than 10 bits of information per sample.

• 12:44Yann LeCun?

  But self-supervised learning, here is a setting. You ideally, we don't know how to do this yet, but ideally you would show a machine a segment of video and then stop the video and ask the machine to predict what's going to happen next.

• 12:59Yann LeCun?

  And so you let the machine predict and then you let time go by and show the machine what actually happened and hope the machine will, you know, learn to do a better job at predicting next time around. There's a huge amount of information you give the machine because it's an entire video clip of, you know, of the future after the video clip you fed it in the first place.

• 13:22Lex Fridman?

  So both for language and for vision, there's a subtle, seemingly trivial construction, but maybe that's representative of what is required to create intelligence, which is filling the gap.

• 13:35Yann LeCun?

  Filling the gaps.

• 13:38Lex Fridman?

  It sounds dumb, but can you, it's, it is possible you could solve all of intelligence in this way, just for both language, just give a sentence and continue it or give a sentence and there's a gap in it, some words blanked out and you fill in what words go there. For vision, you give a sequence of images and predict what's going to happen next or you fill in what happened in between. Do you think it's possible that formulation alone?

• 14:10Lex Fridman?

  As a signal for self-supervised learning can solve intelligence for vision and language?

• 14:15Yann LeCun?

  I think that's our best shot at the moment. So whether this will take us all the way to, you know, human level intelligence or something, or just cat level intelligence is not clear. But among all the possible approaches that people have proposed, I think it's our best shot. So I think this idea of an intelligent system filling in the blanks, either predicting the future, inferring the past, filling in missing information. You know, I'm currently filling the blank of what is behind your head and what your head looks like and, you know, from the back. Because I have, you know, basic knowledge about how humans are made. And I don't know if you're gonna, you know, what are you gonna say at which point you're gonna speak?

• 14:59Yann LeCun?

  Whether you're gonna move your head this way or that way, which way you're gonna look. But I know you're not gonna just dematerialize and reappear three meters down the hall. You know, because I know what's possible and what's impossible according to intuitive physics.

• 15:12Lex Fridman?

  Well, you have a model of what's possible, what's impossible. And then you'd be very surprised if it happens and then you'll have to reconstruct your model.

• 15:19Yann LeCun?

  Right, so that's the model of the world. It's what tells you, you know, what fills in the blanks. So given your partial information about the state of the world, given by your perception, your model of the world fills in the missing information. And that includes predicting the future, predicting the past, you know, filling in things you don't immediately perceive.

• 15:40Lex Fridman?

  And that doesn't have to be purely generic vision or visual information or generic language. You can go to specifics like predicting what control decision you make when you're driving in a lane. You have a sequence of images from a vehicle. And then you could, you have information, if you record it on video, where the car ended up going. So you can go back in time and predict where the car went based on the visual information. That's very specific, domain-specific.

• 16:11Yann LeCun?

  Right, but the question is whether we can come up with sort of a generic method for, you know, training machines to do this kind of prediction or filling in the blanks. So right now,

• 16:24Yann LeCun?

  this type of approach has been unbelievably successful in the context of natural language processing. Every modern natural language processing is pre-trained in self-supervised manner to fill in the blanks. You show it a sequence of words, you remove 10% of them, and then you train some gigantic neural net to predict the words that are missing. And once you've pre-trained that network, you can use the internal representation learned by it as input to, you know, something that you trained supervised or whatever. That's been incredibly successful, not so successful in images, although it's making progress. And it's based on sort of manual data augmentation. We can go into this later, but what has not been successful yet is

• 17:08Yann LeCun?

  training from video. So getting a machine to learn to represent the visual world, for example, by just watching video. Nobody has really succeeded in doing this.

• 17:16Lex Fridman?

  Okay, well, let's kind of give a high-level overview. What's the difference in kind and in difficulty between vision and language? So you said people haven't been able to really kind of crack the problem of vision open in terms of self-supervised learning, but that may not be necessarily because it's fundamentally more difficult. Maybe like when we're talking about achieving, like passing the Turing test in the full spirit of the Turing test in language might be harder than vision. That's not obvious. So in your view, which is harder? Or perhaps are they just the same problem? When the farther we get to solving each, the more we realize it's all the same thing. It's all the same cake.

• 17:59Yann LeCun?

  I think what I'm looking for are methods that make them look essentially like the same cake, but currently they're not. And the main issue with learning word models or learning predictive models is that

• 18:14Yann LeCun?

  the prediction is never a single thing, because the world is not entirely predictable. It may be deterministic or stochastic. We can get into the philosophical discussion about it. But even if it's deterministic, it's not entirely predictable.

• 18:29Yann LeCun?

  And so if I play a short video clip and then I ask you to predict what's going to happen next, there is many, many plausible continuations for that video clip. And the number of continuation grows with the interval of time that you're asking the system to make a prediction for. And so one big question with self-supervised learning is how you represent this uncertainty? How you represent multiple discrete outcomes? How you represent a sort of continuum of possible outcomes, et cetera? And, you know, if you are a sort of a classical machine learning person, you say, oh, you just represent a distribution, right?

• 19:11Yann LeCun?

  And that we know how to do when we're predicting words, missing words in the text. Because you can have a neural net give a score for every word in a dictionary. It's a big, you know, it's a big list of numbers, you know, maybe a hundred thousand or so. You can turn them into a probability distribution that gives, that tells you when I say a sentence, you know, the, you know, the cat is chasing the blank in the kitchen. You know, there are only a few words that make sense there, you know? It could be a mouse, or it could be a lizard spot, or you know, something like that, right?

• 19:43Yann LeCun?

  And, and if I, if I say the, the blank is chasing the blank in the savanna, you also have a bunch of plausible options for those two words, right?

• 19:52Yann LeCun?

  That, that, that, because you, you have kind of a, you know, underlying reality that you can refer to, to sort of fill in those, those blanks.

• 20:00Yann LeCun?

  So you, you cannot say for sure in the savanna, if it's a, you know, a lion or a cheetah or whatever. You cannot know if it's a zebra or a gnu or, you know, whatever, wildebeest, the same thing.

• 20:14Yann LeCun?

  But, uh,

• 20:17Yann LeCun?

  but you can represent the uncertainty by just a long list of numbers. Now, If I do the same thing with video and I ask you to predict a video clip, it's not a discrete set of potential frames. You have to have somewhere representing a sort of infinite number of plausible continuations of multiple frames in a, you know, high dimensional continuous space. And we just have no idea how to do this properly.

• 20:42Lex Fridman?

  Finite, high dimensional. So like you.

• 20:45Yann LeCun?

  It's finite, high dimensional, yes.

• 20:47Lex Fridman?

  Just like the words, they try to get it to, uh, down to a small finite set of like under a million, something like that.

• 20:56Yann LeCun?

  Something like that.

• 20:56Lex Fridman?

  I mean, it's kind of ridiculous that we're doing a distribution of every single possible word for language and it works. It feels like that's a really dumb way to do it. Like there seems to, there, there, it seems to be like there should be some more compressed representation of the distribution of the words.

• 21:17Yann LeCun?

  You're right about that.

• 21:18Lex Fridman?

  And so.

• 21:18Yann LeCun?

  I agree.

• 21:19Lex Fridman?

  Do you have any interesting ideas about how to represent all of the reality in a compressed way such that you can form a distribution over it?

• 21:25Yann LeCun?

  That's one of the big questions, you know, how do, how do you do that? Right, I mean, what's, what's kind of, you know, another thing that, that really is stupid about, um, I shouldn't say stupid, but like simplistic about current approaches to self-supervising in, in, uh, NLP and text is that not only do you represent a giant distribution of words, but for multiple words that are missing, those distributions are essentially independent of each other. And, you know, you don't pay too much of a price for this, so you, so you can't, so the, you know, the system, you know, in the, the sentence that I gave earlier, uh, if it gives a certain probability for a lion and, and cheetah, and then a certain probability

• 22:06Yann LeCun?

  for, uh, you know, gazelle, uh, wildebeest and, and zebra, uh, those two probabilities are independent of each other. And it's not the case that those things are independent lions actually attack like bigger animals than cheetahs. So, you know, there is a huge independence hypothesis in this process, which is not actually true. The reason for this is that we don't know how to represent properly distributions over combinatorial sequences of symbols, essentially, because the number grows exponentially with the length of the symbols. And so we have to use tricks for this, but those techniques can, you know, get around,

• 22:48Yann LeCun?

  like, don't even deal with it. So, so the big question is, like, would there be some sort of abstract latent representation of text that would say that, you know, when I, when I switch lion for gazelle, lion for cheetah, I also have to switch zebra for gazelle?

• 23:06Lex Fridman?

  Yeah, so this independence assumption, let me throw some criticism at you that I often hear and see how you respond. So this kind of filling in the blanks is just statistics? You're not learning anything like the deep underlying concepts. You're just mimicking stuff from the past. You're not learning anything new such that you can use it to generalize about the world. Or, okay, let me just say the crude version, which is is just statistics. It's not intelligence. What do you have to say to that? What do you usually say to that if you kind of hear this kind of thing?

• 23:44Yann LeCun?

  I don't get into those discussions because they are kind of pointless. So, first of all, it's quite possible that intelligence is just statistics. It's just statistics of a particular kind.

• 23:54Lex Fridman?

  Yes. Well, this is the philosophical question. It's kind of is, is it possible that intelligence is just statistics?

• 24:02Yann LeCun?

  Yeah. But what kind of statistics? So, if you are asking the question...

• 24:09Yann LeCun?

  Are the model of the world, the models of the world that we learn, do they have some notion of causality? Yes. So if the criticism comes from people who say, you know, current machine learning system don't care about causality, which by the way is wrong.

• 24:24Yann LeCun?

  You know, I agree with them. Yeah, you should, you know, your model of the world should have your actions as one of your, of the inputs. And that will drive you to learn causal models of the world where you know what, you know, what intervention in the world will cause what result. Or you can do this by observation of other agents acting in the world and observing the effect. Other humans, for example. So I think, you know, at some level of description, intelligence is just statistics. But that doesn't mean you don't, you don't, you know, you won't have models that have, you know, deep mechanistic explanation for what goes on. The question is, how do you learn them? That's, that's the question I'm interested in.

• 25:06Yann LeCun?

  Because, you know, a lot of people who actually voiced their criticism, say that those mechanistic model has to, have to come from someplace else. They have to come from human designers. They have to come from, I don't know what. And obviously, we learn them.

• 25:21Yann LeCun?

  Or if we don't learn them as an individual, nature learned them for us using evolution. So regardless of what you think, those processes have been learned somehow.

• 25:32Lex Fridman?

  So if you look at the, the human brain, just like when we humans introspect about how the brain works, it seems like when we think about what is intelligence, we think about the high level stuff, like the models we've constructed, concepts like cognitive science, like concepts of memory and reasoning module, almost like these high level modules. Is there, is this serve as a good analogy? Like, are we ignoring the?

• 26:01Lex Fridman?

  The dark matter, the the basic low level mechanisms, just like we ignore the way the operating system works, we're just using the the the high level software, we're ignoring that at the low level, the neural network might be doing something like statistics. Like me, sorry to use this word probably incorrectly and crudely, but doing this kind of fill in the gap kind of learning and just kind of updating the model constantly in order to be able to support the raw sensory information to predict it and then adjust to the prediction when it's wrong. But like high level, when we look at our brain at the high level, it feels like we're doing like we're playing chess, like we're we're like playing with high level concepts

• 26:44Lex Fridman?

  and we're stitching them together, we're putting them into long term memory, but really what's going underneath is something we're not able to introspect, which is this kind of a simple, large neural network that's just filling in the gaps.

• 26:58Yann LeCun?

  Right, well, okay, so there's a lot of questions and a lot of answers there. Okay, so first of all, there's a whole school of thought in neuroscience, computational neuroscience in particular, that likes the idea of predictive coding, which is really related to the idea I was talking about in self supervised learning. So everything is about prediction, the essence of intelligence is the ability to predict, and everything the brain does is trying to predict, predict everything from everything else. Okay, and that's really sort of the underlying principle, if you want, that self supervised learning is trying to kind of reproduce this idea of prediction as kind of an essential mechanism of task independent learning, if you want. The next step is, what kind of intelligence are you interested in reproducing?

• 27:43Yann LeCun?

  And of course, you know, we all think about, you know, trying to reproduce sort of, you know, high level cognitive processes in humans, but like with machines, we're not even at the level of even reproducing the learning processes in a in a cat brain. You know, the most intelligent or intelligent systems don't have as much common sense as as a house cat.

• 28:05Yann LeCun?

  So how is it that cats learn? And you know, cats don't do a whole lot of reasoning. They certainly have causal models. They certainly have, because you know, many cats can figure out like how they can act on the world to get what they want. They certainly have a fantastic model of intuitive physics. Certainly of the the the dynamics of their own bodies, but but also of praise and things like that, right? So they they're they're pretty smart. They they only do this with about 800 million neurons. We are not anywhere close to reproducing this kind of thing. So to some extent, I I could I could say, let's not even worry about like the high level cognition.

• 28:48Yann LeCun?

  And kind of, you know, long term planning and reasoning that humans can do until we figure out like, you know, can we even reproduce what cats are doing? Now that said, this ability to learn world models, I think, is the key to the possibility of learning machines that can also reason. So whenever I give a talk, I say there are there are three challenges in the three main challenges in machine learning. The first one is, you know, getting machines to learn to represent the world. And I'm proposing self supervised learning. The second is getting machines to reason in ways that are compatible with essentially gradient based learning, because this is what deep learning is all about, really.

• 29:27Yann LeCun?

  And the third one is something we have no idea how to solve, at least I have no idea how to solve, is

• 29:33Yann LeCun?

  can we get machines to learn hierarchical representations of action plans?

• 29:39Yann LeCun?

  You know, like, you know, we know how to train them to learn hierarchical representations of perception. You know, with computational nets and things like that and transformers. But what about action plans? Can we get them to spontaneously learn good hierarchical representations of actions?

• 29:52Lex Fridman?

  Also gradient based.

• 29:54Yann LeCun?

  Yeah, all of that, you know, needs to be somewhat differentiable so that you can apply sort of gradient based learning, which is really what deep learning is about.

• 30:04Lex Fridman?

  So it's background, knowledge, ability to reason in a way

• 30:11Lex Fridman?

  that's differentiable, that is somehow connected, deeply integrated with that background knowledge, or builds on top of that background knowledge, and then given that background knowledge, be able to make hierarchical plans in the world.

• 30:24Yann LeCun?

  So if you take classical optimal control, there's something in classical optimal control called model predictive control, and it's been around since the early 60s. NASA uses that to compute trajectories of rockets. And the basic idea is that you have a predictive model of the rocket, let's say, or whatever system you intend to control, which, given the state of the system at time t and given an action that you're taking in the system, so for a rocket would be thrust and all the controls you can have, it gives you the state of the system at time t plus delta t, right? So basically a differential equation, something like that.

• 31:06Yann LeCun?

  And if you have this model, and you have this model in the form of some sort of neural net, or some sort of set of formula that you can back propagate gradient through, you can do what's called model predictive control, or gradient-based model predictive control. So you have, you can unroll that model in time, you feed it a hypothesized sequence of actions, and then you have some objective function that measures how well, at the end of the trajectory, the system has succeeded or matched what you want it to do. You know, is it a robot arm? Have you grasped the object you want to grasp? If it's a rocket, are you at the right place near the space station? Things like that.

• 31:51Yann LeCun?

  And by backpropagation through time, and again, this was invented in the 1960s by optimal control theorists, you can figure out what is the optimal sequence of actions that will, you know, get my system to the best final state.

• 32:07Yann LeCun?

  So that's a form of reasoning. It's basically planning and a lot of planning systems in robotics are actually based on this. And you can think of this as a form of reasoning. So, you know, to take the example of the teenager driving a car again. You have a pretty good dynamical model of the car. It doesn't need to be very accurate, but you know, again, that if you turn the wheel to the right and there is a cliff, you're going to run off the cliff, right? You don't need to have a very accurate model to predict that. And you can run this in your mind and decide not to do it for that reason. Because you can predict in advance that the result is going to be bad. So you can sort of imagine different scenarios and then, you know, employ or take the first step in the scenario that is most favorable and then repeat the process of planning. That's called receding horizon model predictive control.

• 32:52Yann LeCun?

  So even, you know, all those things have names, you know, going back, you know, decades. And so if you're not, you know, in classical optimal control, the model of the world is not generally learned. There's, you know, sometimes a few parameters you have to identify. That's called systems identification. But, but generally the model is mostly deterministic and mostly built by hand. So the big question of AI, I think the big challenge of AI for the next decade is how do we get machines to learn predictive models of the world that deal with uncertainty and deal with the real world in all this complexity? So it's not just the trajectory of a rocket, which you can reduce to first principles.

• 33:32Yann LeCun?

  It's not, it's not even just the trajectory of a robot arm, which again, you can model by, you know, careful mathematics. But it's everything else, everything we observe in the world, you know, people, behavior, you know, physical systems that involve collective phenomena like water or, or, you know, trees and, you know, branches in a, in a tree or something, or, or like complex things that, you know, humans have no trouble developing abstract representations and predictive model for, but we still don't know how to do with machines.

