Ilya Sutskever: Deep Learning | Lex Fridman Podcast #94 | Transcription

Note: This transcription is split and grouped by topics and subtopics. You can navigate through the Table of Contents on the left. It's interactive. All paragraphs are timed to the original video. Click on the time (e.g., 01:53) to jump to the specific portion of the video.

Introductory Note

Introduction (00:00)

The following is a conversation with Ilya Siskever, co-founder and chief scientist of OpenAI, one of the most cited computer scientists in history with over 165,000 citations. And to me, one of the most brilliant and insightful minds ever in the field of deep learning. There are very few people in this world who I would rather talk to and brainstorm with about deep learning, intelligence, and life in general than Ilya on and off the mic. This was an honor and a pleasure. This conversation was recorded before the outbreak of the pandemic. For everyone feeling the medical, psychological, and financial burden of this crisis, I'm sending love your way. Stay strong. We're in this together. We'll beat this thing. This is the artificial intelligence podcast. If you enjoy it, subscribe on YouTube, review it with five stars and have a podcast. Support on Patreon or simply connect with me on Twitter at Lex Friedman, spelled F-R-I-D-M-A-N. As usual, I'll do a few minutes of ads now and never any ads in the middle that can break the flow of the conversation. I hope that works for you and doesn't hurt the listening experience. This show is presented by CashApp, the number one finance app in the App Store. When you get it, use code Lexpodcast. CashApp, let's you send money to friends by Bitcoin, invest in the stock market with as little as $1. Since CashApp allows you to buy Bitcoin, let me mention that cryptocurrency in the context of the history of money is fascinating. I recommend a cent of money as a great book on this history. Both the book and audiobook are great. Debits and credits on ledgers started around 30,000 years ago. The US dollar created over 200 years ago and Bitcoin, the first decentralized cryptocurrency released just over 10 years ago. So given that history, cryptocurrency is still very much in its early days of development, but it's still aiming to just might redefine the nature of money. So again, if you get CashApp from the App Store or Google Play and use the code Lex podcast, you get $10 and CashApp will also donate $10 to first, an organization that is helping advance robotics and STEM education for young people around the world. And now here's my conversation with the Iliya Satsgather. You were one of the three authors with Alex Kuchowski, Jeff Hinton of the famed Alex Netpaper that is arguably the paper that marked the big catalytic moment that launched the deep learning revolution.

Key Concepts And Developments In Deep Learning

AlexNet paper and the ImageNet moment (02:23)

At that time, take us back to that time. What was your intuition about neural networks, about the representation of power of neural networks? And maybe you could mention how did that evolve over the next few years up to today, over the 10 years? Yeah, I can answer that question. At some point in about 2010 or 2011, I connected two facts in my mind. Basically, the realization was this. At some point, we realized that we can train very large, I shouldn't say very tiny by today's standards, but large and deep neural networks and to end with back propagation. At some point, different people obtained this result. I obtained this result. The first moment in which I realized that deep neural networks are powerful was when James Martin's invented the Hessian Free Optimizer in 2010 and he trained a 10 layer neural network and to end without pre-training from scratch. And when that happened, I thought, this is it. Because if you can train a big neural network, a big neural network can represent very complicated function. Because if you have a neural network with 10 layers, it's as though you allow the human brain to run for some number of milliseconds. Neuron firing is slow. And so in maybe 100 milliseconds, your neurons only fire 10 times. So it's also kind of like 10 layers. And in 100 milliseconds, you can perfectly recognize any object. So I thought, so I already had the idea then that we need to train a very big neural network on lots of supervised data. And then it must succeed because we can find the best neural network. And then there's also theory that if you have more data than parameters, you want to overfit. So Dave, you know that actually this theory is very incomplete and you want to overfit even you have less data than parameters. But definitely if you have more data than parameters, you want to overfit. So the fact that networks were heavily over-parameterized wasn't discouraging to you. So you were thinking about the theory that the number of parameters, the fact there's a huge number of parameters is okay. It's going to be okay. I mean, there was some evidence before that it was okay-ish, but the theory was most, the theory was that if you had a big data set and a big neural net it was going to work, the over-parameterization just didn't really figure much as a problem. I thought, well, with images, you just go and add some data augmentation and it's going to be okay. So where was any doubt coming from? The main doubt was can we train a bigger, if you have enough computer, train a big enough neural net with very propagation. The back propagation, I thought, would work. The thing which wasn't clear was whether there would be enough compute to get a very convincing result. Then at some point, Alex Kojeski wrote these insanely fast CUDA kernels for training convolutional neural nets. That was bam, let's do this, let's get image in it and it's going to be the greatest thing. Was your intuition, most of your intuition from empirical results by you and by others, so like just actually demonstrating that a piece of program can train a 10 layer neural network or was there some pen and paper or marker and whiteboard thinking intuition? Because you just connected a 10 layer large neural network to the brain. So you just mentioned the brain. So in your intuition about neural networks, does the human brain come into play as an intuition builder? Definitely. I mean, you know, you got to be precise with these analogies between neural artificial neural networks in the brain. But there is no question that the brain is a huge source of intuition and inspiration for deep learning researchers since all the way from Rosenblatt in the 60s. Like if you look at the whole idea of a neural network is directly inspired by the brain, you had people like McCollum and Pitts who were saying, Hey, you got these neurons in the brain. And hey, we recently learned about the computer and automata. Can we use some ideas from the computer and automata to design some kind of computational object that's going to be simple, computational and kind of like the brain and they invented the neuron. So they were inspired by it back then. Then you had the convolutional neural network from Fukushima and then later on the con was that, Hey, if you limit the receptive fields of a neural network, it's going to be especially suitable for images as it turned out to be true. So there was a very small number of examples where analogies to the brain were successful. And yeah, I thought, well, probably an artificial neuron is not that different from the brain if it's quite hard enough. So let's just assume it is and roll with it. So we're now at a time where deep learning is very successful. So let us squint less and say, let's open our eyes and say, what to you is an interesting difference between the human brain. Now I know you're probably not an expert, neither in your scientist and your biologist, but loosely speaking, what's the difference between the human brain and artificial neural networks? That's interesting to you for the next decade or two. That's a good question to ask. What is what is an interesting difference between the neurons between the brain and our artificial neural networks? So I feel like today, artificial neural networks, so we all agree that there are certain dimensions in which the human brain vastly outperforms our models. But I also think that there are some ways in which artificial neural networks have a number of very important advantages over the brain. Look, looking at advantages versus disadvantages is a good way to figure out what is the important difference. So the brain uses spikes, which may or may not be important. Yes, that's a really interesting question. Do you think it's important or not? That's one big architectural difference between artificial neural networks. It's hard to tell, but my prior is not very high. And I can tell I can say why. You know, there are people who interested in spike in neural networks. And basically what they figured out is that they need to simulate the non-spiking neural networks in spikes. And that's how they're going to make them work. If you don't simulate the non-spiking neural networks in spikes, it's not going to work because the question is why should it work? And that connects to questions around back propagation and questions around deep learning. You got this giant neural network.