• 34:03Lex Fridman?

  Where do you put in, in these three, maybe in the, in the planning stages? The game theoretic nature of this world where your actions not only respond to the dynamic nature of the world, the environment, but also affect it. So if there's other humans involved, is this is this point number four? Or is it somehow integrated into the hierarchical representation of action in your view?

• 34:28Yann LeCun?

  I think it's integrated. It's just it's just that now your model of the world has to deal with, you know, it just makes it more complicated. Right. The fact that humans are complicated and not easily predictable. That makes your model of the world much more complicated, that much more complicated.

• 34:43Lex Fridman?

  Well, there's a chess. I mean, I suppose chess is an analogy. So multi car tree search. I mean, it there is a I go, you go, I go, you go. Like, Andre Kapata recently gave a talk at MIT about car doors. I think there's some machine learning too, but mostly car doors. And there's a dynamic nature to the car, like the person opening the door checking. And he wasn't talking about that. He was talking about the perception problem of what the ontology of what defines a car door. This big philosophical question. But to me, it was interesting because like it's obvious that the person opening the car doors, they're trying to get out like here in New York. Trying to get out of the car. You slowing down is going to signal something. You speeding up is going to signal something and that's a dance.

• 35:28Lex Fridman?

  It's a asynchronous chess game. I don't know. So it feels like it's not just, I mean, I guess you can integrate all of them into one giant model. Like the entirety of the these little interactions. Because it's not as complicated as chess. It's just like a little dance. We do like a little dance together and then we figure it out.

• 35:51Yann LeCun?

  Well, in some ways, it's way more complicated than chess. Because because it's continuous, it's uncertain in a continuous manner.

• 36:00Lex Fridman?

  It doesn't feel more complicated.

• 36:01Yann LeCun?

  But it doesn't feel more complicated because that's what we're we've evolved to solve. This is the kind of problem we've evolved to solve. And so we're good at it because, you know, nature has made us good at it. Nature has not made us good at chess. We completely suck at chess. Yeah. In fact, that's why we designed it as a game, is to be challenging. And if there is something that, you know, recent progress in chess and Go has made us realize is that humans are really terrible at those things, like really bad. You know, there was a story, right, before AlphaGo that,

• 36:35Yann LeCun?

  you know, the best Go players thought they were maybe two or three stones behind, you know, an ideal player that they would call god. In fact, no, they are like nine or 10 stones behind. I mean, we're just bad. So we're not good at, and it's because we have limited working memory. We're not very good at like doing this tree exploration that, you know, computers are much better at doing than we are. But we are much better at learning differentiable models of the world. I mean, I said differentiable in a kind of, you know, I should say, not differentiable in the sense that, you know, we run backprop through it, but in the sense that our brain has some mechanism for estimating gradients of some kind. Yeah. And that's what, you know, makes us efficient. So if you have an agent

• 37:21Yann LeCun?

  that consists of a model of the world, which, you know, in the human brain is basically the entire front half of your brain,

• 37:30Yann LeCun?

  an objective function, which in humans is a combination of two things. There is your sort of intrinsic motivation module, which is on the basal ganglia, you know, at the base of your brain. That's the thing that measures pain and hunger and things like that, like immediate

• 37:47Yann LeCun?

  feelings and emotions. And then there is, you know, the equivalent of what people in reinforcement training call a critic, which is a sort of module that predicts ahead what the outcome of a of a situation will be. And so it's not a cost function, but it's sort of, not an objective function, but it's sort of a You know, trained predictor of the ultimate objective function. And that also is differentiable. And so if all of this is differentiable, your cost function, your critic, your, you know, your world model, then you can use gradient based type methods to do planning, to do reasoning,

• 38:26Yann LeCun?

  to do learning, you know, to do all the things that would like an intelligent agent to do.

• 38:33Lex Fridman?

  And gradient based learning, like what's your intuition? That's probably at the core of what can solve intelligence. So you don't need like a logic based reasoning in your view.

• 38:47Yann LeCun?

  I don't know how to make logic based reasoning comparable with efficient learning.

• 38:51Lex Fridman?

  Yeah.

• 38:52Yann LeCun?

  And, okay, I mean, there is a big question, perhaps a philosophical question. I mean, it's not that philosophical, but that we can ask is, is that, you know, all the learning algorithms we know from engineering and computer science proceed by optimizing some objective function.

• 39:08Lex Fridman?

  Yeah.

• 39:08Yann LeCun?

  Right? So one question we may ask is, is, does learning in the brain minimize an objective function? It could be, you know, a composite of multiple objective functions, but it's still an objective function.

• 39:23Yann LeCun?

  Second, if it does optimize an objective function, does it do, does it do it by some sort of gradient estimation? You know, it doesn't need to be backprop, but you know, some way of estimating the gradient in efficient manner, whose complexity is on the same order of magnitude as, you know, actually running the inference.

• 39:44Yann LeCun?

  Because you can't afford to do things like, you know, perturbing a weight in your brain to figure out what the effect is. And then sort of, you know, you can do sort of estimating gradient by perturbation. It's, to me, it seems very implausible that the brain uses some sort of, you know, zeroth order, black box, gradient free optimization, because it's so much less efficient than gradient optimization. So it has to have a way of estimating gradients.

• 40:11Lex Fridman?

  Is it possible that some kind of logic reasoning emerges in pockets as a useful, like you said, if the brain is an objective function, maybe it's a mechanism for creating objective functions. It's a mechanism for creating knowledge bases, for example, that can then be queried. Like, maybe it's an efficient representation of knowledge that's learned in a gradient-based way, or something like that.

• 40:35Yann LeCun?

  Well, so I think there is a lot of different types of intelligence. So first of all, I think the type of logical reasoning that we think about, that we are, you know, maybe stemming from, you know, sort of classical AI of the 1970s and 80s. I think humans use that relatively rarely and are not particularly good at it.

• 40:56Lex Fridman?

  But we judge each other based on our ability to solve those rare problems. It's called an IQ test.

• 41:03Yann LeCun?

  I don't think so. Like, I'm not very good at chess.

• 41:06Lex Fridman?

  Yes, I'm judging you this whole time because, well, we actually...

• 41:11Yann LeCun?

  With your, you know, heritage, I'm sure you're good at chess.

• 41:14Lex Fridman?

  No, stereotypes, not all stereotypes are true.

• 41:18Yann LeCun?

  Well, I'm terrible at chess. So, you know, but I think perhaps another type of intelligence that I have is this, you know, ability of sort of building models of the world from, you know, reasoning, obviously, but also data. And those models generally are more kind of analogical, right? So it's reasoning by simulation and by analogy, where you use one model to apply to a new situation, even though you've never seen that situation, you can sort of connect it to a situation you've encountered before. And your reasoning is more, you know, akin to some sort of internal simulation.

• 42:00Yann LeCun?

  So you're kind of simulating what's happening when you're building, I don't know, a box out of wood or something, right? You can imagine in advance, like, what would be the result of, you know, cutting the wood in this particular way? Are you going to use, you know, screws or nails or whatever? When you are interacting with someone, you also have a model of that person and sort of interact with that person. You know, having this model in mind to kind of tell the person what you think is useful to them. So I think this this ability to construct models of the world is basically the essence, the essence of intelligence and the ability to use it then to plan actions

• 42:40Yann LeCun?

  that will fulfill a particular criterion, of course, is necessary as well.

• 42:47Lex Fridman?

  So I'm going to ask you a series of impossible questions as we keep asking, as I've been doing. So so if that's the fundamental sort of dark matter of intelligence, this ability to form a background model, what's your intuition about how much knowledge is required? You know, I think dark matter, you could put a percentage on it of of the composition of the universe, how much of it is dark matter, how much of it is dark energy, how much information do you think is required to be a house cat? So you have to be able to, when you see a box, go in it, when you see a human compute the most evil action, if there's a thing

• 43:29Lex Fridman?

  that's near an edge, you knock it off, all of that, plus the extra stuff you mentioned, which is a great self-awareness of the physics of your of your own body and in the world. How much knowledge is required, do you think, to solve it? I don't even know how to measure an answer to that question.

• 43:47Yann LeCun?

  I'm not sure how to measure it, but whatever it is, it fits in about about 800,000 neurons, 800 million neurons, sorry.

• 43:55Lex Fridman?

  The representation does.

• 43:57Yann LeCun?

  Everything, all knowledge, everything, right? You know, it's less than a billion. A dog is two billion, but a cat is less than one billion. And so multiply that by a thousand and you get the number of synapses. And I think almost all of it is learned through this, you know, a sort of self-supervised learning, although, you know, I think a tiny sliver is learned through reinforcement learning and certainly very little through. You know, classical supervised running, although it's not even clear how supervised running actually works in the biological world. So I think almost all of it is self-supervised running. But it's driven by the sort of ingrained objective functions that a cat or a human

• 44:42Yann LeCun?

  have at the base of their brain, which kind of drives their behavior. So, you know, nature tells us, you're hungry. It doesn't tell us how to feed ourselves. That's something that the rest of our brain has to figure out, right?

• 44:56Lex Fridman?

  Well, it's interesting because there might be more like deeper objective functions underlying the whole thing. So hunger may be some kind of, now you go to like neurobiology, it might be just the brain trying to maintain homeostasis. So hunger is just one of the human perceivable symptoms of the brain being unhappy with the way things are currently.

• 45:22Yann LeCun?

  Right.

• 45:23Lex Fridman?

  Because it could be just like one really dumb objective function at the core.

• 45:26Yann LeCun?

  But that's how, that's how behavior is driven. The fact that, you know, our basal ganglia drives us to do things that are, that are different from, say, an orangutan or certainly a cat is what makes, you know, human nature versus orangutan nature versus cat nature. So for example, you know, our basal ganglia drives us to seek the company of other humans. And that's because nature has figured out that we need to be social animals for our species to survive, and it's true of many primates. It's not true of orangutans. Orangutans are solitary animals. They don't seek the company of others. In fact, they avoid them.

• 46:09Yann LeCun?

  In fact, they scream at them when they come too close because they are territorial. Because for their survival, you know, evolution has figured out that's the best thing. I mean, they are occasionally social, of course, for, you know, reproduction and stuff like that. But, but they're mostly solitary. So, so all of those behaviors are not part of intelligence. You know, people say, oh, you're never going to have intelligent machines because, you know, human intelligence is social. But then you look at orangutans, you look at octopus. Octopus never know their parents. They barely interact with any other and they and they get to be really smart in less than a, less than a year, in like half a year.

• 46:48Yann LeCun?

  You know, in a year they're adults, in two years they're dead. So there are things that we think as humans are intimately linked with intelligence, like social interaction, like language. We think, I think we give way too much importance to language as a substrate of intelligence as humans, because we think our reasoning is so linked with language.

• 47:11Lex Fridman?

  So for, to solve the house cat intelligence problem, you think you could do it on a desert island, you could have.

• 47:18Yann LeCun?

  Pretty much.

• 47:18Lex Fridman?

  You could just have a cat sitting there, um, looking at the waves, at the ocean waves, and figure a lot of it out.

• 47:27Yann LeCun?

  It needs to have sort of, you know, the right set of drives to kind of, you know, get it to do the thing and learn the appropriate things, right? But, uh, like for example, you know, uh, baby humans are, are driven to learn to stand up and walk. Okay, you know, it's not, that's kind of, th- this desire is hardwired. How to do it precisely is not, that's learned. But the desire to.

• 47:51Lex Fridman?

  To walk?

• 47:52Yann LeCun?

  Move around and stand up, uh, that's sort of.

• 47:56Lex Fridman?

  Ah.

• 47:57Yann LeCun?

  It's very simple to hardwire this kind of stuff.

• 48:00Lex Fridman?

  Oh, like the desire to, well, that's interesting. You're hardwired to want to walk? That's not a, there's got to be a deeper need for walking. I, I think it was probably socially imposed by society that you need to walk. All the other bipedal.

• 48:17Yann LeCun?

  No, like a lot of simple animals that, you know, would probably walk without ever watching any other members of the species.

• 48:25Lex Fridman?

  It seems like a scary thing to have to do, because you suck at bipedal walking at first. It seems crawling is much safer, much more like, why are you in a hurry?

• 48:37Yann LeCun?

  Well, because, because you have this thing that drives you to do it, you know, which is sort of part of, uh, the sort of. Human development.

• 48:47Lex Fridman?

  Is that understood actually what?

• 48:48Yann LeCun?

  Not entirely, no.

• 48:49Lex Fridman?

  What is, what's the reason to get on two feet? It's really hard, like most animals don't get on two feet.

• 48:54Yann LeCun?

  Well, they get on four feet, you know, many mammals get on four feet.

• 48:57Lex Fridman?

  Yeah.

• 48:58Yann LeCun?

  Very quickly, some of them extremely quickly.

• 49:00Lex Fridman?

  But I don't, you know, like from the last time I've interacted with a table, that's much more stable than a thing than two legs. It's just a really hard problem.

• 49:08Yann LeCun?

  Yeah, I mean, birds have figured it out with two feet.

• 49:11Lex Fridman?

  Well, technically, we can go into ontology, they have four, I guess they have two feet.

• 49:16Yann LeCun?

  They have two feet.

• 49:17Lex Fridman?

  Chickens.

• 49:18Yann LeCun?

  You know, dinosaurs had two feet, many of them.

• 49:20Lex Fridman?

  Allegedly.

• 49:23Lex Fridman?

  I'm just now learning that T. rex was eating grass, not other animals. T. rex might have been a friendly, friendly pet. What do you think about, I don't know if you looked at the test for general intelligence that Francois Chollet put together? I don't know if you got a chance to look at that kind of thing, like what's your intuition about how to solve like an IQ type of test?

• 49:45Yann LeCun?

  I don't know, I think it's so outside of my radar screen that it's not really relevant, I think in the short term.

• 49:53Lex Fridman?

  I guess one way to ask, another way, perhaps more closer to what to your work is like, how do you solve MNIST with very little example data?

• 50:04Yann LeCun?

  That's right, and that's the answer to this probably is self-supervised learning, just learn to represent images and then learning, you know, to recognize handwritten digits on top of this will only require a few samples. We observe this in, in humans, right? You, you show a young child a picture book with a couple of pictures of an elephant and that's it. The child knows what an elephant is. And we, we, we see this today with practical systems that we, you know, we train image recognition systems with

• 50:33Yann LeCun?

  enormous amounts of, of images, either, either completely self-supervised or very weakly supervised. For example, you can train a neural net to predict whatever hashtag people type on Instagram. Right, then you can do this with billions of images because there's billions per day that are showing up. So the amount of training data there is essentially unlimited. And then you take the output representation, you know, a couple of layers down from the output of what the system learned and feed this as input to a classifier for any object in the world that you want, and it works pretty well. So that's transfer learning, okay, or weekly supervised transfer learning.

• 51:13Yann LeCun?

  People are making very, very fast progress using self-supervised learning for, with this kind of scenario as well. And, you know, my guess is that that's, that's going to be the, the future.

• 51:24Lex Fridman?

  For self-supervised learning, how much cleaning do you think is needed for filtering

• 51:32Lex Fridman?

  malicious signal or what's a better term, but like a lot of people use hashtags on Instagram to get like good SEO that doesn't fully represent the contents of the image. Like they'll put a picture of a cat and hashtag it with like science, awesome, fun. I don't know all kinds. Why would you put science? That's not very good SEO.

• 51:54Yann LeCun?

  The way, the way my, my colleagues who worked on this project at, at Facebook now Meta, Meta AI a few years ago, dealt with this is that they only selected something like 17,000 tags that correspond to kind of physical things or, or situations like, you know, that has some visual content.

• 52:14Yann LeCun?

  So, you know, you wouldn't have like hashtag TBT or anything like that.

• 52:19Lex Fridman?

  Oh, so they, they keep a very select set of hashtags, is what you're saying?

• 52:23Yann LeCun?

  Yeah.

• 52:23Lex Fridman?

  Okay.

• 52:24Yann LeCun?

  But it's still, it's still in the order of, you know, 10 to 20,000. So it's fairly large.

• 52:29Lex Fridman?

  Okay. Can you tell me about data augmentation? What the heck is data augmentation? And how is it used? Maybe contrast of learning for, for video? What are some cool ideas here?

• 52:42Yann LeCun?

  Right. So data augmentation, I mean, first data augmentation, you know, is the idea of artificially increasing the size of your training set by distorting the images that you have in ways that don't change the nature of the image, right? So you take, you, you do MNIST, you can do data augmentation on MNIST. And people have done this since the 1990s, right? You take a MNIST digit and you shift it a little bit or you change the size or rotate it, skew it. You know, etc.

• 53:08Lex Fridman?

  Add noise.

• 53:10Yann LeCun?

  Add noise, etc. And it it works better. If you train a supervised classifier with augmented data, you're going to get better results. Now, it's become really interesting over the last couple of years because a lot of self-supervised learning techniques to pre-train vision systems are based on data augmentation. And the the basic techniques is originally inspired by techniques that I worked on in the early 90s and Jeff Hinton worked on also in the early 90s. There was sort of parallel work. I used to call this Siamese network. So basically, you take two identical copies of the same network. They share the same weights. And you show two different views of the same object.

• 53:53Yann LeCun?

  Either those two different views may have been obtained by data augmentation, or maybe it's two different views of the same scene from a camera that you moved or at different times or something like that, right? Or two pictures of the same person, things like that. And then you train this neural net, those two identical copies of this neural net, to produce an output representation, a vector, in such a way that the representation for those two images are as close to each other as possible, as identical to each other as possible, right? Because you want the system to basically learn a function that will that will be invariant, that will not change, whose output will not change when you transform those inputs in those in those particular ways, right?

• 54:36Yann LeCun?

  So that's easy to do. What's complicated is how do you make sure that when you show two images that are different, the system will produce different things? Because if you don't have a specific provision for this, the system will just ignore the inputs. When you train it, it will end up ignoring the input and just produce a constant vector that is the same for every input, right?

• 54:55Lex Fridman?

  Yes.

• 54:55Yann LeCun?

  That's called a collapse. Now, how do you avoid collapse? So there's two ideas. One idea that I proposed in the early 90s with my colleagues at Bell Labs, Jane Bromley and a couple of other people, which we now call contrastive learning, which is to have negative examples, right? So you have pairs of images that you know are different. And you show them to the network and those two copies. And then you you push the two output vectors away from each other. And it will eventually guarantee that things that are semantically similar produce similar representations and things that are different produce different representations.

• 55:32Yann LeCun?

  We actually came up with this idea for a project of doing a signature verification. So we would collect signature signatures from like multiple signatures on the same person and then train a neural net to produce the same representation. And then, you know, force the system to produce different representations for different signatures.

• 55:53Yann LeCun?

  This was actually the the problem was proposed by people from what was a subsidiary of AT&T at the time called NCR. And they were interested in storing a representation of the signature on the 80 bytes of the magnetic strip of a credit card. So we came up with this idea of having a neural net with 80 outputs, you know, that we would quantize on bytes so that we could encode the

• 56:15Lex Fridman?

  And that encoding was then used to compare whether the signature matches or not.

• 56:19Yann LeCun?

  That's right. So then you would, you know, sign, it would run through a neural net and then you would compare the output vector to whatever is stored on your card.

• 56:25Lex Fridman?

  Did it actually work?

• 56:26Yann LeCun?

  It worked, but they ended up not using it.

• 56:30Yann LeCun?

  Because nobody cares, actually. I mean, the American, you know, financial payment system is incredibly lax in that respect compared to Europe.

• 56:39Lex Fridman?

  Oh, with the signatures? What's the purpose of signatures, anyway? This is very

• 56:43Yann LeCun?

  Nobody looks at them. Nobody cares.

• 56:44Lex Fridman?

  And it's, yeah.

• 56:46Yann LeCun?

  Yeah, no. So so that that's contrastive learning, right? So you need positive and negative pairs. And the problem with that is that, you know, even though I had the original paper on this, I'm actually not very positive about it because it doesn't work in high dimension. If your representation is high dimensional, there's just too many ways for two things to be different. And and so you would need lots and lots and lots of negative pairs. So there is a particular implementation of this, which is relatively recent from actually the Google Toronto Group, where, you know, Jeff Hinton is the senior member there. It's called SIMCLR, S I M C L R. And it, you know, basically a particular way of implementing this idea of contrastive learning, the particular objective function.

• 57:30Yann LeCun?

  Now, what I'm much more enthusiastic about these days is non-contrasty methods. So other ways to guarantee that

• 57:40Yann LeCun?

  the representations would be different for different, uh, different inputs.

• 57:46Yann LeCun?

  And it's actually based on an idea that Jeff Hinton proposed in the early 90s with his student at the time, Sue Becker. And it's based on the idea of maximizing the mutual information between the outputs of the two systems. You only show positive pairs, you only show pairs of images that you know are somewhat similar. And you train the two networks to be informative, but also to be as informative of each other as possible. So basically, one representation has to be predictable from the other, essentially.

• 58:17Yann LeCun?

  And, you know, he proposed that idea had, you know, a couple of papers in the early 90s, and then nothing was done about it for decades. And I kind of revived this idea together with my postdocs at FAIR.

• 58:30Yann LeCun?

  Particularly a postdoc called Stefan Doni, who is now a junior professor in Finland at University of Alto.

• 58:37Yann LeCun?

  We came up with something called, that we called Barlow Twins. And it's a particular way of maximizing the information content of a vector, you know, using some hypotheses. And we have kind of another version of it that's more recent now called VICREG, V I C R E G. That that means variance, invariance, covariance, regularization. And it's the thing I'm the most excited about in machine learning in the last 15 years. I mean, I'm not, I'm really, really excited about this.

• 59:06Lex Fridman?

  What kind of data augmentation is useful for that non-contrastive learning method? Are we talking about, does that not matter that much? Or it seems like a very important part of the step.

• 59:17Yann LeCun?

  Yeah.

• 59:18Lex Fridman?

  How you generate the images that are similar but sufficiently different?

• 59:21Yann LeCun?

  Yeah, that's right. It's an important step, and it's also an annoying step because you need to have that knowledge of what data augmentation

• 59:28Yann LeCun?

  you can do that do not change the nature of the of the object. And so the standard scenario, which, you know, a lot of people working in this area are using, is you use the type of distortion. So so basically, you do geometric distortion. So one basically just shifts the image a little bit, it's called cropping. Another one kind of changes the scale a little bit. Another one kind of rotates it. Another one changes the colors. You know, you can do a shift in color balance or something like that. Saturation, another one sort of blurs it, another one adds noise. So you have like a catalog of kind of standard things and people try to use the same ones for different algorithms so that they can compare. But some algorithms, some self-supervised algorithms actually can deal with much bigger,

• 1:00:12Yann LeCun?

  like more aggressive data augmentation and some don't. So that kind of makes the whole thing difficult. But but that's the kind of distortions we're talking about. And so you you you you train with those distortions. And then you you chop off the last layer, a couple of layers of the of the the network. And you use the representation as input to a classifier. You train the classifier on ImageNet, let's say, or whatever, and measure the performance. And interestingly enough, the methods that are really good at eliminating the information that is irrelevant, which is the distortions between those images, do a good job at eliminating it.

• 1:00:54Yann LeCun?

  And as a consequence, you cannot use those the representations in those systems for things like object detection and localization, because that information is gone. So the type of data augmentation you need to do depends on the task you want eventually the system to to solve. And the type of data augmentation, standard data augmentation that we use today are only appropriate for object recognition or image classification. They're not appropriate for things like.

• 1:01:19Lex Fridman?

  Can you help me out and understand what why the localization is? So you're saying it's just not good at the negative, like classifying the negative? So that's why it can't be used for the localization?

• 1:01:29Yann LeCun?

  No, it's just that you train the system, you know, you you you give it an image and then you give it the same image shifted and scaled. You tell it that's the same image.

• 1:01:38Lex Fridman?

  Yeah.

• 1:01:39Yann LeCun?

  So the system basically is trained to eliminate the information about position and size. So now, and now you want to use that.

• 1:01:46Lex Fridman?

  Oh.

• 1:01:47Yann LeCun?

  To like figure out where an object is and what size it is.

• 1:01:49Lex Fridman?

  Like a bounding box, like to be able to actually, okay. It can still find, it can still find the object in the image, it's just not very good at finding the exact boundaries of that object. Interesting, interesting. Which, you know, that's an interesting sort of philosophical question. How important, how important is object localization anyway? We're like obsessed by measuring, like image segmentation, obsessed by measuring perfectly knowing the boundaries of objects when arguably.

• 1:02:19Lex Fridman?

  That's not that essential to understanding what are the contents of the scene.

• 1:02:25Yann LeCun?

  On the other hand, I think evolutionarily, the first vision systems in animals were basically all about localization, very little about recognition. And in the human brain, you have two separate pathways for recognizing the nature of a scene or an object and localizing objects. So you use the first pathway called eventual pathway for, you know, telling what you're looking at. The other pathway, the dorsal pathway, is used for navigation, for grasping, for everything else. And, you know, basically, a lot of the things you need for survival are localization and detection.

• 1:03:03Lex Fridman?

  Is similarity learning or contrastive learning, or these non-contrastive methods, the same as understanding something? Just because you know a distorted cat is the same as a non-distorted cat, does that mean you understand what it means to be a cat?

• 1:03:18Yann LeCun?

  To some extent. I mean, it's a superficial understanding, obviously.

• 1:03:22Lex Fridman?

  But like, what is the ceiling of this method, do you think? Is this just one trick on the path to doing self-supervised learning? Can we go really, really far?

• 1:03:32Yann LeCun?

  I think we can go really far. So if we figure out how to use techniques of that type, perhaps very different, but, you know, the same nature, to train a system from video, to do video prediction, essentially.

• 1:03:47Yann LeCun?

  I think we'll have a path, you know, towards, you know, I wouldn't say unlimited, but a path towards some level of, you know, physical common sense in machines. And I also think that that ability to learn how the world works from a sort of high-throughput channel like vision is a necessary step towards sort of real artificial intelligence. In other words, I believe in grounded intelligence. I don't think we can train a machine to be intelligent purely from text. Because I think the amount of information about the world that's contained in text is tiny compared to what we need to know.

• 1:04:33Yann LeCun?

  So for example, let's, and yeah, I know people have attempted to do this for thirty years, right, the SAG project and things like that, right, of basically kind of writing down all the facts that are known and hoping that some sort of common sense will emerge. I think it's basically hopeless. But let me take an example. You take an object. I describe a situation to you. I take an object, I put it on the table, and I push the table. It's completely obvious to you that the object will be pushed with the table, right? Because it's sitting on it. There is no text in the world, I believe, that explains this. And so if you train a machine as powerful as it could be, you know, your GPT-5000,

• 1:05:14Yann LeCun?

  or whatever it is, it's never going to learn about this. That information is just not present in any text.

• 1:05:22Lex Fridman?

  Well, the question, like with the SAG project, the dream, I think, is to have like, like ten million, say, facts like that, that give you a head start, like a parent guiding you. Now, we humans don't need a parent to tell us that the table will move, sorry, the smartphone will move with the table. But we get a lot of guidance in other ways. So it's possible that we can give it a quick shortcut.

• 1:05:50Yann LeCun?

  What about cat? The cat knows that.

• 1:05:52Lex Fridman?

  No, but they evolved, so.

• 1:05:54Yann LeCun?

  No, they learn like us.

• 1:05:56Lex Fridman?

  The, sorry, the physics of stuff?

• 1:05:59Yann LeCun?

  Yeah.

• 1:06:00Lex Fridman?

  Well, yeah, so you're saying it's, so you're putting a lot of intelligence onto the nurture side, not the nature.

• 1:06:09Yann LeCun?

  Yes.

• 1:06:10Lex Fridman?

  We seem to have, you know, there's a very inefficient, arguably, process of evolution that got us from bacteria to who we are today. Started at the bottom, now we're here. So the question is how, okay, so the question is how fundamental is that, the nature of the whole hardware? And then, is there any way to shortcut it if it's fundamental? If it's not, if it's most of intelligence, most of the cool stuff we've been talking about is mostly nurture, mostly trained, we figure it out by observing the world, we can form that big, beautiful, sexy background model that you're talking about just by sitting there. Then, okay, then you need to, then like maybe.

• 1:06:55Lex Fridman?

  It is all supervised learning all the way down. Self-supervised learning.

• 1:07:00Yann LeCun?

  Whatever it is that makes, you know, human intelligence different from other animals, which, you know, a lot of people think is language and logical reasoning and this kind of stuff. It cannot be that complicated because it only popped up in the last million years.

• 1:07:14Lex Fridman?

  Yeah.

• 1:07:16Yann LeCun?

  And, you know, it, it, and it only involves, you know, less than 1% of a genome, right? Which is the difference between human genome and chimps or whatever. So it can't be that complicated, you know, it can't be that fundamental. I mean, the, most of the so complicated stuff already exists in cats and dogs and, you know, certainly primates, non-human primates.

• 1:07:39Lex Fridman?

  Yeah, that little thing with humans might be just something about social interaction and ability to maintain ideas across like a collective of people. It's, it sounds very dramatic and very impressive, but it probably isn't mechanistically speaking.

• 1:07:55Yann LeCun?

  It is, but we're not there yet. Like, you know, we, we have, I mean, this is number 634, you know, in the list of problems we have to solve.

• 1:08:05Lex Fridman?

  So, basic physics of the world is, is, is number one. What do you, um, just a quick tangent on data augmentation. So, a lot of it is hardcoded versus learned. Do you have any intuition that maybe there could be some weird data augmentation? Like generative type of data augmentation, like doing something weird to images which then improves the, the similarity learning process? So not just kind of dumb, simple distortions, but by you shaking your head, just saying that even simple distortions are enough?

• 1:08:42Yann LeCun?

  I think, no, I think data augmentation is a temporary necessary evil.

• 1:08:47Lex Fridman?

  Huh.

• 1:08:48Yann LeCun?

  So what people are working on now is, is two things. One is the type of self-supervised learning.

• 1:08:55Yann LeCun?

  Like trying to to translate the type of self-supervised learning people use in language, translating these two images, which is basically a denoising autoencoder method. Right. So you you take an image, you you block, you mask some parts of it. Uh, and then you you train some giant neural net to reconstruct the parts that you've that are that are missing. And until very recently, there was no, there was no working methods for that. Uh, all the autoencoder type methods for images weren't producing very good representation. But there's a paper now coming out of the FAIR group at Menlo Park that actually works very well. So that doesn't require the augmentation, that requires only masking. Okay.

• 1:09:36Lex Fridman?

  Only masking for images. Uh, okay.

• 1:09:40Yann LeCun?

  Right. So you mask a part of the image and you train a system, which, you know, in this case is a transformer, because you you can, you can, the transformer represents the image as uh, non-overlapping patches. So it's easy to mask patches and things like that.

• 1:09:55Lex Fridman?

  Okay. Then my question transfers to that problem, the masking. Like, why should the mask be a square or rectangle?

• 1:10:01Yann LeCun?

  Yeah. So it doesn't matter. Like, you know, I think we're gonna come up probably in the future with sort of,

• 1:10:08Yann LeCun?

  you know, ways to mask that are, you know, kind of random, essentially. I mean, there are random already, but.

• 1:10:14Lex Fridman?

  No, no, but like something that's challenging, like optimally challenging. So like, I mean, maybe it's a metaphor that doesn't apply, but you're, it seems like there's a data augmentation or masking, there's an interactive element with it. Like, you're almost like playing with an image.

• 1:10:33Yann LeCun?

  Yeah.

• 1:10:34Lex Fridman?

  And like, it's like the way we play with an image in our minds.

• 1:10:37Yann LeCun?

  No, but it's like dropout. It's like bosom machine training. You, um, you know, every, every, every time you see a percept, you also, you you can, you can perturb it in some way. And then, uh, the the principle of the training procedure is to minimize the difference of the output of the representation between the the clean version and the corrupted version, essentially, right? And you you and you can do this in real time, right? So, you know, bosom machine work like this, right? You you you show a percept, uh, you tell the machine that's a good combination of activities of your input neurons. Uh, and then you either, uh,

• 1:11:17Yann LeCun?

  Let them go their merry way without clamping them to values, or you only do this with a subset. And what you're doing is you're training the system so that the the stable state of the entire network is the same regardless of whether it sees the entire input or whether it sees only part of it. You know, denoising autoencoder method is basically the same thing, right? You're you're training a system to reproduce the input, the complete input, and filling the blanks, regardless of which which parts are are missing. And that's really the underlying principle. And you could imagine sort of even in the brain, some sort of neural principle where, you know, neurons can oscillate, right? So they they take their activity and then temporarily they kind of shut off to, you know, force the rest of the system

• 1:12:01Yann LeCun?

  to basically reconstruct the input without their help, you know? And and and I mean, you you could imagine, you know, more or less biologically plausible processes.

• 1:12:13Lex Fridman?

  Something like that. And I guess with this denoising autoencoder and masking and data augmentation, you don't have to worry about being super efficient. You could just do as much as you want.

• 1:12:25Yann LeCun?

  Yeah.

• 1:12:25Lex Fridman?

  And and get better over time. Because I was thinking like you might want to be clever about the way you do all these procedures, you know? But that's only it's somehow costly to do every iteration, but it's not really.

• 1:12:39Yann LeCun?

  Not really. Maybe. And then there is, you know, data augmentation without explicit data augmentation. There's data augmentation by weighting, which is, you know, the the sort of video prediction.

• 1:12:51Yann LeCun?

  You're observing a video clip, observing the, you know, the the continuation of that video clip. And try you try to learn a representation using those joint embedding architectures in such a way that the representation of the future clip is easily predictable from the representation of the of the observed clip.

• 1:13:10Lex Fridman?

  Do you think YouTube has enough raw data from which to learn how to be a cat?

• 1:13:18Yann LeCun?

  I think so.

• 1:13:19Lex Fridman?

  So the amount of data is not the constraint.

• 1:13:23Yann LeCun?

  No, it would require some selection, I think.

• 1:13:26Lex Fridman?

  Some.

• 1:13:27Yann LeCun?

  Some selection of, you know, maybe the right type of data.

• 1:13:30Lex Fridman?

  Don't go down the rabbit hole of just cat videos that might, you might need to watch some lectures or something.

• 1:13:36Yann LeCun?