Cost functions (08:33)

Why should it work at all? Why should the learning rule work at all? It's not a self-evident question, especially if you, let's say if you were just starting in the field and you read the very early papers, you can say, hey, people are saying, let's build neural networks. That's a great idea because the brain is a neural network. So it would be useful to build neural networks. Yeah. Now let's figure out how to train them. It should be possible to train them probably, but how? And so the big idea is the cost function. That's the big idea. The cost function is a way of measuring the performance of the system, according to some measure. By the way, that is a big, actually, let me think. Is that, is that a one, a difficult idea to arrive at? And how big of an idea is that? That there's a single cost function. Sorry, let me take a pause. Is supervised learning a difficult concept to come to? I don't know. All concepts are very easy in retrospect. Yeah, that's what it seems trivial now. But I, because the reason I asked that and we'll talk about it, because is there other things? Is there things that don't necessarily have a cost function, or maybe have many cost functions, or maybe have dynamic cost functions, or maybe a totally different kind of architectures? Because we have to think like that in order to arrive at something new, right? So the only, so the good examples of things which don't have clear cost functions are GANs. Right, again, you have a game. So instead of thinking of a cost function, where you want to optimize, where you know that you have an algorithm gradient descent, which will optimize the cost function. And then you can reason about the behavior of your system in terms of what it optimizes. With a GAN, you say, I have a game, and I'll reason about the behavior of the system in terms of the equilibrium of the game. But it's all about coming up with these mathematical objects that help us reason about the behavior of our system. Right, that's really interesting. Yeah, so GAN is the only way. It's kind of the cost function is emergent from the comparison. I don't know if it has a cost function. I don't know if it's meaningful to talk about the cost function of GAN. It's kind of like the cost function of biological evolution, or the cost function of the economy. It's you can talk about regions to which it can go towards, but I don't think. I don't think the cost function analogy is the most useful. So evolution doesn't. That's really interesting. So if evolution doesn't really have a cost function, like a cost function based on its. Something akin to our mathematical conception of a cost function, then do you think cost functions in deep learning are holding us back? Yeah, I so you just kind of mentioned that cost function is a nice first performed idea. Do you think that's a good idea? Do you think it's an idea will go past? So self play starts to touch on that a little bit in reinforcement learning systems. That's right. Self play and also ideas around exploration where you're trying to take action. That's that surprise a predictor. I'm a big fan of course functions. I think cost functions are great and they service really well. And I think that whenever we can do things with cost functions, we should. And you know, maybe there is a chance that we will come up with some yet another profound way of looking at things that will involve cost functions in a less central way. But I don't know. I think cost functions are, I mean, I would not better guess against cost functions. Is there other things about the brain that pop into your mind that might be different and interesting for us to consider in designing artificial neural networks? So we talked about spiking a little bit. I mean, one one thing which may potentially be useful. I think people, neuroscientists have figured out something about the learning rule of the brain or talking about spike time independent plasticity. And it would be nice if some people were to study that in simulation. Wait, sorry, spike time independent plasticity. Yeah, that's a STD. It's a particular learning rule that uses spike time in to figure out how to determine how to update the synapses. So it's kind of like if a synapse fires into the neuron before the neuron fires, then it shrinks in the synapse and if the synapse fires into the neuron shortly after the neuron fired, then it weakens the synapse, something along this line. I'm 90% sure it's right. So if I said something wrong here, don't get too angry, but you sounded really well saying it. But the timing, that's one thing that's missing. The temporal dynamics is not captured. I think that's like a fundamental property of the brain is the timing of the signals. Well, you're requiring neural networks. But you think of that as this, I mean, that's a very crude simplified, what's that called? There's a clock, I guess, to recurrent neural networks. This seems like the brain is the general, the continuous version of that, the generalization where all possible timings are possible. And then within those timings is contained some information.