  No.

• 1:13:37Lex Fridman?

  How meta would that be if it like watches lectures about intelligence and then learns, watches your lectures in NYU and learns from that how to be intelligent?

• 1:13:48Yann LeCun?

  I don't think that would be enough.

• 1:13:52Lex Fridman?

  What's your, do you find multimodal learning interesting? We've been talking about visual language, like combining those together, maybe audio, all those kinds of things.

• 1:14:00Yann LeCun?

  There's a lot of things that I find interesting in the short term, but are not addressing the important problem that I think are really kind of the big challenges. So I think, you know, things like multitask learning, continual learning, you know,

• 1:14:15Yann LeCun?

  adversarial issues, I mean, those have, you know, great practical interests in the relatively short term, possibly, but I don't think they're fundamental, you know, active learning, even to some extent reinforcement learning. I think those things will become either obsolete or useless or easy once we figure out how to do self-supervised representation learning or learning predictive word models. And so I think that's what, you know, the entire community should be focusing on. At least people who are interested in sort of fundamental questions or, you know, really kind of pushing the envelope of AI towards the next, the next stage. But of course, there is like a huge amount of, you know, very interesting work to do in sort of practical questions that have, you know, short term impact.

• 1:15:00Lex Fridman?

  Well, you know, it's, it's difficult to talk about the temporal scale because all of human civilization will eventually be destroyed because the, the, the sun will die out. And even if Elon Musk has successful multi-planetary colonization across the galaxy, eventually the entirety of it would just become giant black holes and.

• 1:15:22Yann LeCun?

  That's gonna take a while though.

• 1:15:24Lex Fridman?

  So, but, but what I'm saying is then that logic can be used to say it's all meaningless. I'm saying all that to say that multitask learning. Might be you're calling it practical or pragmatic or whatever. That might be the thing that achieves something very akin to intelligence. While we're trying to solve the more general problem of self supervised learning of background knowledge. So the reason I bring that up, maybe one way to ask that question, I've been very impressed by what Tesla Autopilot team is doing. I don't know if you got any chance to glance at this particular one example of multitask learning, where they're literally taking the problem, like, I don't know, Charles Darwin

• 1:16:08Lex Fridman?

  starts studying animals, they're studying the problem of driving and asking, okay, what are all the things you have to perceive? And the way they're solving it is one, there's an ontology where you're bringing that to the table. So you're formulating a bunch of different tasks. It's like over 100 tasks or something like that, that they're involved in driving. And then they're deploying it and then getting data back from people that run into trouble. And they're trying to figure out, do we add tasks? Do we, like, we focus on each individual task separately?

• 1:16:37Yann LeCun?

  Sure.

• 1:16:38Lex Fridman?

  In fact, half, so the, I would say, I would classify Andre Karpathy's talking to as, so one was about doors and the other one about how much ImageNet sucks. He kept going back and forth on those two topics, which ImageNet sucks, meaning you can't just use a single benchmark. There's so, like, you have to have like a giant suite of benchmarks to understand how well your system actually works.

• 1:17:02Yann LeCun?

  I agree with him. I mean, he's a very sensible guy. Now, okay, it's very clear that if you're faced with a, an engineering problem that you need to solve in a relatively short time, particularly if you have Elon Musk breathing down your neck, you're going to have to take shortcuts, right? You, you, you might think about the, the fact that the, the right thing to do and the long term solution involves, you know, some fancy self supervisioning, but you have, you know, Elon Musk breathing down your neck and, you know, this involves, you know, human lives. And so you, you have to basically just do the systematic engineering and, you know,

• 1:17:44Yann LeCun?

  fine tuning and refinements and trial and error and all that stuff. There's nothing wrong with that. That's, that's called engineering. That's called, you know, Putting technology out in the in the world and and you have to kind of ironclad it before before you do this, you know.

• 1:18:05Yann LeCun?

  So much for, you know, grand grand ideas and principles.

• 1:18:10Yann LeCun?

  But, you know, I'm placing myself sort of, you know, some.

• 1:18:15Yann LeCun?

  You know, upstream of this, you know, quite a bit upstream of this.

• 1:18:17Lex Fridman?

  You're a Plato, think about platonic forms. You're.

• 1:18:21Yann LeCun?

  It's not platonic, because eventually I want that stuff to get used, but it's okay if it takes five or 10 years for the community to realize this is the right thing to do. I've I've done this before. It's been the case before that, you know, I've made that that case. I mean, if you look back in the mid 2000s, for example, and you ask yourself the question, okay, I want to recognize cars or faces or whatever.

• 1:18:46Yann LeCun?

  You know, I can use convolutional nets, so I can use sort of more conventional kind of computer vision techniques, you know, using interest point detectors or sift dense features and, you know, sticking an SVM on top. At that time, the data sets were so small that those methods that use more hand engineering worked better than ConvNets. There was just not enough data for ConvNets. And ConvNets were were a little a little slow with the kind of hardware that was available at the time. And there was a sea change when, basically, when, you know, data sets became bigger and and GPUs became available. That that's what, you know, the two of the main factors that basically made people change their change their mind.

• 1:19:29Yann LeCun?

  And you can you can look at the history of like all sub branches of AI or pattern recognition. And there is a similar trajectory followed by techniques where people start by, you know, engineering the hell out of it. You know, be it optical character recognition, speech recognition, computer vision, like image recognition in general, natural language understanding, like, you know, translation, things like that, right? You start to engineer the hell out of it. You start to acquire all the knowledge, the prior knowledge you know about image formation, about, you know, the shape of characters, about, you know, morphological operations, about like feature extraction, Fourier transforms,

• 1:20:14Yann LeCun?

  you know, Verneke moments, you know, whatever, right? People have come up with thousands of ways of representing images so that they could be easily classified afterwards. Same for speech recognition, right? There is, you know, it took decades for people to figure out a good font and to pre-process speech signals so that, you know, the information about what is being said is preserved, but most of the information about the identity of the speaker is gone. You know, Castrol coefficients or whatever, right? And same for text, right? You do named entity recognition and you parse and you you do tagging of the parts of speech and, you know, you do this sort of tree representation of clauses and all that stuff, right?

• 1:20:59Yann LeCun?

  Before you can do anything. So that's how it starts, right? Just engineer the hell out of it. And then you start having data and maybe you have more powerful computers. Maybe you know something about statistical learning. So you start using machine learning and it's usually a small sliver on top of your kind of handcrafted system where, you know, you extract features by hand. Okay, and now, you know, nowadays, the standard way of doing this is that you train the entire thing end to end with a deep learning system and it learns its own features and, and, you know, speech recognition systems nowadays, OCR systems are completely end to end. It's, you know, it's some giant neural net that takes raw waveforms and produces a sequence of characters coming out.

• 1:21:43Yann LeCun?

  And it's just a huge neural net, right? There's no, you know, Markov model, there's no language model that is explicit other than, you know, something that's ingrained in the, in the sort of neural language model, if you want. Same for translation, same for all kinds of stuff. So you see this continuous evolution from

• 1:22:01Yann LeCun?

  You know, less and less handcrafting and more and more learning. And I think it's true in biology as well.

• 1:22:12Lex Fridman?

  So, I mean, we might disagree about this, maybe not. In this one little piece at the end, you mentioned active learning. It feels like active learning, which is the selection of data and also the interactivity, needs to be part of this general network. You cannot just be an observer to do self-supervised learning. You have to, well, I don't, self-supervised learning is just a word, but I would, whatever this giant stack of a neural network that's automatically learning, it feels, my intuition is that you have to have a system, whether it's a physical robot or a digital robot,

• 1:22:52Lex Fridman?

  that's interacting with the world and doing so in a flawed way and improving over time. In order to do form the self-supervised learning well, you can't just give it a giant sea of data.

• 1:23:06Yann LeCun?

  Okay, I agree and I disagree.

• 1:23:08Lex Fridman?

  Okay.

• 1:23:09Yann LeCun?

  I agree in the sense that I think, I agree, I agree in two ways. The first, the first way I agree is that if you want, and you certainly need a causal model of the world that allows you to predict the consequences of your actions, to train that model, you need to take actions, right? You need to be able to act in a world and see the effect for you to be, to learn causal models of the world.

• 1:23:30Lex Fridman?

  Well, so that's not, that's not obvious because you can observe others.

• 1:23:33Yann LeCun?

  You can observe others.

• 1:23:34Lex Fridman?

  And you can infer that they're similar to you and then you can learn from that.

• 1:23:37Yann LeCun?

  Yeah, but then you have to kind of hardwire that part, right? And, you know, mirror neurons and all that stuff, right? So, and it's not clear to me how you would do this in a machine. So, so I think the, the action part would be necessary for having causal models of the world. The second reason it may be necessary, or at least more efficient, is that active learning basically, you know, goes for the jiggler of what you, what you don't know, right? Is, is, you know, obvious areas of uncertainty about your, your world and about the, how the world behaves. And you can resolve this uncertainty by systematic exploration of that part that

• 1:24:21Yann LeCun?

  you don't you don't know. And if you know that you don't know, then you know, it makes you curious. You kind of look into situations that and you know, across the animal world, different species at different levels of curiosity, right? Depending on how they're built, right? So, you know, cats and rats are incredibly curious, dogs not so much, I mean, less.

• 1:24:42Lex Fridman?

  Yeah, so it could be useful to have that kind of curiosity.

• 1:24:45Yann LeCun?

  So it'd be useful, but curiosity just makes the process faster. It doesn't make the process exist. The so what process, what learning process is it that active learning makes more efficient? And I'm asking that first question, you know, you know, we haven't answered that question yet. So, you know, I worry about active learning once this question is.

• 1:25:10Lex Fridman?

  So it's the more fundamental question to ask. And if active learning or interaction increases the efficiency of the learning, see, sometimes it becomes very different if the increase is several orders of magnitude, right? Like.

• 1:25:26Yann LeCun?

  That's true.

• 1:25:27Lex Fridman?

  But fundamentally, it's still the same thing and building up the intuition about how to in a self supervised way to construct background models, efficient or inefficient, is is the core problem. What do you think about Yosher Benjos talking about consciousness and all of these kinds of concepts?

• 1:25:46Yann LeCun?

  Okay, I don't know what consciousness is, but.

• 1:25:52Lex Fridman?

  It's a good opener.

• 1:25:53Yann LeCun?

  And to some extent, a lot of the things that are said about consciousness remind me of the questions people were asking themselves in the 18th century or 17th century when they discovered that, you know, how the eye works and the fact that the image at the back of the eye was upside down, right? Because you have a lens and and so on your retina, the image that forms is an image of the world, but it's upside down. How is it that you see right side up? And, you know, with what we know today in science, you know, we realize this question doesn't make any sense, or is kind of ridiculous in some way, right? So I think a lot of what is said about consciousness is of that nature. Now, that said, there is a lot of really smart people that for whom I have a lot of respect who are talking

• 1:26:36Yann LeCun?

  about this topic, people like David Chalmers, who is a colleague of mine at NYU.

• 1:26:41Yann LeCun?

  I have kind of a unorthodox folk speculative hypothesis about consciousness. So we're talking about the study of world model. And I think, you know, our entire prefrontal cortex basically is the engine for a world model. But when we are attending at a particular situation, we're focused on that situation. We basically cannot attend to anything else. And that seems to suggest that we basically have only one world model engine in our prefrontal cortex. That engine is configurable to the situation at hand.

• 1:27:24Yann LeCun?

  So we are building a box out of wood or we are, you know, driving down the highway, playing chess. We, we, we basically have a single model of the world that we're configuring to the situation at hand, which is why we can only attend to one task at a time. Now, if there is a task that we do repeatedly, it, it goes from the sort of deliberate reasoning using model of the world and prediction, and perhaps something like model predictive control, which I was talking about earlier, to something that is more subconscious, that becomes automatic. So I don't know if you've ever played against a chess grandmaster. You know, I get wiped out in, you know, 10, 10 plays, right?

• 1:28:05Yann LeCun?

  And, you know, I have to think about my move for, you know, like 15 minutes. And the person in front of me, the grandmaster, you know, would just like react within seconds, right? You know, he doesn't need to think about it. That's become part of subconscious because, you know, it's basically just pattern recognition at this point. Same, you know, you, the first few hours, you drive a car, you're really attentive, you can't do anything else. And then after 20, 30 hours of practice, 50 hours, you know, it's subconscious. You can talk to the person next to you, you know, things like that, right? Unless the situation becomes unpredictable and then you have to stop talking. So that suggests you only have one model in your head. And it might suggest the idea that consciousness basically is the module

• 1:28:50Yann LeCun?

  that configures this world model of yours. You know, you need to have some sort of executive kind of overseer that configures your world model for the situation at hand. And that that leads to kind of the really curious concept that consciousness is not a consequence of the power of our minds, but of the limitation of our brains. But because we have only one world model, we have to be conscious. If we had as many world models as there are situations we encounter, then we could do all of them simultaneously, and we wouldn't need this sort of executive control that we call consciousness.

• 1:29:26Lex Fridman?

  Yeah, interesting. And somehow, maybe that executive controller, I mean, the hard problem of consciousness, there's some kind of chemicals in biology that's creating a feeling, like it feels to experience some of these things. That's kind of like the hard question is, what the heck is that? And why is that useful? Maybe the more pragmatic question, why is it useful to feel like this is really you experiencing this versus just like information being processed? It could be just a very nice side effect of the way we evolved. That's just very useful to feel a sense of ownership

• 1:30:10Lex Fridman?

  to the decisions you make, to the perceptions you make, to the model you're trying to maintain. Like you own this thing, and it's the only one you got. And if you lose it, it's going to really suck. And so you should really send the brain some signals about it. What ideas do you believe might be true that most, or at least many people disagree with you with? Let's say in the space of machine learning.

• 1:30:35Yann LeCun?

  Well, it depends who you're talking about. But I think.

• 1:30:39Yann LeCun?

  So certainly, there is a bunch of people who are nativist, right? Who think that a lot of the basic things about the world are kind of hardwired in our minds. Things like, you know, the world is three dimensional, for example. Is that hardwired? Things like, you know, object permanence. Is it something that we learn, you know, before the age of three months or so? Or are we born with it? And there are, you know, very, you know, wide disagreements among the cognitive scientists for this. I think those things are actually very simple to learn. You know, is it the case that the oriented edge detectors in V1 are learned? Or are they hardwired? I think they are learned. They might be learned before birth, because it's really easy to generate signals from the retina

• 1:31:23Yann LeCun?

  that actually will train edge detectors. So, and again, those are things that can be learned within minutes of opening your eyes, right? I mean, you know, since the 1990s, we have algorithms that can learn oriented edge detectors completely unsupervised with the equivalent of a few minutes of real time. So, so those things have to be learned. And there's also those, you know, MIT experiments where you kind of plug the optical nerve on the auditory cortex of a baby ferret, right? And that auditory cortex becomes a visual cortex, essentially. So, you know, clearly, there's learning taking place there. So, you know, I think a lot of what people think are so basic that they need to be hardwired, I think a lot of those things are learned because they are easy to learn.

• 1:32:07Lex Fridman?

  So you put a lot of value in the power of learning. What kind of things do you suspect might not be learned? Is there something that could not be learned?

• 1:32:18Yann LeCun?

  So your intrinsic drives are not learned. There are the things that, you know, make humans human or make, you know, cats different from dogs, right? It's the basic drives that are kind of hardwired in our basal ganglia. I mean, there are people who are working on this kind of stuff. It's called intrinsic motivation in the context of reinforcement learning. So these are objective functions where the reward doesn't come from the external world. It's computed by your own brain. Your own brain computes whether you're happy or not, right? It measures your degree of comfort or uncomfort. And, and because it's your brain computing this, presumably it knows also how to estimate gradients of this, right?

• 1:33:00Yann LeCun?

  So, so it's easier to learn when your objective is intrinsic.

• 1:33:08Yann LeCun?

  So that has to be hardwired. The critic that makes long-term prediction of the outcome, which is the eventual result of this, that's learned. And perception is learned and your model of the world is learned. But let me take an example of why the critic, I mean, example of how the critic may be learned, right? If I, if I come to you, you know, I reach across the table and I pinch your arm, right? Complete surprise for you. You would not have expected this from me.

• 1:33:38Lex Fridman?

  I was expecting that the whole time, but yes, right. Let's say for the sake of the story, yes.

• 1:33:42Yann LeCun?

  Okay, your bezel ganglia is gonna light up because it's gonna hurt, right? And now your model of the world includes the fact that I may pinch you if I approach my, my...

• 1:33:56Lex Fridman?

  Don't trust humans.

• 1:33:58Yann LeCun?

  Right. My hand to your arm, so if I try again, you're gonna recoil. And that's your critic, your predictive, you know, your predictor of your ultimate pain system that predicts that something bad is gonna happen and you recoil to avoid it.

• 1:34:15Lex Fridman?

  So even that can be learned?

• 1:34:16Yann LeCun?

  That is learned, definitely. This is what allows you also to, you know, define some goals, right? So the fact that, you know, you're a school child, you wake up in the morning and you go to school and, you know, it's not because you necessarily like waking up early and going to school, but you know that there is a long-term objective you're trying to optimize.

• 1:34:37Lex Fridman?

  So Ernest Becker, I'm not sure if you're familiar with him, the philosopher, he wrote the book Denial of Death, and his idea is that one of the core motivations of human beings is our terror of death, our fear of death. That's what makes us unique from cats. Cats are just surviving. They do not have a deep, like, cognizance, introspection that...

• 1:35:01Lex Fridman?

  Over the horizon is the end. And he says that, I mean, there's a terror management theory that just all these psychological experiments that show that basically this idea that all of human civilization, everything we create is kind of trying to forget if even for a brief moment that we're going to die. When, when do you think humans understand that they're going to die? Is it learned early on also? Like?

• 1:35:30Yann LeCun?

  I don't know to at what point, I mean, it's a, it's a question like, you know, at what point do you realize that, you know, what death really is? And I think most people don't actually realize what death is, right? I mean, most people believe that you go to heaven or something, right?

• 1:35:43Lex Fridman?

  So the, the pushback on that, what Ernest Becker says and Sheldon Solomon, all of those folks, and I find those ideas a little bit compelling is that there is moments in life, early in life, a lot of this fun happens early in life when you are, when you do deeply experience the terror of this realization and all the things you think about, about religion, all those kinds of things that we kind of think about more like teenage years and later, we're talking about way earlier.

• 1:36:14Yann LeCun?

  No, it's like seven or eight years, something like that.

• 1:36:15Lex Fridman?

  You realize, holy crap, this is like the mystery, the terror, like it's almost like you're a little prey, a little baby deer sitting in the darkness of the jungle or the woods, looking all around you, there's darkness full of terror. I mean, that's, that realization says, okay, I'm going to go back in the comfort of my mind where there's a, where there is a deep meaning, where there's a maybe like pretend I'm immortal and however way, however kind of idea I can construct to help me understand that I'm immortal. Religion helps with that. You can, you can delude yourself in all kinds of ways, like lose yourself in the busyness of each day, have little goals in mind, all those kinds of things to think that it's going to go on forever.

• 1:37:00Lex Fridman?

  And you kind of know you're going to die, yeah, and it's going to be sad, but you don't really understand that you're going to die. And so that's, that's their idea. And I find that compelling. Because it does seem to be a core, unique aspect of human nature that we were able to think that we're going, we're able to really understand that this life is finite. That seems important.

• 1:37:22Yann LeCun?

  There's, there's a bunch of different things there. So first of all, I don't think there is a qualitative difference between, between us and cats in the term. I think the difference is that we just have a better long-term ability to predict, you know, in the long term. And so we have a better understanding of how the world works. So we have better understanding of, you know, finiteness of life and things like that.

• 1:37:42Lex Fridman?

  So we have a better planning engine than cats?

• 1:37:45Yann LeCun?

  Yeah.

• 1:37:46Lex Fridman?

  Okay. But what's the motivation for, for planning that far?

• 1:37:50Yann LeCun?

  Well, I think it's just a side effect of the fact that we have just a better planning engine because it makes us, as I said, you know, the essence of intelligence is the ability to predict. And so the, because we're smarter, as a side effect, we also have this ability to kind of make predictions about our own future existence or lack thereof.

• 1:38:09Lex Fridman?

  Okay.

• 1:38:10Yann LeCun?

  You say religion helps with that. I think religion hurts, actually. It makes people worry about like, you know, what's going to happen after their death, et cetera. If you believe that, you know, you just don't exist after death. Like, you know, it solves completely the problem, at least.

• 1:38:24Lex Fridman?

  You're saying if you don't believe in God, you don't worry about what happens after death?

• 1:38:29Yann LeCun?

  Yeah.

• 1:38:30Lex Fridman?

  I don't know.

• 1:38:30Yann LeCun?

  You only worry about, about, you know, this life because that's the only one you have.

• 1:38:36Lex Fridman?

  I think it's, well, I don't, I don't know. If I were to say what Ernest Becker says, and I would say I agree with him more

• 1:38:43Lex Fridman?

  than not, is you do deeply worry. If you, if you believe there's no God, there's still a deep worry, like of the mystery of it all. Like, how does that make any sense? That it just ends. I don't think we can truly understand that this right, I mean, so much of our life, the consciousness, the ego is invested in this, in this being. And then

• 1:39:09Yann LeCun?

  Science keeps bringing humanity down from its pedestal. And that's just another example of it.

• 1:39:16Lex Fridman?

  That's wonderful, but for us individual humans, we don't like to be brought down from a pedestal. You're saying like, but see, you're fine with it because, well, so what Ernest Becker would say is you're fine with it because that's just a more peaceful existence for you, but you're not really fine. You're hiding from, in fact, some of the people that experienced the deepest trauma earlier in life, they often, before they seek extensive therapy, will say that I'm fine. It's like when you talk to people who are truly angry, how are you doing? I'm fine. The question is, what's going on?

• 1:39:49Yann LeCun?

  I had a near death experience. I had a very bad motorbike accident when I was 17. So, but that didn't have any impact on my reflection on that topic.

• 1:40:02Lex Fridman?

  So I'm basically just playing a bit of devil's advocate, pushing back and wondering, is it truly possible to accept death? And the flip side, that's more interesting, I think, for AI and robotics is how important is it to have this as one of the suite of motivations is to not just avoid falling off the roof or something like that, but ponder the, the end of the ride. If you listen to the Stoics, it's, it's a great motivator. It adds a sense of urgency. So maybe to truly fear death or be cognizant of it

• 1:40:43Lex Fridman?

  might give a deeper meaning and urgency to the moment to live fully.

• 1:40:50Yann LeCun?

  Maybe I don't disagree with that. I mean, I think what motivates me here is,

• 1:40:58Yann LeCun?

  you know, knowing more about, about human nature. I mean, I think human nature and human intelligence is a big mystery. It's a scientific mystery in addition to, you know, philosophical and et cetera. But, you know, I'm a true believer in science, so

• 1:41:15Yann LeCun?

  and, and, and I do have kind of a belief that for complex systems like, like the brain and the mind. The the way to understand it is to try to reproduce it with, you know, artifacts that you build, because you know what's essential to it when you try to build it. You know, the same way, I've used this analogy before with you, I believe. The same way, we we only started to understand aerodynamics when we started building airplanes, and that helped us understand how birds fly. You know, so I I think there's kind of a similar process here where we don't have a theory of a full theory of intelligence, but building, you know, intelligent artifacts will help us perhaps develop some, you know, underlying theory that encompasses not just artificial

• 1:42:00Yann LeCun?

  implements, but also human and biological intelligence in general.

• 1:42:05Lex Fridman?

  So you're an interesting person to ask this question about sort of all kinds of different other intelligent entities or intelligences. What are your thoughts about kind of like the touring or the Chinese room question?

• 1:42:21Lex Fridman?

  If we create an AI system that exhibits a lot of properties of intelligence and consciousness, how comfortable are you thinking of that entity as intelligent or conscious? So so you're trying to build now systems that have intelligence and there's metrics about their performance, but that metric is external.

• 1:42:44Yann LeCun?

  Okay.

• 1:42:44Lex Fridman?

  So how are you, are you okay calling a thing intelligent? Or are you going to be like most humans and be, once again, unhappy to be brought down from a pedestal of consciousness slash intelligence?

• 1:42:56Yann LeCun?

  No, I'm I'll be very happy to understand more about human nature, human mind, and human intelligence through the construction of machines that have similar abilities. And if a consequence of this is to bring down humanity one notch down from its already low pedestal, I'm just fine with it. That's just the reality of life. So I'm I'm fine with that. Now, you're asking me about things that, opinions I have that a lot of people may disagree with. I think.

• 1:43:33Yann LeCun?

  If we think about the design of autonomous intelligence system, so assuming that we are somewhat successful at some level of getting machines to learn models of the world, predictive models of the world. We have, we build intrinsic motivation, objective functions to drive the behavior of that system. The system also has perception modules that allows it to estimate the state of the world and then have some way of figuring out a sequence of actions that, you know, to optimize a particular objective. If it has a critic of the type that I was describing before, the thing that makes you recoil your arm the second time I try to pinch you. Intelligent autonomous machine will have emotions. I think emotions are an integral part of autonomous intelligence.

• 1:44:18Yann LeCun?

  If you have an intelligent system that is driven by intrinsic motivation, by objectives, if it has a critic that allows it to predict in advance whether the outcome of a, of a situation is going to be good or bad, it's going to have emotions, it's going to have fear.

• 1:44:35Lex Fridman?

  Yes.

• 1:44:36Yann LeCun?

  When it predicts that the outcome is going to, is going to be bad and, and something to avoid, it's going to have elation when it predicts it's going to be good. If it has drives to relate with humans, you know, in some ways, the way humans have, you know, it's, it's going to be social, right? And so it's going to have emotions about attachment and, and things of that type. So, so I think, you know, the, the sort of sci-fi thing where, you know, you see commander Data like having an emotion chip that he can turn off, right? I think that's ridiculous.

• 1:45:13Lex Fridman?

  So, I mean, here's the difficult philosophical, social question. Do you think there will be a time, like a civil rights movement for robots, where, okay, forget the movement, but a discussion, like the Supreme Court?

• 1:45:31Lex Fridman?

  That particular kinds of robots, you know, particular kinds of systems

• 1:45:38Lex Fridman?

  deserve the same rights as humans because they can suffer just as humans can. All those kinds of things.

• 1:45:46Yann LeCun?

  Well, perhaps, perhaps not. Like, imagine that humans were, that you could, you know, die and be restored. Like, you know, you could be sort of, you know, be 3D reprinted and, you know, your brain could be reconstructed in its finest details.

• 1:46:02Lex Fridman?

  Our ideas of rights will change in that case. If you can always just, there's always a backup. You could always restore. Maybe like the importance of murder will go down one notch.

• 1:46:13Yann LeCun?

  That's right. But also, the, your, your, you know, desire to do dangerous things like, you know, doing skydiving or, or, you know,

• 1:46:25Yann LeCun?

  or, or, you know, race car driving, you know, car racing or that kind of stuff, you know, would probably increase. Or, or, you know, airplane robotics or that kind of stuff, right? It would be fine to do a lot of those things or explore, you know, dangerous areas and things like that. It would kind of change your relationship. So now it's very likely that robots would be like that because, you know, they'll be based on perhaps technology that is somewhat similar to, to this technology and you can, you can always have a backup.

• 1:46:54Lex Fridman?

  So it's possible, I don't know if you like video games, but there's a, there's a game called Diablo and.

• 1:47:01Yann LeCun?

  Oh, my, my sons are huge fans of this.

• 1:47:03Lex Fridman?

  Yes.

• 1:47:06Yann LeCun?

  In fact, they made a game that's inspired by it.

• 1:47:09Lex Fridman?

  Awesome. Like built a game?

• 1:47:11Yann LeCun?

  My, my three sons have a game design studio between them. Yeah.

• 1:47:14Lex Fridman?

  That's awesome.

• 1:47:15Yann LeCun?

  They came out with a game.

• 1:47:16Lex Fridman?

  Like it just came out?

• 1:47:17Yann LeCun?

  Last year. No, this was last year, early last year, about a year ago.

• 1:47:20Lex Fridman?

  That's awesome. But so in Diablo, there's a something called hardcore mode, which if you die, there's no. You're gone.

• 1:47:27Yann LeCun?

  Right.

• 1:47:27Lex Fridman?

  That's it. And so it's possible with AI systems for them to be able to operate successfully and for us to treat them in a certain way for because they have to be integrated in human society. They have to be able to die, no copies allowed. In fact, copying is illegal. It's possible with humans as well, like cloning will be illegal, even what's possible.

• 1:47:50Yann LeCun?

  Uh, but cloning is not copying, right? I mean, you don't reproduce the the mind of the person and experience.

• 1:47:55Lex Fridman?

  Right.

• 1:47:56Yann LeCun?

  It's just a delayed twin, so.

• 1:47:58Lex Fridman?

  But then it's, but we were talking about with computers that you will be able to copy.

• 1:48:02Yann LeCun?

  Right.

• 1:48:02Lex Fridman?

  You'll be able to perfectly save pickle the the the mind state. And it's possible that that will be illegal because that goes against that that will destroy the motivations of the system.

• 1:48:17Yann LeCun?

  Okay, so let's say you you have a domestic robot.

• 1:48:21Lex Fridman?

  Yeah.

• 1:48:21Yann LeCun?

  Okay, sometime in the future.

• 1:48:23Lex Fridman?

  Yes.

• 1:48:24Yann LeCun?

  And a domestic robot, you know, comes to you kind of somewhat pre-trained. You know, it can do a bunch of things.

• 1:48:30Lex Fridman?

  Yes.

• 1:48:30Yann LeCun?

  But it has a particular personality that makes it slightly different from the other robots because that makes them more interesting. And then because it's, you know, it's lived with you for five years. You've you've grown some attachment to it and vice versa. And it's learned a lot about you. Or maybe it's not a household robot. Maybe it's a maybe it's a virtual assistant that lives in your, you know, augmented reality glasses or whatever, right?

• 1:48:53Lex Fridman?

  Yes.

• 1:48:54Yann LeCun?

  Uh, you know, the the her movie type thing, right?

• 1:48:58Yann LeCun?

  And that system, to some extent, the the intelligence in that system is a bit like your child or maybe your PhD student in the sense that there's a lot of you in that in that machine now, right?

• 1:49:11Lex Fridman?

  Yeah.

• 1:49:12Yann LeCun?

  And so if it were a living thing, you you would do this for free if you want, right? If it's your child, your child can, you know, then live his or her own own life. And, you know, the fact that they learn stuff from you doesn't mean that you have any ownership of it, right?

• 1:49:28Lex Fridman?

  Yeah.

• 1:49:28Yann LeCun?

  But if it's a robot that you've trained, perhaps you have some

• 1:49:33Lex Fridman?

  Yeah.

• 1:49:34Yann LeCun?

  intellectual property claim about.

• 1:49:36Lex Fridman?

  Intellectual property? Oh, I thought you meant like a permanent value in the sense that's part of you is in.

• 1:49:42Yann LeCun?

  Well, there is permanent value, right? So you would lose a lot if that robot were to be destroyed and you you had no backup. You would lose a lot, you lose a lot of investment, you know, kind of like a. You know, a person dying, you know, that a friend of yours dying or a coworker or something like that.

• 1:50:00Lex Fridman?

  But also you have like intellectual property rights in the sense that that that system is fine tuned to your particular existence. So that's now a very unique instantiation of that original background model, whatever it was that arrived.

• 1:50:16Yann LeCun?

  And then there are issues of privacy, right? Because now imagine that that robot has its own kind of volition and decides to work for someone else.

• 1:50:24Lex Fridman?

  Yes.

• 1:50:24Yann LeCun?

  Or kind of, you know, thinks life with you is sort of untenable or whatever.

• 1:50:29Lex Fridman?

  Right.

• 1:50:31Yann LeCun?

  Now, all the things that that system learned from you,

• 1:50:36Yann LeCun?

  you know, how can you like, you know, delete all the personal information that that system knows about you?

• 1:50:41Lex Fridman?

  Yeah.

• 1:50:42Yann LeCun?

  I mean, that would be kind of an ethical question, like, you know, can you erase the the mind of a intelligent robot to protect your your privacy?

• 1:50:51Lex Fridman?

  Yeah.

• 1:50:52Yann LeCun?

  You can't do this with humans. You can ask them to shut up, but that you don't have complete power over them.

• 1:50:57Lex Fridman?

  Can't erase humans. Yeah, it's the problem with the relationships, you know, that you break up, you can't, you can't erase the other human. With robots, I think it will have to be the same thing with robots, that that risk that there has to be some risk to our interactions to truly experience them deeply, it feels like. So you have to be able to lose your robot friend and that robot friend to go tweeting about how much of an asshole you are.

• 1:51:25Yann LeCun?

  But then are you allowed to, you know, murder the robot to protect your private information?

• 1:51:30Lex Fridman?

  Yeah, probably not.

• 1:51:31Yann LeCun?

  If the robot decides to leave?

• 1:51:31Lex Fridman?

  I have this intuition that for robots with with certain, like, it's almost like a regulation. If you declare your robot to be, let's call it sentient or something like that, like, this this robot is designed for human interaction, then you're not allowed to murder these robots. It's the same as murdering other humans.

• 1:51:50Yann LeCun?

  Well, but what about you do a backup of the robot that you preserve on a on a hard drive or the equivalent in the future?

• 1:51:55Lex Fridman?

  That might be illegal. It's like it's like piracy, piracy, piracy is illegal.

• 1:52:00Yann LeCun?

  But it's your own, it's your own robot, right?

• 1:52:01Lex Fridman?

  But you can't, you don't.

• 1:52:03Yann LeCun?

  But then, but then you can wipe out.

• 1:52:06Yann LeCun?

  His brain, so the this robot doesn't know anything about you anymore. But you still have technically it's still in existence because you backed it up.

• 1:52:13Lex Fridman?

  And then there'll be these great speeches at the Supreme Court by saying, oh, sure, you can erase the mind of the robot, just like you can erase the mind of a human. We both can suffer. There'll be some epic like Obama type character with a speech that we we like the robots and the humans are the same. We can both suffer. We can both hope. We can both all of those, all those kinds of things. Raise families, all that kind of stuff. It's it's interesting for these, just like you said, emotion seems to be a fascinatingly powerful aspect of human, human interaction, human robot interaction. And if they're able to exhibit emotions at the end of the day, that's probably going to have us deeply consider

• 1:52:57Lex Fridman?

  human rights, like what we value in humans, what we value in other animals. That's why robots and AI is great. It makes us ask really good questions.