Recurrent neural networks (13:39)

You think recurrent neural networks, the recurrence in recurrent neural networks can capture the same kind of phenomena as the timing that seems to be important for the brain in the firing of neurons in the brain. I mean, I think recurrent neural networks are amazing and they can do anything we'd want them to, if you'd want a system to do right now, recurrent neural networks have been superseded by transformers. But maybe one day they'll make a comeback. Maybe they'll be back. We'll see. Let me, in a small tangent, say, do you think they'll be back? So so much of the breakthroughs recently that we'll talk about on natural language processing and language modeling has been with transformers that don't emphasize recurrence. Do you think recurrence will make a comeback? Well, some kind of recurrence, I think very likely. Recurring neural networks for as they're typically thought of for processing sequences, I think it's also possible. What is to you a recurrent neural network? In generally speaking, I guess, what is a recurrent neural network? You have a neural network, which maintains a high dimensional hidden state. And then when an observation arrives, it updates its high dimensional hidden state through its connections in some way. So do you think, you know, that's what like expert systems did, right? Symbolic AI, the knowledge based growing a knowledge base is maintaining a hidden state, which is its knowledge base and is growing it by some question. Prossent. Do you think of it more generally in that way or is it simply, is it the more constrained form that of a hidden state with certain kind of gating units that we think of as today with LSCMs and that? I mean, the hidden state is technically what you describe there. The hidden state that goes inside the LSTM or the RNN or something like this. But then what should be contained, you know, if you want to make the expert system analogy, I'm not mean, you could say that the knowledge is stored in the connections. And then the short term processing is done in the in the hidden state. Yes. Could you say that? Yeah. So sort of, do you think there's a future of building large scale knowledge bases within the neural network? Definitely. So we're going to pause on that confidence because I want to explore that. But let me zoom back out and ask back to the history of ImageNet. Neural networks have been around for many decades, as you mentioned.

Key ideas that led to success of deep learning (16:19)

What do you think were the key ideas that led to their success, that image net moment and beyond the success in the past 10 years? OK. So the question is to make sure I didn't miss anything. The key idea is that led to the success of deep learning over the past 10 years. Exactly. Even though the fundamental thing behind deep learning has been around for much longer. So the key idea about deep learning, or rather the key fact about deep learning before deep learning started to be successful is that it was underestimated. People who worked in machine learning simply didn't think that neural networks could do much. People didn't believe that large neural networks could be trained. People thought that, well, there was lots of there was a lot of debate going on in machine learning about what are the right methods and so on. People were arguing because there were no there were no there was no way to get hard facts. And by that, I mean, there were no benchmarks, which were truly hard that if you do really well on them, then you can say, look. Here is my system. That's when you switch from. That's when this field becomes a little bit more of an engineering field. So in terms of deep learning to answer the question directly, the ideas were all there. The thing that was missing was a lot of supervised data and a lot of compute. Once you have a lot of supervised data and a lot of compute, then there is a third thing which is needed as well. And that is conviction. Conviction that if you take the right stuff, it already exists and apply and mix with a lot of data and a lot of compute that it will in fact work. And so that was the missing piece. It was you had the unit, the data, you needed the compute, which showed up in terms of GPUs and you needed the conviction to realize that you need to mix them together. So that's really interesting. So I guess the presence of compute and the presence of supervised data allowed the empirical evidence to do the convincing of the majority of the computer science community. So I guess there's a key moment with a Jitendra, Malik and Alex, Aliyasha Ephros, who were very skeptical. And then there's a Jeffrey Hinton that was the opposite of skeptical. And there was a convincing moment and I think ImageNet served as that moment. That's right. And they represented this kind of or the big pillars of computer vision community, kind of the wizards got together. And then all of a sudden there was a shift and it's not enough for the ideas to all be there and the compute to be there. It's for to convince the cynicism that existed. That's interesting. The people just didn't believe for a couple of decades. Yeah, well, but it's more than that. It's kind of when when put this way, it sounds like, well, you know, those silly people who didn't believe what were they what was they missing. But in reality, things were confusing because neural networks really did not work on anything. And they were not the best method on pretty much anything as well. And it was pretty rational to say, yeah, this stuff doesn't have any traction. And that's why you need to have these very hard tasks which are which produce undeniable evidence. And that's how we make progress. And that's why the field is making progress today because we have these hard benchmarks, which represent true progress. And so and this is why we are able to avoid endless debate. So incredibly you've contributed some of the biggest recent ideas in AI in computer vision, language, natural language processing, reinforcement learning, sort of everything in between.