• 1:53:06Yann LeCun?

  The hard questions. Yeah. But you you ask about you asked about the Chinese room type argument, you know, is it real? If it looks real?

• 1:53:12Lex Fridman?

  Yeah.

• 1:53:13Yann LeCun?

  I think the Chinese room argument is a ridiculous one. So.

• 1:53:17Lex Fridman?

  So so for people who don't know Chinese room is you can, I don't even know how to formulate it well, but basically you can mimic the behavior of an intelligent system by just following a giant algorithm code book that tells you exactly how to respond in exactly each case. But is that really intelligent? It's like a giant lookup table. When this person says this, you answer this. When this person says this, you answer this. And if you understand how that works, you have this giant, nearly infinite lookup table. Is that really intelligence? Because intelligence seems to be a mechanism that's much more interesting and complex than this lookup table.

• 1:53:56Yann LeCun?

  I don't think so. So the I mean, the real question comes down to, do you think, you know, you can you can mechanize intelligence in some way, even if that involves learning? And the answer is, of course, yes, there's no question. There's a second question then, which is, assuming you can reproduce intelligence in sort of different hardware than biological hardware, you know, like computers.

• 1:54:24Yann LeCun?

  Can you, you know, match human intelligence in?

• 1:54:30Yann LeCun?

  All the domains in which humans are intelligent. Is it possible, right? So that's the hypothesis of a strong AI. The answer to this, in my opinion, is an unqualified yes. This will well happen at some point. There's no question that machines at some point will become more intelligent than humans in all domains where humans are intelligent. This is not for tomorrow. It's going to take a long time, regardless of what, you know, Elon and others have claimed or believed. This is a lot harder than many of those guys think it is. And many of those guys who thought it was simpler than that years, you know, five years ago, now think it's hard because it's been five years and they realize it's going to take a lot longer.

• 1:55:15Yann LeCun?

  That includes a bunch of people at DeepMind, for example, but.

• 1:55:18Lex Fridman?

  Oh, interesting. I haven't actually touched base with the DeepMind folks, but some of it, Elon or Demis Hassabis, I mean, sometimes in your role, you have to kind of create deadlines that are nearer than farther away to kind of create an urgency. Because, you know, you have to believe the impossible is possible in order to accomplish it. And there's, of course, a flip side to that coin, but it's a weird. You can't be too cynical if you want to get something done.

• 1:55:44Yann LeCun?

  Absolutely. I agree with that. But, I mean, you have to inspire people, right? To work on sort of ambitious things. So, you know, it's certainly a lot harder than we believe, but there's no question in my mind that this will happen. And now, you know, people are kind of worried about what does that mean for humans? They are going to be brought down from their pedestal, you know, a bunch of notches with that. And, you know, is that going to be good or bad? I mean, it's just going to give more power, right? It's an amplifier for human intelligence, really.

• 1:56:18Lex Fridman?

  So, speaking of doing cool, ambitious things, FAIR, the Facebook AI research group, has recently celebrated its eighth birthday. Or. Maybe you can correct me on that. Looking back, what has been the successes, the failures, the lessons learned from the eight years of FAIR? And maybe you can also give context of where does the newly minted Meta AI fit into how does it relate to FAIR?

• 1:56:44Yann LeCun?

  Right, so let me tell you a bit about the organization of all this. Yeah, FAIR was created almost exactly eight years ago. It wasn't called FAIR yet. It took that name a few, a few months later. And at the time I joined Facebook, there was a group called AI Group that had about 12 engineers and a few scientists, like, you know, 10 engineers and two scientists or something like that. I ran it for three and a half years as a director. You know, hired the first few scientists and kind of set up the culture and organized it, you know, explained to the Facebook leadership what fundamental research was about and how it can work within industry and how it needs to be open and everything.

• 1:57:29Yann LeCun?

  And I think it's been an unqualified success in the sense that FAIR has simultaneously produced, you know, top level research and advanced the science and the technology, provided tools, open source tools like PyTorch and many others.

• 1:57:48Yann LeCun?

  But at the same time has had a direct or mostly indirect impact on Facebook at the time, now Meta. In the sense that a lot of systems that are, that Meta is built around now are, are based on research projects that started at FAIR. And so if you were to take out, you know, deep learning out of Facebook services now and Meta more generally, I mean, the company would literally crumble. I mean, it's completely built around, around AI these days. And it's really essential to the operations. So what happened after three and a half years is that I changed role, I became chief scientist.

• 1:58:32Yann LeCun?

  So I'm not doing day to day management of FAIR anymore. I'm more of a kind of, you know, think about strategy and things like that. And I carry my, I conduct my own research. I have, you know, my own kind of research group, working on self-supervised learning and things like this, which I didn't have time to do when I was director.

• 1:58:50Yann LeCun?

  So now FAIR is run by Joel Pinault and Antoine Bord together. Because FAIR is kind of split in two now. There's something called FAIR Labs, which is sort of bottom-up, science-driven research. And FAIR Excel, which is slightly more organized for bigger projects that require a little more kind of focus and more engineering support and things like that. So Joel needs FAIR Lab and Antoine Bord leads FAIR Excel.

• 1:59:14Lex Fridman?

  Where are they located?

• 1:59:16Yann LeCun?

  It's de-located all over. So there's no question that the leadership of the company believes that this was a very worthwhile investment. And what that means is that it's there for the long run. Right, so there is, if you want to talk in these terms, which I don't like, there's a business model, if you want, where FAIR, despite being a very fundamental research lab, brings a lot of value to the company, either mostly indirectly through other groups.

• 1:59:51Yann LeCun?

  Now, what happened three and a half years ago when I stepped down is, was also the creation of Facebook AI, which was basically a larger organization that covers FAIR, so FAIR is included in it, but also has other organizations that are focused on applied research or advanced development of AI technology that is more, you know, focused on the products of the company.

• 2:00:16Lex Fridman?

  So less emphasis on fundamental research.

• 2:00:18Yann LeCun?

  Less fundamental, but it's still research. I mean, there's a lot of papers coming out of those organizations and the people are awesome and, you know, wonderful to interact with. And, but it serves as kind of... A way to kind of scale up if you want sort of AI technology, which you know may be very experimental and and sort of lab prototypes into things that are usable.

• 2:00:42Lex Fridman?

  So FAIR is a subset of Meta AI. Is FAIR become like KFC? It'll just keep the F. Nobody cares what the F stands for?

• 2:00:51Yann LeCun?

  Will know soon enough by probably probably by the end of the of twenty twenty one.

• 2:00:57Lex Fridman?

  This is not a giant change, Mayor. FAIR.

• 2:01:00Yann LeCun?

  Well, Mayor doesn't sound too good, but you know, the the brand people are kind of deciding on this. And they've been hesitating for for a while now and they, you know, they tell us they're going to come up with an answer as to whether FAIR is going to change name or whether we're going to change just the meaning of the F.

• 2:01:15Lex Fridman?

  That's a good call. I will keep FAIR and change the meaning of the F.

• 2:01:18Yann LeCun?

  That would be my preference, you know, I would tend, I would turn the F into fundamental.

• 2:01:22Lex Fridman?

  Oh, that's good.

• 2:01:23Yann LeCun?

  AI research.

• 2:01:24Lex Fridman?

  Oh, that's really good. Yeah.

• 2:01:25Yann LeCun?

  Within Meta AI. So this would be

• 2:01:27Lex Fridman?

  Meta FAIR.

• 2:01:27Yann LeCun?

  Meta FAIR.

• 2:01:28Lex Fridman?

  Yeah.

• 2:01:28Yann LeCun?

  But you know, people will call it FAIR, right?

• 2:01:30Lex Fridman?

  Yeah, exactly. I like it.

• 2:01:31Yann LeCun?

  And now Meta AI is part of the Reality Lab. So, you know,

• 2:01:41Yann LeCun?

  Meta now, the new Facebook, it's called Meta, and it's kind of divided into, you know, Facebook, Instagram, WhatsApp.

• 2:01:52Yann LeCun?

  And

• 2:01:54Yann LeCun?

  Reality Lab. And Reality Lab is about, you know, AR, VR, you know, telepresence, communication technology and stuff like that. It's kind of the, you can think of it as the sort of a combination of sort of new products and and technology part of of Meta.

• 2:02:13Lex Fridman?

  Is that where the touch sensing for robots I saw that you were posting about? That's

• 2:02:17Yann LeCun?

  But touch sensing for robots is part of FAIR, actually. That's that's a

• 2:02:20Lex Fridman?

  Oh, it is. Okay, cool.

• 2:02:22Yann LeCun?

  This is also the, no, but there is the the other way, the the haptic glove, right?

• 2:02:27Lex Fridman?

  Yes.

• 2:02:28Yann LeCun?

  That has like

• 2:02:28Lex Fridman?

  That's more Reality Lab.

• 2:02:29Yann LeCun?

  That's that's Reality Lab research.

• 2:02:32Lex Fridman?

  Yeah, the lab research. By the way, the touch sensors is super interesting. Like integrating that modality into the whole sensing suite is very interesting. So what do you think about the metaverse? What do you think about this whole kind of expansion of the view of the role of Facebook and Meta in the world?

• 2:02:52Yann LeCun?

  Well, Metaverse really should be thought of as the next step in the internet, right? Sort of trying to kind of, you know, make the experience more compelling of, you know, being connected either with other people or with content.

• 2:03:11Yann LeCun?

  And, you know, we are evolved and trying to evolve in, you know, 3D environments where, you know, we can see other people, we can talk to them when we're near them or, you know, and other viewers are far away, can't hear us, you know, things like that, right? So there's a lot of social conventions that exist in the real world that we can try to transpose. Now, what is going to be eventually the... How compelling is it going to be? Like, you know, is it going to be the case that people are going to be willing to do this if they have to wear, you know, a huge pair of goggles all day? Maybe not.

• 2:03:48Lex Fridman?

  But then again, if the experience is sufficiently compelling, maybe so.

• 2:03:52Yann LeCun?

  Or if the device that you have to wear is just basically a pair of glasses, you know, technology makes sufficient progress for that.

• 2:04:00Yann LeCun?

  You know, AR is a much easier concept to grasp that you're going to have, you know, augmented reality glasses that basically contain some sort of, you know, virtual assistant that can help you in your daily lives.

• 2:04:12Lex Fridman?

  But at the same time with the AR, you have to contend with reality. With VR, you can completely detach yourself from reality, so it gives you freedom. It might be easier to design worlds in VR.

• 2:04:22Yann LeCun?

  Yeah, but you can imagine, you know, the metaverse being...

• 2:04:26Lex Fridman?

  A mix.

• 2:04:27Yann LeCun?

  A mix, right? Or like you can have objects that exist in the metaverse that, you know, pop up on top of the real world or only exist in virtual reality.

• 2:04:36Lex Fridman?

  Okay, let me ask the hard question.

• 2:04:39Yann LeCun?

  Because all of this was easy.

• 2:04:40Lex Fridman?

  This was easy. The Facebook, now Meta, the social network, has been painted by the media as a net negative for society, even destructive and evil at times. You've pushed back against this, defending Facebook. Can you explain your defense?

• 2:04:58Yann LeCun?

  Yeah, so the description, the company that is being described in the, in some media, is not the company we know when we work inside. And, you know, it could be claimed that a lot of employees are uninformed about what really goes on in the company. But, you know, I'm a vice president. I mean, I have a pretty good vision of what goes on. You know, I don't know everything, obviously. I'm not involved in, in, in everything, but certainly not in decision about like, you know, content moderation or anything like this. But, but I have, you know, some decent vision of what, what goes on. And this evil that is being described, I just don't see it. And then, you know, I think there is an easy story to buy, which is that, you know, all the bad things in the, in the world

• 2:05:43Yann LeCun?

  and, you know, the, the reason your friend believe crazy stuff. You know, there's an easy scapegoat, right? In the, in, in, in, in social media, in general, Facebook in particular. But you have to look at the data, like, is it the case that Facebook, for example, polarizes people politically? Are there academic studies that show this? Is it the case that, you know, teenagers think of themselves less if they use Instagram more? Is it the case that, you know, people get more riled up against, you know, opposite sides in a, in a debate or political opinion if they, if they are more on Facebook or if they are less?

• 2:06:27Yann LeCun?

  And study after study show that none of this is true. This is independent studies by academic. They're not funded by Facebook or Meta. You know, studied by Stanford, by some of my colleagues at NYU, actually, with whom I have no connection. You know, there's a, a study recently, they, they, they paid people, I think it was in.

• 2:06:50Yann LeCun?

  In former Yugoslavia, I'm not exactly sure in what part, but they paid people to not use Facebook for a while in the period before the anniversary of the Sibernica massacres, right? So people get riled up, like, you know, should we have a celebration? I mean, a memorial kind of celebration for it or not? So they paid a bunch of people to not use Facebook for a few weeks. It turns out that those people ended up being more polarized than they were at the beginning, and the people who were more on Facebook were less polarized. There's a study, you know, from Stanford of economists at Stanford

• 2:07:32Yann LeCun?

  that tried to identify the causes of increasing polarization in the US. And it's been going on for 40 years before, you know, Mark Zuckerberg was born.

• 2:07:42Lex Fridman?

  Yeah.

• 2:07:43Yann LeCun?

  Continuously. And so if there is a cause, it's not Facebook or social media. So you could say if social media just accelerated. But no, I mean, it's basically a continuous evolution by some measure of polarization in the US. And then you compare this with other countries like the west half of Germany, because you can't go 40 years in the east side, or Denmark, or other countries. And they use Facebook just as much. And they're not getting more polarized, they're getting less polarized. So if you want to look for, you know, a causal relationship there, you can find a scapegoat, but you can't find a cause. Now, if you want to fix the problem, you have to find the right cause.

• 2:08:25Yann LeCun?

  And what rise me up is that people now are accusing Facebook of bad deeds that are done by others, and those others are, were not doing anything about them. And by the way, those others include the owner of the Wall Street Journal, in which all of those papers were published.

• 2:08:39Lex Fridman?

  So I should mention that I'm talking to Shrep, Mike Shrepfer on this podcast, and also Mark Zuckerberg, and probably these are conversations I can have with them. Because it's very interesting. To me, even if Facebook has some measurable negative effect, you can't just consider that in isolation. You have to consider about all the positive ways it connects us. So like, every technology-

• 2:08:59Yann LeCun?

  It connects people, there's no question.

• 2:09:01Lex Fridman?

  You can't just say like, there's an increase in division. Yes, probably Google's search engine has created increasing division. But you have to consider about how much information are brought to the world. Like, I'm sure Wikipedia created more division if you just look at the division. But you have to look at the full context of the world, and did it make a better world?

• 2:09:21Yann LeCun?

  I mean, the printing press has created more division, right?

• 2:09:23Lex Fridman?

  Exactly.

• 2:09:25Yann LeCun?

  So, you know, when the printing press was invented, the first books that were printed were things like the Bible, and that allowed people to read the Bible by themselves, not get the message uniquely from priests in Europe. And that created, you know, the Protestant movement, and 200 years of religious persecution and wars. So that's a bad side effect of the printing press. You know, social networks aren't being nearly as bad as the printing press. But nobody would say the printing press was a bad idea.

• 2:09:55Lex Fridman?

  Yeah, a lot of this perception, and there's a lot of different incentives operating here. Maybe a quick comment, since you're one of the top leaders at Facebook and at Meta, sorry, that's in the tech space. I'm sure Facebook involves a lot of incredible technological challenges that need to be solved. A lot of it probably is in the computer infrastructure, the hardware, the... I mean, it's just a huge amount. Maybe can you give me context about how much of Shrep's life is AI? And how much of it is low-level compute? How much of it is flying all around, doing business stuff? And the same with Zuckerberg, Mark Zuckerberg?

• 2:10:34Yann LeCun?

  They really focus on AI. I mean, certainly, in the-

• 2:10:39Yann LeCun?

  In the run up of the creation affair and for, you know, at least a year after that, if not more, Mark was very, very much focused on AI and was spending quite a lot of effort on it. And that's his style. When he gets interested in something, he reads everything about it. You know, he read some of my papers, for example, before I joined.

• 2:11:01Yann LeCun?

  And so he learned a lot about it.

• 2:11:03Lex Fridman?

  He said he liked notes.

• 2:11:05Yann LeCun?

  Right.

• 2:11:08Yann LeCun?

  And, you know, Shrep was really to it also. I mean, Shrep is really kind of,

• 2:11:16Yann LeCun?

  you know, has something I've tried to preserve also despite my not so young age, which is a sense of wonder about science and technology. And he certainly, he certainly has that. He's also a wonderful person. I mean, in terms of like, as a manager, like dealing with people and everything, Mark also, actually. I mean, they're very like, you know, very human people. In the case of Mark, it's shockingly human, you know, given his, his trajectory.

• 2:11:47Yann LeCun?

  I mean, the personality of him that is painted in the press is just completely wrong.