What's harder to solve: language or vision? (19:57)

Maybe not GANS. Is there there may not be a topic you haven't touched? And of course, the the fundamental science of deep learning. What is the difference to you between vision, language, and as in reinforcement learning action as learning problems? And what are the commonalities? Do you see them as all interconnected? Are they fundamentally different domains that require different approaches? OK, that's a good question. Machine learning is a field with a lot of unity, a huge amount of unity. In fact, what do you mean by unity? Like overlap of ideas, overlap of ideas, overlap of principles. In fact, there is only one or two or three principles, which are very, very simple. And then they apply in almost the same way, in almost the same way to the different modalities of the different problems. And that's why today when someone writes a paper on improving optimization of deep learning and vision, it improves the different NLP applications and it improves the different reinforcement learning applications. Reinforcement learning. So I would say that computer vision and NLP are very similar to each other. Today, they differ in that they have slightly different architectures. We use transformers in NLP and we use convolutional neural networks in vision. But it's also possible that one day this will change and everything will be unified with a single architecture. Because if you go back a few years ago in natural language processing, there were a huge number of architectures for every different, tiny problem had its own architecture. Today, there is just one transformer for all those different tasks. And if you go back in time even more, you had even more and more fragmentation and every little problem in AI had its own little sub-specialization and sub, you know, little set of collection of skills, people who would know how to engineer the features. Now it's all been subsumed by deep learning. We have this unification. And so I expect vision to become unified with natural language as well. Or rather, I shouldn't say expect. I think it's possible. I don't want to be too sure because I think the convolutional neural networks is very computationally efficient. A rel is different. A rel does require slightly different techniques because you really do need to take action. You really need to do something about exploration. Your variance is much higher. But I think there is a lot of unity even there. And I would expect, for example, that at some point there will be some broader unification between a rel and supervised learning where somehow the rel will be making decisions to make the supervised learning go better. And it will be, I imagine, one big black box and you just throw every, you know, you shovel things into it and it just figures out what to do with whatever you shovel at it. I mean, reinforcement learning has some aspects of language and vision combined almost. There's elements of a long-term memory that you should be utilizing and there's elements of a really rich sensory space. So it seems like the, it's like the union of the two or something like that. I'd say something slightly differently. I'd say the reinforcement learning is neither, but it naturally interfaces and integrates with the two of them. You think action is fundamentally different? So yeah, what is interesting about, what is unique about policy of learning to act? Well, so one example, for instance, is that when you learn to act, you are fundamentally in an on-stationary world because as your actions change, the things you see start changing. You experience the world in a different way and this is not the case for the more traditional static problem where you have some distribution and you just apply a model to that distribution. You think it's a fundamentally different problem or is it just a more difficult generalization of the problem of understanding? I mean, it's a question of definitions almost. There is a huge amount of commonality for sure. You take gradients, you take gradients, you try to approximate gradients in both cases. In the case of reinforcement learning, you have some tools to reduce the variance of the gradients. You do that. There's lots of commonality. You use the same neural net in both cases. You compute the gradient, you apply atom in both cases. So, I mean, there's lots in common for sure, but there are some small differences which are not completely insignificant. It's really just a matter of your point of view, what frame of reference, how much do you want to zoom in or out as you look at these problems. Which problem do you think is harder? So, people like Nochamsky believe that language is fundamental to everything. So, it underlies everything. Do you think language understanding is harder than visual scene understanding or vice versa? I think that asking if a problem is hard is likely wrong. I think the question is a little bit wrong and I want to explain why. So, what does it mean for a problem to be hard? Okay, the non-interesting dumb answer to that is there's a benchmark and there's a human level performance on that benchmark and how is the effort required to reach the human level benchmark. So, from the perspective of how much until we get to human level on a very good benchmark. Yeah, I understand what you mean by that. So, what I was going to say that a lot of it depends on once you solve a problem it stops being hard and that's always true. But something is harder not depends on what a tool can do today. So, today through human level, language understanding and visual perception are hard in the sense that there is no way of solving the problem completely in the next three months. So, I agree with that statement. Beyond that, my guess would be as good as yours, I don't know. You don't have a fundamental intuition about how hard language understanding is. I know I changed my mind. I'd say language is probably going to be hard. I mean, it depends on how you define it. Like, if you mean absolute top-notch 100% language understanding, I'll go with language. But then if I show you a piece of paper with letters on it, is that you see what I mean? You have a vision system, you say it's the best human level vision system. I open a book and I show you letters. Really to understand how these letters form into word and sentences and meaning is this part of the vision problem. Where does vision end and language begin? Yeah, so Chomsky would say it starts at language. So, vision is just a little example of the kind of structure and fundamental hierarchy of ideas that's already represented in our brains somehow that's represented through language. But where does vision stop and language begin? That's a really interesting question. So, one possibility is that it's impossible to achieve really deep understanding in either images or language without basically using the same kind of system. So, you're going to get the other for free. I think it's pretty likely that yes, if we can get one, our machine learning is probably that good that we can get the other. But it's not one hundred percent sure. And also, I think a lot of it really does depend on your definitions. Definitions of perfect vision. Because reading is vision, but should it count? Yeah, to me, so my definition is if a system looked at an image and then a system looked at a piece of text and then told me something about that and I was really impressed. That's relative. You'll be impressed for half an hour and then you're going to say, "Well, I mean, all the systems do that, but here's the thing they don't do." Yeah, but I don't have that with humans. Humans continue to impress me. Is that true? Well, the ones, okay, so I'm a fan of monogamy. So, I like the idea of marrying somebody being with them for several decades. So, I believe in the fact that yes, it's possible to have somebody continuously giving you pleasurable, interesting, witty, new ideas, friends. Yeah, I think so. They continue to surprise you. The surprise, it's that injection of randomness seems to be a nice source of continued inspiration, like the wit, the humor. I think that that would be a very subjective test, but I think if you have enough humans in the room. Yeah, I understand what you mean. Yeah, I feel like I misunderstood what you meant by impressing you. I thought you meant to impress you with its intelligence, with how good value it understands an image. I thought you meant something like, "I'm going to show it a really complicated image and it's going to get it right and you're going to say, 'Wow, that's really cool, our systems of, you know, a January 20 to any have not been doing that.'" Yeah, no, I think it all boils down to the reason people click like on stuff on the internet, which is like it makes them laugh. So, it's like humor or wit or insight. I'm sure we'll get that as well. So, forgive the romanticized question, but looking back to you, what is the most beautiful or surprising idea in deep learning or AI in general you've come across?