• 2:11:51Lex Fridman?

  Yeah, but you have to know how to play the press. So that's, I put some of that responsibility on him to, you have to, it's like, you know, like the director, the conductor of an orchestra. You have to play the press and the public in a certain kind of way where you convey your true self to them. If there is a depth and kindness to you.

• 2:12:13Yann LeCun?

  And he's probably not the best at it. So, yeah.

• 2:12:18Lex Fridman?

  You have to learn. And it's sad to see, and I'll talk to him about it, but Shrep is slowly stepping down. It's always sad to see folks sort of be there for a long time and slowly. I guess time is sad.

• 2:12:32Yann LeCun?

  I think he's done the thing he set out to do and, you know, he's got, you know,

• 2:12:41Yann LeCun?

  family priorities and stuff like that. And I understand, you know, after 13 years or something.

• 2:12:49Lex Fridman?

  It's been a good run.

• 2:12:50Yann LeCun?

  Which in Silicon Valley is basically a lifetime.

• 2:12:54Lex Fridman?

  Yeah.

• 2:12:54Yann LeCun?

  You know, because, you know, it's dog years.

• 2:12:56Lex Fridman?

  So in Europe, the conference just wrapped up. Let me just go back to something else. You posted the paper you co-authored was rejected from Europe's. As you said proudly in quotes, rejected.

• 2:13:10Yann LeCun?

  It was a joke.

• 2:13:10Lex Fridman?

  Yeah, I know.

• 2:13:13Lex Fridman?

  Can you describe this paper? And like, what was the idea in it? And also, maybe this is a good opportunity to ask, what are the pros and cons? What works and what doesn't about the review process?

• 2:13:25Yann LeCun?

  Yeah, let me talk about the paper first. I'll talk about the review process afterwards. The paper is called VCRAG. So this is, I mentioned that before, variance in variance, covariance regularization. And it's a technique, a non-contrasting learning technique for what I call joint embedding architectures. So SMI's nets are an example of joint embedding architecture. So joint embedding architecture is.

• 2:13:51Yann LeCun?

  Let me back up a little bit, right? So if you want to do self-supervised learning, you can, you can do it by prediction. So let's say you want to train a system to predict video, right? You show it a video clip and, and you train the system to predict the next, the continuation of that video clip. Now, because you need to handle uncertainty, because there are many, you know, many continuations that are plausible. You need to have, you need to handle this in some way. You need to have a way for the system to be able to produce multiple predictions. And the way, the only way I know to do this is through what's called a latent variable. So you have some sort of hidden vector of a variable that you can vary over a set or draw from a distribution.

• 2:14:34Yann LeCun?

  And as you vary this vector over a set, the output, the prediction varies over a set of plausible predictions. Okay, so that's called, I call this a generative latent variable model.

• 2:14:46Lex Fridman?

  Got it.

• 2:14:46Yann LeCun?

  Okay. Now, there is an alternative to this to handle uncertainty. And instead of directly predicting the, the next frames of the, of the, of the clip, you also run those through another neural net.

• 2:15:03Yann LeCun?

  So you now have two neural nets, one that looks at the, the, you know, the initial segment of the video clip. And another one that looks at the, the continuation during training, right? And what, what you're trying to do is learn a representation. Of those two video clips that is maximally informative about the video clips themselves, but is such that you can predict the representation of the second video clip from the representation of the first one easily, okay? And you can sort of formalize this in terms of maximizing mutual information, some stuff like that, but it doesn't matter.

• 2:15:40Yann LeCun?

  What you want is informative representations of the two video clips that are mutually predictable. What that means is that there's a lot of details in the second video clip that are irrelevant. You know, I, let's say, a video clip consists in, you know, a camera panning the scene. There's going to be a piece of that room that is going to be revealed, and I can somewhat predict what that room is going to look like, but I may not be able to predict the details of the texture of the ground and where the tiles are ending and stuff like that, right? So those are irrelevant details that perhaps my representation will eliminate.

• 2:16:21Yann LeCun?

  And so what I need is to train this second neural net in such a way that whenever the continuation video clip varies over all the plausible continuations, the representation doesn't change.

• 2:16:37Lex Fridman?

  Got it. So it's the, yeah, yeah, got it. Over the space of representations, doing the same kind of thing as you're doing with similarity learning.

• 2:16:46Yann LeCun?

  Right. So these are two ways to handle multimodality in a prediction, right? In the first way, you parameterize the prediction with a latent variable, but you predict pixels, essentially, right? In the second one, you don't predict pixels, you predict an abstract representation of pixels, and you guarantee that this abstract representation has as much information as possible about the input, but sort of, you know, drops all the stuff that you really can't predict, essentially. I used to be a big fan of the first approach, and in fact, in this paper with the Chan-Mitra, this blog post, The Dark Matter Intelligence, I was kind of advocating for this. And in the last year and a half, I've completely changed my mind. I'm now a big fan of the second one.

• 2:17:27Yann LeCun?

  And it's because of a small collection of algorithms that have been proposed over the last year and a half or so, two years, to do this, including V-Creg, its predecessor called Babel Twins, which I mentioned, a method from our friends at DeepMind called BYOL. And and and there's a bunch of others now that can work similarly. So they're all based on this idea of joint embedding. Some of them have an explicit criterion. There is an approximation of mutual information. Some others would be well, well, work, but we don't really know why. And there's been like lots of theoretical papers about why be where that works. No, it's not that because we take it out and it still works and you know, blah, blah, blah.

• 2:18:08Yann LeCun?

  I mean, so there's like a big, a big debate, but, um, uh, but the important point is that we now have a collection of non-contrasting joint embedding methods, which I think is the best thing since sliced bread. So I'm super excited about this because I think it's our best shot for techniques that would allow us to kind of build predictive world models. And at the same time, learn hierarchical representations of the world where what matters about the world is preserved and what is irrelevant is eliminated.

• 2:18:36Lex Fridman?

  By the way, the representations of before and after is in the space in a sequence of images or is it for single images?

• 2:18:44Yann LeCun?

  Uh, it would be either for a single image for a sequence. It doesn't have to be images. This could be applied to text. This could be applied to just about any signal. I'm looking at, you know, I'm looking for methods that are generally applicable that are not specific to, you know, one particular modality. You know, it could be audio or whatever.

• 2:18:59Lex Fridman?

  Got it. So what's the story behind this paper? This, this paper is what is describing one of the, one such method?

• 2:19:05Yann LeCun?

  This, this Vikrak method. So this is co-authored, uh, the first author is a student, uh, called Adrien Bard, who is a resident PhD student at Fair Paris. Was co-advised by me and Jean-Ponce, um, who's a professor at Economie Supérieure, also a research director at INRIA. Uh, so this is wonderful program in France where PhD students can basically do their PhD in industry. And that's kind of what, what's happening here.

• 2:19:32Yann LeCun?

  And this paper is a follow-up on, uh, the, the, this Barlow-Twin paper by my former postdoc, now Stephane Doni, uh, with Li Jing and, and Yurich Montar and a bunch of other people from, from, from Fair. And one of the main criticisms from reviewers is that Vikrak is not different enough from Barlow-Twin's. But, you know, my impression is that it's, you know, Barlow-Twin's with a few bugs fixed essentially. And, uh, in the end, this is what people reuse.

• 2:20:05Lex Fridman?

  Right, so.

• 2:20:06Yann LeCun?

  But you know, I'm used to stuff.

• 2:20:08Lex Fridman?

  Yeah.

• 2:20:09Yann LeCun?

  That I submit being rejected for.

• 2:20:10Lex Fridman?

  So it might be rejected and actually exceptionally well cited because people use it.

• 2:20:14Yann LeCun?

  Well, it's already cited like a bunch of times.

• 2:20:16Lex Fridman?

  So I mean, the, the, the question is then to the deeper question about peer review and conferences. I mean, computer science as a field is kind of unique that the conference is highly prized. That's one.

• 2:20:27Yann LeCun?

  Right.

• 2:20:28Lex Fridman?

  And it's interesting because the peer review process there is similar, I suppose, to journals, but it's accelerated significantly, or not significantly, but it goes fast. And it's a nice way to get stuff out quickly, to peer review it quickly, go to present it quickly to the community. So not quickly, but quicker.

• 2:20:47Yann LeCun?

  Yeah.

• 2:20:47Lex Fridman?

  But nevertheless, it has many of the same flaws of peer review because it's a limited number of people look at it. There's bias and the following like that. If you, if you want to do new ideas, you're going to get pushed back.

• 2:21:00Lex Fridman?

  There's self-interested people that kind of can infer who submitted it and kind of, you know, be cranky about it, all that kind of stuff.

• 2:21:09Yann LeCun?

  Yeah, I mean, there's a lot of, you know, social phenomena there. There's one social phenomenon, which is that because the field has been growing exponentially, the vast majority of people in the field are extremely junior.

• 2:21:21Lex Fridman?

  Yeah.

• 2:21:22Yann LeCun?

  So as a consequence, and that's just a consequence of the field growing, right? So as the number of, as the size of the field kind of starts saturating, you, you will have less of that problem of reviewers being very inexperienced. A consequence of this is that, you know, young reviewers, I mean, there's a phenomenon which is that reviewers try to make their life easy and to make their life easy when reviewing a paper is very simple. You just have to find a flaw in the paper, right? So basically they see their task as finding flaws in papers and most papers have flaws, even the good ones.

• 2:21:58Lex Fridman?

  Yeah.

• 2:22:00Yann LeCun?

  So it's, it's easy to, you know, to do that. Your, your, your job is easier as a reviewer if you just focus on this. But what's important is. Like, is there a new idea in that paper that is likely to influence? It doesn't matter if the experiments are not that great, if the protocol is, you know, so, so, you know, things like that, as long as there is a worthy idea in it that will influence the way people think about the problem. Even if they make it better, you know, eventually, I think that's, that's really what, what makes a paper useful. And so this combination of social phenomena creates a, a, a disease

• 2:22:44Yann LeCun?

  that has plagued, you know, other fields in the past, like speech recognition, where basically, you know, people chase numbers on, on benchmarks. And, and it's much easier to get a paper accepted if it brings an incremental improvement on a sort of mainstream, well-accepted method or problem. And those are, to me, boring papers. I mean, they're not useless, right? Because industry strives on, on those kind of progress. But they're not the one that I'm interested in, in terms of like new concepts and new ideas. So papers that are really trying to strike kind of new advances generally don't make it. Now thankfully, we have Archive.

• 2:23:26Lex Fridman?

  Archive, exactly. And then there's open review type of situations where you, and then, I mean, Twitter is a kind of open review. I, I'm a huge believer that review should be done by thousands of people, not two people.

• 2:23:37Yann LeCun?

  I agree.

• 2:23:38Lex Fridman?

  And so Archive, like, do you see a future where a lot of really strong papers, it's already the present, but a growing future where it'll just be Archive? And you're presenting an ongoing, continuous conference called Twitter slash the internet slash Archive Sanity. Andre just released a new version. So just not, you know, not being so elitist about this particular gating.

• 2:24:05Yann LeCun?

  It's not a question of being elitist or not. It's a question of being basically.

• 2:24:11Yann LeCun?

  Recommendation and CIO approvals for people who don't see themselves as having the ability to do so by themselves, right? And so it saves time, right? If you rely on other people's opinion and you trust those people or those groups to evaluate a paper for you, that saves you time because, you know, you don't have to like scrutinize the paper as much, you know, it is brought to your attention. I mean, there's a whole idea of sort of, you know, collective recommender system, right? So I actually thought about this a lot, you know, about ten, fifteen years ago, because there were discussions at NIPS and, you know, and we're about to create ICLIA with Yashab Enjo. And so I wrote a document

• 2:24:55Yann LeCun?

  kind of describing a reviewing system, which basically was, you know, you post your paper on some repository, let's say Archive, or now it could be Open Review. And then you can form a reviewing entity, which is equivalent to a reviewing board, you know, of a journal or a program committee of a conference. You have to list the members. And then that group, reviewing entity, can choose to review a particular paper, spontaneously or not. There is no exclusive relationship anymore between a paper and a venue or reviewing entity. Any reviewing entity can review any paper or may choose not to. And then, you know,

• 2:25:37Yann LeCun?

  give an evaluation. It's not published, not published, it's just an evaluation and a comment, which would be public, signed by the reviewing entity. And if it's signed by a reviewing entity, you know, it's one of the members of reviewing entity. So if the reviewing entity is, you know, Lex Fridman's, you know, preferred papers, right? You know, it's Lex Fridman writing the review.

• 2:25:57Lex Fridman?

  Yes. What, so for me, one, that's a beautiful system, I think. But what's in addition to that, it feels like there should be a reputation system for the reviewers.

• 2:26:08Yann LeCun?

  Absolutely. For the reviewing entities, not the reviewers individually.

• 2:26:12Lex Fridman?

  The reviewing entities, sure. But even that, within that, the reviewers too, because There's another thing here. It's not just the reputation. It's an incentive for an individual person to do great. Right now, in the academic setting, the incentive is kind of internal, just wanting to do a good job. But honestly, that's not a strong enough incentive to do a really good job in reading a paper and finding the beautiful amidst the mistakes and the flaws and all that kind of stuff. Like, if you're the person that first discovered a powerful paper and you get to be proud of that discovery, then that gives a huge incentive to you.

• 2:26:49Yann LeCun?

  That's a big part of my proposal, actually, where I describe that as, you know, if your evaluation of papers is predictive of future success,

• 2:26:58Lex Fridman?

  Yes, yes.

• 2:26:59Yann LeCun?

  then your reputation should go up as a reviewing entity. So, yeah, exactly. I mean, I even had a master's student, who was a master's student in library science and computer science, actually kind of work out exactly how that should work with formulas and everything, but...

• 2:27:17Lex Fridman?

  So, in terms of implementation, do you think that's something that's doable?

• 2:27:20Yann LeCun?

  I mean, I've been sort of, you know, talking about this to sort of various people, like, you know, Andrew McCallum, who started Open Review. And the reason why we picked Open Review for ICLIA initially, even though it was very early for them, is because my hope was that ICLIA was eventually going to kind of inaugurate this type of system. So, ICLIA kept the idea of open reviews, so where the reviews are, you know, published with a paper, which I think is very useful. But in many ways, that's kind of reverted to kind of more of a conventional type conferences for everything else. And that, I mean, I don't run ICLIA, I'm just the president of the foundation,

• 2:28:03Yann LeCun?

  but, you know, people who run it should make decisions about how to run it, and I'm not going to tell them, because they are volunteers and I'm really thankful that they do that. So, but I'm saddened by the fact that we're not being innovative enough.

• 2:28:19Lex Fridman?

  Yeah, me too. I hope that changes, yeah. Because of communication of science broadly, but communication of computer science ideas. Is how you make those ideas have impact, I think.

• 2:28:30Yann LeCun?

  Yeah, and I think, you know, a lot of this is because people have in their minds kind of an objective, which is, you know, fairness for authors and the ability to count points, basically, and give credits accurately. But that comes at the expense of the progress of science. So to some extent, we're slowing down the progress of science.

• 2:28:54Lex Fridman?

  And are we actually achieving fairness?

• 2:28:56Yann LeCun?

  And we're not achieving fairness, you know, we still have biases, you know, we're doing, you know, double-blind review, but, you know, the biases are still there, there are different kinds of biases.

• 2:29:08Lex Fridman?

  You write that the phenomenon of emergence, collective behavior exhibited by a large collection of simple elements, interaction, is one of the things that got you into neural nets in the first place. I love cellular automata, I love simple interacting elements and the things that emerge from them. Do you think we understand how complex systems can emerge from such simple components that interact simply?

• 2:29:33Yann LeCun?

  No, we don't. It's a big mystery also, it's a mystery for physicists, it's a mystery for biologists.

• 2:29:39Yann LeCun?

  You know, how is it that the universe around us seems to be increasing in complexity and not decreasing? I mean, that is a kind of curious property of physics, that despite the second law of thermodynamics, we seem to be, you know, evolution and learning and et cetera seems to be kind of, at least locally, to increase complexity, not decrease it. So perhaps the ultimate purpose of the universe is to just get more complex.

• 2:30:11Lex Fridman?

  Have these, I mean, small pockets of beautiful complexity. Does that, the cellular automata, do these kinds of emergence of complex systems give you some intuition or guide your understanding of machine learning systems and neural networks and so on? Or are these, for you right now, disparate concepts?

• 2:30:31Yann LeCun?

  Well, it got me into it. You know, I discovered the existence of the perceptron when I was a college student,

• 2:30:40Yann LeCun?

  you know, by a really good book, and it was a debate between Chomsky and Piaget. Seymour Pepperd from MIT was kind of singing the praise of the perceptron in that book. And the first time I heard about the learning machine, right? So I started digging the literature and I found those papers, those books, which were basically transcription of, you know, workshops or conferences from the 50s and 60s about self-organizing systems. So there were, there was a series of conferences on self-organizing systems and these books on this. Some of them are, you can actually get them at the Internet Archive, you know, the digital version. And there are like fascinating articles in there by, there's a guy whose name has been largely forgotten, Heinz von Förster.

• 2:31:24Yann LeCun?

  He's a German physicist who immigrated to the US and worked on self-organizing systems in the 50s. And in the 60s, he created, at University of Illinois, Benet-Champagne, he created the Biological Computer Laboratory, ECL, which was, you know, all about neural nets. Unfortunately, that was kind of towards the end of the popularity of neural nets. So that lab never kind of strived very much. But, but he wrote a bunch of papers about self-organization and about the mystery of self-organization. An example he has is you take, imagine you are in space, there's no gravity. You have a big box with magnets in it, okay? You know, kind of rectangular magnets with North Pole on one end, South Pole on the other end.

• 2:32:08Yann LeCun?

  You shake the box gently and the magnets will kind of stick to themselves and probably form a complex structure, you know, spontaneously. You know, that could be an example of self-organization. But, you know, you have lots of examples. Neural nets are an example of self-organization to, you know, in many respects. And it's a, it's a bit of a mystery, you know, how, like what, what is possible with this? You know, pattern formation in physical systems, in chaotic system and things like that. You know, you know, the emergence of life, you know, things like that. So, you know, how does, how does that happen? It's a, it's a big puzzle for physicists as well.

• 2:32:44Lex Fridman?

  It feels like understanding this, the mathematics of emergence in some constrained situations might help us create intelligence, like help us add a little spice to the systems. Because you seem to be able to In complex systems with emergence to be able to get a lot from little. And so that seems like a shortcut to get big leaps in performance. But.

• 2:33:13Yann LeCun?

  But there's a missing theory, a concept that we are, we don't have.

• 2:33:16Lex Fridman?

  Yeah.

• 2:33:17Yann LeCun?

  And it's, it's something also I've been fascinated by since my undergrad days. And it's how you measure complexity. Right, so we don't actually have good ways of measuring, or at least we don't have good ways of interpreting the measures that we, we have at our disposal. Like, how do you measure the complexity of something? Right? So there is all those things, you know, like, you know, Kolmogorov, Chaiting, Solomonov, complexity of, you know, the length of the shortest program that would generate a bit string can be thought of as the complexity of that bit string. Right? I've been fascinated by that concept. The problem with that is that that complexity is defined up to a constant, which can be very large.

• 2:33:57Lex Fridman?

  Right.

• 2:33:58Yann LeCun?

  There's, there are similar concepts that are derived from, you know, you know, Bayesian probability theory, where, you know, the complexity of something is the negative log of its probability, essentially, right? And you have a complete equivalence between the two things. And there you would think, you know, the probability is something that's well defined mathematically, which means complexity is well defined. But it's not true. You need to have a model of, of, of the distribution. And you may need to have a prior if you're doing Bayesian inference. And the prior plays the same role as the choice of the computer with which you measure Kolmogorov complexity. And so every measure of complexity we have has some arbitrariness.

• 2:34:38Yann LeCun?

  You know, an additive constant, which is, can be arbitrarily large. And so, you know, how can we come up with a good theory of how things become more complex if we don't have a good measure of complexity?

• 2:34:48Lex Fridman?

  Yeah. Which we need for, this is one way that people study this in the space of biology. The people that study the origin of life or try to recreate life in, in the laboratory. And the more interesting one is the alien one, is when we go to other planets, how do we recognize that it's life? Because, you know, complexity, we associate complexity, maybe some level of mobility with life. You know, we have to be able to like have concrete algorithms for like measuring the level of complexity we see in order to know the difference between life and non-life.

• 2:35:25Yann LeCun?

  And the problem is that complexity is in the eye of the beholder. So let me give you an example. If I, if I give you an image of the MNIST digits, right, and I flip through MNIST digits. There is some, obviously some structure to it because local structure, you know, neighboring pixels are correlated across the entire data set. Now imagine that

• 2:35:50Yann LeCun?

  I, I apply a random permutation to all the pixels. A fixed random permutation. Now I show you those images, they will look, you know, really disorganized to you, more complex. In fact, they're not more complex in absolute terms, they're exactly the same as originally, right? And if you knew what the permutation was, you know, you could undo the permutation. Now, imagine I give you special glasses that undo that permutation. Now, all of a sudden, what looked complicated becomes simple.

• 2:36:19Lex Fridman?

  Right.

• 2:36:20Yann LeCun?

  So if you have two, if you have, you know, humans on one end, and then another race of aliens that sees the universe with permutation glasses.

• 2:36:27Lex Fridman?

  Yeah, with the permutation glasses.

• 2:36:30Yann LeCun?

  What we perceive as simple to them is horribly complicated, it's probably heat.

• 2:36:34Lex Fridman?

  Yeah, heat, yeah.

• 2:36:35Yann LeCun?

  Okay, and what they perceive as simple to us is, is random fluctuation, it's heat.

• 2:36:41Lex Fridman?

  Yeah. Truly in the eye of the beholder. Depends what kind of glasses you're wearing. Depends what kind of algorithm you're running in your perception system.

• 2:36:50Yann LeCun?

  So I don't think we'll have a theory of intelligence, self-organization, evolution, things like this, until we have a good handle on a notion of complexity, which we know is in the eye of the beholder.

• 2:37:04Lex Fridman?

  Yeah, it's sad to think that we might not be able to detect or interact with alien species because we're wearing different glasses.

• 2:37:12Yann LeCun?

  Because their notion of locality might be different from ours.

• 2:37:14Lex Fridman?

  Yeah.

• 2:37:14Yann LeCun?

  This actually connects with fascinating questions in physics at the moment, like modern physics, quantum physics, like, you know, questions about like, you know, can we recover the information that's lost in a black hole and things like this, right? And that relies on notions of complexity, which, you know, I find this fascinating.

• 2:37:33Lex Fridman?

  Can you describe your personal quest to build an expressive electronic wind instrument, EWI? What is it? What does it take to build it?

• 2:37:46Yann LeCun?

  Well, I'm a tinkerer. I like building things. I like building things with combinations of electronics and, you know, mechanical stuff.

• 2:37:54Yann LeCun?

  You know, I have a bunch of different hobbies, but, you know, probably my first one was little, was building model airplanes and stuff like that, and I still do that to some extent. But also electronics, I taught myself electronics before I studied it. And the reason I taught myself electronics is because of music. My cousin was an aspiring electronic musician and he had an analog synthesizer and I was, you know, basically modifying it for him and building sequencers and stuff like that, right, for him. I was in high school when I was doing this.

• 2:38:24Lex Fridman?

  Hence the interest in like progressive rock, like 80s. Like, what's the greatest band of all time, according to Yalun Kuhn?

• 2:38:31Yann LeCun?

  Oh, there's too many of them, but, you know, it's a combination of, you know, Mahavishnu Orchestra, Weather Report, yes, Genesis, you know,

• 2:38:44Lex Fridman?

  Yes, Genesis.

• 2:38:45Yann LeCun?

  Pre-Peter Gabriel, Gentle Giant, you know, things like that.

• 2:38:51Lex Fridman?

  Great. Okay, so this love of electronics and this love of music combined together.

• 2:38:56Yann LeCun?

  Right, so I was actually trained to play Baroque and Renaissance music and I played in an orchestra when I was in high school and first year of college. And I played the recorder, trombone, a little bit of oboe, you know, things like that. So I'm a wind instrument player, but I always wanted to play improvised music, even though I don't know anything about it. And the only way I figured, you know, short of like learning to play saxophone, was to play electronic wind instruments. So they behave, you know, the fingering is similar to a saxophone, but, you know, you have wide variety of sound because you control the synthesizer with it. So I had a bunch of those, you know, going back to the late eighties,

• 2:39:37Yann LeCun?

  from either Yamaha or Akai, they're both kind of the main manufacturers of those, so they were classically, you know, going back several decades. But I've never been completely satisfied with them because of lack of expressivity.

• 2:39:53Yann LeCun?

  And, you know, those things, you know, are somewhat expressive. I mean, they measure the breath pressure, they measure the lip pressure and, you know, you have various parameters, you can vary it with fingers, but they, they're not really as expressive as an acoustic instrument, right? You hear John Coltrane play two notes and you hear, you know it's John Coltrane, you know, it's got a unique sound. Or Mize Davis, right? You can hear it's Mize Davis playing the trumpet because the, the sound reflects their, you know, physiology, basically, the shape of the vocal tract

• 2:40:30Yann LeCun?

  kind of shapes the, the sound. So how, how do you do this with an electronical instrument? And I was, many years ago, I met a guy called David Wessel. He, he was a professor at Berkeley and created the the Center for like, you know, Music Technology there. And he was interested in that question.

• 2:40:48Yann LeCun?

  And so I kept kind of thinking about this for many years. And finally, because of COVID, you know, I was at home. I was in my workshop. My workshop serves also as my kind of Zoom room and, and home office and

• 2:40:59Lex Fridman?

  This is in New Jersey?

• 2:41:00Yann LeCun?

  In New Jersey. And I started really being serious about, you know, building my own EWE instrument.

• 2:41:07Lex Fridman?

  What else is going on in that New Jersey workshop? Is there some, is there some crazy stuff you built that just, or, or like left on the workshop floor? Left behind?

• 2:41:17Yann LeCun?

  A lot of crazy stuff is, you know, electronics with, built with microcontrollers of various kinds. And, you know, weird flying contraptions.

• 2:41:28Lex Fridman?

  So you still love flying?

• 2:41:30Yann LeCun?

  It's a family disease. My dad got me into it when I was a kid. And he was building model airplanes when he was a kid. And he was a mechanical engineer. He taught himself electronics also, so he built his early radio control systems in the late 60s, early 70s. And so that's what got me into, I mean, he got me into kind of, you know, engineering and science and technology.

• 2:41:54Lex Fridman?

  Do you also have an interest and appreciation of flight in other forms like with drones, quadruplers? Or do you, is it model airplane?

• 2:42:04Yann LeCun?

  You know, before drones were, you know, kind of a consumer product. You know, I built my own, you know, with also building a microcontroller with gyroscopes and accelerometers for stabilization, writing the firmware for it, you know. And then when it became kind of a standard thing you could buy, it was boring, you know, I stopped doing it. It was not fun anymore.

• 2:42:25Lex Fridman?

  Yeah, you were doing it before it was cool.

• 2:42:28Yann LeCun?

  Yeah.

• 2:42:28Lex Fridman?

  What advice would you give to a young person today in high school and college? That dreams of doing something big like Yann Le Cun, like let's talk in the space of intelligence. Dreams of having a chance to solve some fundamental problem in space of intelligence, both for their career and just in life, being somebody who was a part of creating something special.

• 2:42:52Yann LeCun?

  So try to get interested by big questions. Things like, you know, what is intelligence? What is the universe made of? What's life all about? Things like that.

• 2:43:07Yann LeCun?

  Like even like crazy big questions like what's time? Like nobody knows what time is.

• 2:43:15Yann LeCun?

  And then learn basic things like basic methods, either from math, from physics or from engineering. Things that have a long shelf life. Like if you have a choice between like, you know, learning, you know, mobile programming on iPhone or quantum mechanics, take quantum mechanics.

• 2:43:39Yann LeCun?

  Because you're going to learn things that you have no idea exist. And you may not, you may never be a quantum physicist, but you'll learn about path integrals and path integrals are used everywhere. It's the same formula that you use for, you know, Bayesian integration and stuff like that.

• 2:43:55Lex Fridman?

  So the ideas, the little ideas within quantum mechanics or within some of these kind of more solidified fields will have a longer shelf life, that you'll somehow use indirectly in your work.

• 2:44:08Yann LeCun?

  Learn classical mechanics, like you'll learn about Lagrangian, for example,

• 2:44:13Yann LeCun?

  which is like a huge, hugely useful concept, you know, for all kinds of different things. Learn statistical physics, because all the math that comes out of, you know, for machine learning basically comes out of, was figured out by statistical physicists in the, you know, late 19th, early 20th century, right? So, and for some of them, actually more recently, by people like Giorgio Peruzzi, who just got the Nobel Prize for the replica method, among other things. It's used for a lot of different things, you know, variational inference. That math comes from statistical physics. So, so a lot of those kind of, you know, basic courses, you know, you'll,

• 2:44:57Yann LeCun?

  if you do electrical engineering, you take signal processing, you'll learn about Fourier transforms. Again, something super useful is at the basis of things like graph neural nets, which is an entirely new sub-area of, you know, AI, machine learning, deep learning, which I think is super promising for all kinds of applications. Something very promising, if you're more interested in applications, is the applications of AI, machine learning and deep learning to science. Or to science that can help solve big problems in the world. I have colleagues at Meta, at FAIR, who started this project called Open Catalyst. And it's, it's an open project collaborative.

• 2:45:36Yann LeCun?

  And the idea is to use deep learning to help design new chemical compounds or materials that would facilitate the separation of hydrogen from oxygen. If you can efficiently separate oxygen from hydrogen with electricity. You solve climate change. It's as simple as that because you you cover, you know, some random desert with solar panels. And you have have them work all day, produce hydrogen, and then you shoot the hydrogen wherever it's needed. You don't need anything else.

• 2:46:10Yann LeCun?

  You know, you you have controllable power that's, you know, can be transported anywhere. So if we if we have a large scale, efficient energy storage technology like producing hydrogen, we solve climate change. Here's another way to solve climate change is figuring out how to make fusion work. Now, the problem with fusion is that you make a super hot plasma and the plasma is unstable and you can't control it. Maybe with deep learning, you can find controllers that will stabilize plasma and make, you know, practical fusion reactors. I mean, that's very speculative, but you know, it's worth trying because, you know, the payoff is huge. There is a group at Google working on this led by John Platt.

• 2:46:53Lex Fridman?

  So control, convert as many problems in science and physics and biology and chemistry into a into a learnable problem and see if a machine can learn it.

• 2:47:03Yann LeCun?

  Right, I mean, there's properties of, you know, complex materials that we don't understand from first principle, for example. Right, so, you know, if we could design new, you know, new materials, we could make more efficient batteries, you know, we could make maybe faster electronics. We could, I mean, there's a lot of things we can imagine. Doing or, you know, lighter materials for for cars or airplanes or things like that, maybe better fuel cells. I mean, there's all kinds of stuff we can imagine. If we had good fuel cells, hydrogen fuel cells, we could use them to power airplanes and, you know, transportation wouldn't be, or cars. We wouldn't have a emission problem, CO2 emission problems for for air transportation anymore.

• 2:47:46Yann LeCun?

  So there's a lot of those things, I think, where AI, you know, can be used. And this is not even talking about all the sort of medicine biology and and everything like that, right? You know, like protein folding, you know, figuring out like how could you design a protein that is sticks to another protein that a particular site? Because that's how you design drugs in the end. So you know, deep learning would be useful all of this. And those are kind of, you know, would be sort of enormous progress if we could use it for that. Here's an example. If you take, this is like from recent material physics, you take a monoatomic layer of graphene, right? So it's just carbon on an hexagonal mesh and you make this single, single atom

• 2:48:29Yann LeCun?

  thick. You put another one on top. You twist them by some magic number of degrees, three degrees or something. It becomes superconductor. Nobody has any idea why.

• 2:48:42Lex Fridman?

  I want to know how that was discovered, but that's the kind of thing that machine learning can actually discover, these kinds of things.

• 2:48:47Yann LeCun?

  Well, maybe not, but, but there is a hint perhaps that with machine learning, we could train a system to basically be a phenomenological model of some complex emergent phenomenon, which, you know, superconductivity is one of those, where, you know, this collective phenomenon is too difficult to describe from first principles with the current, you know, the usual sort of, you know, reductionist type method. But we could have deep learning systems that predict the properties of a system from a description of it after being trained with sufficiently many samples. This guy Pascal Fois at EPFL, he has a startup

• 2:49:29Yann LeCun?

  company that where he basically trained a convolutional net essentially to predict the aerodynamic properties of solids. And you can generate as much data as you want by just running computational free dynamics, right? So you give a like a wing, airfoil or something, shape of some kind, and you run computational free dynamics, you get as a result, the drag and, you know, lift and all that stuff, right? And you can, you can generate lots of data, train a neural net to make those predictions. And now what you have is a differentiable model of, let's say, drag and lift as a function of the shape of that solid. And so you can do backgrade and descent, you can optimize the shape. So you get the properties you want.

• 2:50:16Lex Fridman?

  Yeah, that's incredible. That's incredible. And on top of all that, probably you should read a little bit of literature and a little bit of history for inspiration and for wisdom. Because after all, all of these technologies will have to work in the human world.

• 2:50:32Yann LeCun?

  Yes.

• 2:50:32Lex Fridman?

  And the human world is complicated.

• 2:50:34Yann LeCun?

  It is, sadly.

• 2:50:37Lex Fridman?

  Yeah, and this is an amazing conversation. I'm really honored that you talked with me today. Thank you for all the amazing work you're doing at FAIR, at Meta, and thank you for being so passionate after all these years about everything that's going on. You're a beacon of hope for the machine learning community. And thank you so much for spending your valuable time with me today. That was awesome.

• 2:50:57Yann LeCun?

  Thanks for having me on. That was a pleasure.

• 2:51:00Lex Fridman?

  Thanks for listening to this conversation with Jan Leakun. To support this podcast, please check out our sponsors in the description. And now, let me leave you with some words from Isaac Asimov. "Your assumptions are your windows on the world. Scrub them off every once in a while, or the light won't come in." Thank you for listening and hope to see you next time.