We're massively underestimating deep learning (29:35)

So, I think the most beautiful thing about deep learning is that it actually works. And I mean it because you got these ideas, you got the little neural network, you got the back propagation algorithm, and then you got some theories as to, you know, this is kind of like the brain, so maybe if you make it large, if you make the neural network large and you train it a lot of data, then it will do the same function of the brain does. And it turns out to be true, that's crazy. And now we just train these neural networks and you make them larger and they keep getting better. And I find it unbelievable. I find it unbelievable that this whole AI stuff with neural networks works. Have you built up an intuition of why are there a little bits and pieces of intuitions of insights, of why this whole thing works? I mean, some definitely, while we know that optimization, we now have good, you know, we've had lots of empirical, you know, huge amounts of empirical reasons to believe that optimization should work on all most problems we care about. Do you have insights of what? So you just said empirical evidence is most of your sort of empirical evidence kind of convinces you it's like evolution is empirical. It shows you that, look, this evolutionary process seems to be a good way to design organisms that survive in their environment, but it doesn't really get you to the insides of how the whole thing works. Well, I think it's a good analogy is physics. You know how you say, hey, let's do some physics calculation and come up with some new physics theory and make some prediction. But then you got around the experiment. You know, you got around the experiment. It's important. So it's a bit the same here, except that maybe some, sometimes the experiment came before the theory. But it still is the case, you know, you have some data and you come up with some predictions. You say, yeah, let's make a big neural network. Let's train it and it's going to work much better than anything before it. And it will, in fact, continue to get better as you make it larger. And it turns out to be true. That's, that's amazing when a theory is validated with this, you know, it's not a mathematical theory. It's more of a biological theory almost. So I think there are not terrible analogies between deep learning and biology. I would say it's like the geometric mean of biology and physics. That's deep learning. The geometric mean of biology and physics. I think I'm going to need a few hours throughout my head around that. Because just to find the geometric, just to find the set of what biology represents. Well, biology in biology, things are really complicated. And there's a really, really, it's really hard to have good predictive theory. And in physics, the theories are too good. In physics, people make these super precise theories, which make these amazing predictions. And in machine learning, we're kind of in between. Kind of in between. But it'd be nice if machine learning somehow helped us discover the unification of the two, as opposed to serving the in between. But you're right. That's, you're, you're kind of trying to juggle both. So do you think there are still beautiful and mysterious properties in your networks that are yet to be discovered? Definitely. I think that we are still massively underestimating deep learning. What did you go look like? Like what? Fine, you, I would have done it. Yeah. So, but if you look at all the progress from the past 10 years, I would say most of it, I would say there've been a few cases where some were things that felt like really new ideas showed up. But by and large, it was every year we thought, okay, deep learning goes this far. Nope. It actually goes further. And then the next year, okay. Now this is, this is big deep learning. We are really done. Nope. Goes further. It just keeps going further each year. So that means that we keep underestimating. We keep not understanding it. A surprising properties all the time. Do you think it's getting harder and harder to make progress? Need to make progress? It depends on what we mean. I think the field will continue to make very robust progress for quite a while. I think for individual researchers, especially people who are doing research, it can be harder because there is a very large number of researchers right now. I think that if you have a lot of compute, then you can make a lot of very interesting discoveries. But then you have to deal with the challenge of managing a huge compute, a huge compute cluster to run your experiments. It's a little bit harder. So I'm asking all these questions that nobody knows the answer to, but you're one of the smartest people I know. So I keep asking, so let's imagine all the breakthroughs that happen in the next 30 years in deep learning. Do you think most of those breakthroughs can be done by one person with one computer? Sort of in a space of breakthroughs, do you think compute will be compute and large efforts will be necessary? I mean, it can't be sure. When you say one computer, you mean how large? Ah, you're clever. I mean, one GPU. I see. I think it's pretty unlikely. I think it's pretty unlikely. I think that there are many. The stack of deep learning is starting to be quite deep. If you look at it, you've got all the way from the ideas, the systems to build the data sets, the distributed programming, the building, the actual cluster, the GPU programming, putting it all together. So now the stack is getting really deep and I think it becomes, it can be quite hard for a single person to become, to be world class in every single layer of the stack. What about what like Vladimir Vapnik really insists on is taking MNIST and trying to learn from very few examples, so being able to learn more efficiently. Do you think that there'll be breakthroughs in that space that would may not need the huge compute? I think that will be a very, I think there will be a large number of breakthroughs in general that you're looking at the huge amount of compute. So I maybe should clarify that. I think that some breakthroughs will require a lot of compute and I think building systems which actually do things will require a huge amount of compute. That one is pretty obvious. If you want to do x and x requires a huge neural net, you've got to get a huge neural net. But I think there will be lots of, I think there is lots of room for very important work being done by small groups and individuals. Can you maybe sort of on the topic of the science that's deep learning, talk about one of the recent papers that you've released, the deep double descent, where bigger models and more data hurt?

Deep double descent (36:04)

I think it's a really interesting paper. Can you describe the main idea? Yeah, definitely. So what happened is that some, over the years, some small number of researchers noticed that it is kind of weird that when you make the neural network logic, it works better and it seems to go in contradiction with statistical ideas. And then some people made an analysis showing that actually you got this double descent bump. And what we've done was to show that double descent occurs for pretty much all practical deep learning systems and that it will be also, so can you step back? What's the x axis and the y axis of a double descent plot? Okay, great. So you can look, you can do things like, you can take a neural network and you can start increasing its size slowly while keeping your data set fixed. So if you increase the size of the neural network slowly and if you don't do early stopping, that's a pretty important detail. Then when the neural network is really small, you make it larger, you get a very rapid increase in performance. Then you continue to make it larger and at some point performance you'll get worse and it gets the worst exactly at the point at which it achieves zero training or precisely zero training loss. And then as you make it large, it starts to get better again. And it's kind of counterintuitive because you'd expect deep learning phenomena to be monotonic. And it's hard to be sure what it means, but it also occurs in the case of linear classifiers and the intuition basically boils down to the following. When you have a large data set and a small model, then small tiny random, so basically what is overfitting? Overfitting is when your model is somehow very sensitive to the small random unimportant stuff in a data set, in the training data set precisely. So if you have a small model and you have a big data set and there may be some random, you know, some training cases are randomly in the data set and others may not be there. But the small model is kind of insensitive to this randomness because it's the same, there is pretty much no uncertainty about the model when the data set is large. So okay, so at the very basic level to me, it is the most surprising thing that neural networks don't overfit every time very quickly before ever being able to learn anything, the huge number of parameters. So here is, so there is one way, okay, so maybe, so let me try to give the explanation and maybe that will be, that will work. So you got a huge neural network, let's suppose you've got a, you have a huge neural network, you have a huge number of parameters. Now let's pretend everything is linear, which is not, let's just pretend. Then there is this big subspace, a neural network achieved zero error. And as a GT is going to find approximately the point, approximately the point with the smallest norm in that subspace. And that can also be proven to be insensitive to the small randomness in the data when the dimensionality is high. But when the dimensionality of the data is equal to the dimensionality of the model, then there is a one-to-one correspondence between all the data sets and the models. So small changes in the data set actually lead to large changes in the model. And that's why performance gets worse. So this is the best explanation, more or less. So then it would be good for the model to have more parameters, so to be bigger than the data. That's right. But only if you don't early stop. If you introduce early stop in your regularization, you can make the double-ascent bump almost completely disappear. What is early stop? Early stop is when you train your model and you monitor your test your validation performance. And then if at some point validation performance starts to get worse, you say, okay, let's stop training. We are good. We are good enough. So the magic happens after after that moment. So you don't want to do the early stop. Well, if you don't do the early stop and you get this very, you get a very pronounced double descent. Do you have any intuition why this happens? Double descent? Sorry, are you stopping? No, the double descent. So the, yeah, so I try, let's see, the intuition is basically is this, that when the data set has as many degrees of freedom as the model, then there is a one to one correspondence between them. And so small changes to the data set lead to noticeable changes in the model. So your model is very sensitive to all the randomness. It is unable to discard it. Whereas it turns out that when you have a lot more data than parameters or a lot more parameters than data, the resulting solution will be insensitive to small changes in the data set. So it's able to definitely put discard the small changes, the randomness, exactly. The spurious correlation if you don't want. Jeff Hinton suggested we need to throw back propagation. We already kind of talked about this a little bit, but he suggested we need to throw away back propagation and start over.

Backpropagation (41:20)

I mean, of course, some of that is a little bit, um, wit and humor. But what do you think? What could be an alternative method of training neural networks? Well, the thing that he said precisely is that to the extent that you can't find back propagation in the brain, it's worth seeing if we can learn something from how the brain learns, but back propagation is very useful and you should keep using it. Oh, you're saying that once we discover the mechanism of learning in the brain or any aspects of that mechanism, we should also try to implement that in your networks. If it turns out that we can't find back propagation in the brain, if we can find back propagation in the brain. Well, so I guess your answer to that is back propagation is pretty damn useful. So why are we complaining? I mean, I personally am a big fan of back propagation. I think it's a great algorithm because it solves an extremely fundamental problem, which is finding a neural circuit subject to some constraints. And I don't see that problem going away. So that's why I, I really, I think it's pretty unlikely that you'll have anything which is going to be dramatically different. It could happen, but I wouldn't bet on it right now. So let me ask a sort of big picture question. Do you think can, do you think neural networks can be made to reason?

Can neural networks be made to reason? (42:42)

Why not? Well, if you look, for example, let alpha go or alpha zero, the neural network of alpha zero plays go, which, which we all agree is a game that requires reasoning, better than 99.9% of all humans, just the neural network without this search, just the neural network itself. Doesn't that give us an existence proof that neural networks can reason? To push back and disagree a little bit, we all agree that Go is reasoning. I think I agree. I don't think it's a trivial. So obviously, reasoning like intelligence is a, is a loose gray area term a little bit. Maybe you disagree with that. But yes, I think it has some of the same elements of reasoning. Reasoning is almost like akin to search, right? There's a sequential element of stepwise consideration of possibilities and sort of building on top of those possibilities in a sequential manner until you arrive at some insight. So yeah, I guess playing goes kind of like that. And when you have a single neural network doing that without search, that's kind of like that. So there's an existence proof in a particular constraint environment that a process akin to what many people call reasoning exists. But more general kind of reasoning. So off the board, there is one other existence proof, boy, which one us humans? Yes. Okay. All right. So do you think the architecture that will allow neural networks to reason will look similar to the neural network architectures we have today? I think it's real. I think, well, I don't want to make two overly definitive statements. I think it's definitely possible that the neural networks that will produce the reasoning breakthroughs of the future will be very similar to the architecture that exists today. Maybe a little bit more recurrent, maybe a little bit deeper. But these neural heads are so insanely powerful. Why wouldn't they be able to learn to reason? Humans can reason. So why can't neural networks? Do you think the kind of stuff we've seen neural networks do is a kind of just weak reasoning? So it's not a fundamentally different process. Again, this is stuff we don't nobody knows the answer to. So when it comes to our neural networks, I would think which I would say is that neural networks are capable of reasoning. But if you train a neural network on a task, which doesn't require reasoning, it's not going to reason. This is a well known effect where the neural network will solve exactly the is will solve the problem that you pose in front of it in the easiest way possible. Right. That takes us to the, to one of the brilliant sort of ways you describe neural networks, which is you refer to neural networks as the search for small circuits and maybe general intelligence as the search for small programs, which I found is a metaphor very compelling. Can you elaborate on that difference? Yeah. So the thing which I said precisely was that if you can find the shortest program that outputs the data in your, at your disposal, then you will be able to use it to make the best prediction possible. And that's a theoretical statement, which can be proven mathematically. Now you can also prove mathematically that it is that finding the shortest program with generates some data is not, is not a computable operation. No finite amount of compute can do this. So then with neural networks, neural networks are the next best thing that actually works in practice. We are not able to find the best, the shortest program which generates our data. But we are able to find, you know, a small, but now, now that statement should be amended, even a large circuit, which fits our data in some way. Well, I think what you meant by the small circuit is the smallest needed circuit. Well, I see the thing, the thing which I would change now back, back then I really have, I haven't fully internalized the over parameter, the over parameterized results. The, the things we know about over parameterized neural nets. Now I would phrase it as a large circuit that whose weights contain a small amount of information, which I think is what's going on. If you imagine the training process of a neural network as you slowly transmit entropy from the data set to the parameters, then somehow the amount of information in the weights ends up being not very large, which we would explain by the generalize so well. So that's, that the large circuit might be one that's helpful for the regular, for the generalization. Yeah, something like this. But do you see their, do you see it important to be able to try to learn something like programs? I mean, if we can definitely, I think it's kind of the answer is kind of yes, if we can do it. We should do things that we can do it. It's, it's the reason we are pushing on deep learning, the fundamental reason the, the, the root cause is that we are able to train them. So in other words, training comes first. We've got our pillar, which is the training pillar. And now we are trying to contort our neural networks around the training pillar. We got a state trainable. This isn't in, this is an invariant. We cannot violate. And so being trainable means starting from scratch, knowing nothing, you can actually pretty quickly converge towards knowing a lot or even slowly. But it means that given the resources at your disposal, you can train the neural net and get it to achieve useful performance. Yeah, that's a pillar we can't move away from. That's right. Because if you can, if you say, hey, let's find the shortest program, well, we can't do that. So it doesn't matter how useful that would be, we can do it. So we want. So do you think you kind of mentioned that the neural network is a good finding small circuits or large circuits? Do you think then the matter of finding small programs is just the data? No. So the, sorry, not the size or character, the quality, the type of data, sort of ask, giving it programs. Well, I think the thing is that right now finding, there are no good precedence of people successfully finding programs really well. And so the way you'd find programs is you train a deep neural network to do it basically, right? Which is, which is the right way to go about it. But there's not good illustrations that has hasn't been done yet. But in principle, it should be possible. Can you elaborate another bit? What's your insight in principle? And put another way, you don't see why it's not possible? Well, it's kind of like more, it's more a statement of, I think that it's unviced to bet against deep learning. And if it's a cognitive function that humans seem to be able to do, then it doesn't take too long for some deep neural net to pop up that can do it to. Yeah, I'm there with you. I can, I've stopped betting against neural networks at this point, because I continue to surprise us. What about long-term memory? Can neural networks have long-term memory or something like knowledge basis?

Long-term memory (50:35)

So being able to aggregate important information over long periods of time that would then serve as useful sort of representations of state that you can make decisions by. So have a long-term context based on what you make into the decision. So in some sense, the parameters already do that. The parameters are an aggregation of the day of the neural of the entirety of the neural net experience. And so they count as long as long as long-term knowledge. And people have trained various neural nets to act as knowledge bases and, you know, investigated with invest people have investigated language to models as knowledge bases. So there is work, there is work there. Yeah, but in some sense, do you think in every sense, do you think there's a, it's, it's all just a matter of coming up with a better mechanism of forgetting the useless stuff and remembering the useful stuff? Because right now, I mean, there's not been mechanisms that do remember really long-term information. What do you mean by that precisely? Precisely. I like, I like the word precisely. So I'm thinking of the kind of compression of information that knowledge bases represent. Sort of creating a, now I apologize for my sort of human centric thinking about what knowledge is because neural networks aren't interpretable necessarily with the kind of knowledge they have discovered. But a good example for me is knowledge bases being able to build up over time something like the knowledge that Wikipedia represents. It's a really compressed, uh, knowledge base. Obviously not the actual Wikipedia or the language, but like a semantic web, the dream that semantic web represented. So it's a really nice compressed knowledge base or something akin to that in a non interpretable sense as neural networks would have. Well, the neural networks would be not interpretable if you look at their rates, but their outputs should be very interpretable. Okay, so yeah, how do you, how do you make very smart neural networks like language models interpretable? Well, you asked them to generate some text and the text will generally be interpretable. Do you find that the epitome of interpretability? Like, can you do better? Like, can you, uh, because you can't, okay, I'd like to know what does it know and what doesn't know? I would like the neural network to come up with examples where it's completely dumb and examples where it's completely brilliant. And the only way I know how to do that now is to generate a lot of examples and use my human judgment. But it would be nice if a new now we had some aware self awareness about, you know, 100% I'm a big believer in self awareness and I think that I think, I think neural net self awareness will allow for things like the capabilities like the ones you described, like for them to know what they know and what they don't know and for them to know where to invest to increase their skills most optimally. And to your question of interpretability, there are actually two answers to that question. One answer is, you know, we have the neural net so we can analyze the neurons and we can try to understand what the different neurons and different layers mean. And you can actually do that and open AI has done some work on that. But there is a different answer which is that, I would say that's the human centric answer where you say, you know, you look at a human being, you can't read, you know, how do you know what a human being is saying? And you ask them, you say, Hey, what do you think about this? What do you think about that? And you get some answers. The answers you get are sticky in the sense you already have a mental model. You already have an yeah, mental model of that human being. You already have an understanding of like a big conception of what it of that human being, how they think, what they know, how they see the world and then everything you ask, you're adding on to that. And that stickiness seems to be that's one of the really interesting qualities of the human being is that information is sticky. You don't you seem to remember the useful stuff, aggregate it well and forget most of the information that's not useful, that process. But that's also pretty similar to the process and neural networks too, is just that neural networks are much crappier at this time. It doesn't seem to be fundamentally that different. But just stick on reasoning for a little longer. You said why not? Why can't there reason? What's a good impressive feat benchmark to you of reasoning that you'll be impressed by it. You're not also able to do. Is that something you already have in mind? Well, I think writing writing really good code, I think, proving really hard theorems, solving open-ended problems without the box solutions. And sort of theorem type mathematical problems. Yeah, I think those ones are a very natural example as well. You know, if you can prove an unproven theorem, then it's hard to argue, don't reason. And so by the way, and this comes back to the point about the hard results. You know, if you've got the hard, if you have machine learning, deep learning as a field is very fortunate because we have the ability to sometimes produce these unambiguous results. And when they happen, the debate changes, the conversation changes. It's a conversation, we have the ability to produce conversation change in results. conversation. And then of course, just like you said, people kind of take that for granted, say that wasn't actually a hard problem. Well, I mean, at some point, we'll probably run out of hard problems. Yeah, that whole mortality thing is kind of a sticky problem that we haven't quite figured out. Maybe we'll solve that one. I think one of the fascinating things in your entire body of work, but also the work at OpenAI recently, one of the conversation changes has been in the world of language models.

Language models (56:37)

Can you briefly kind of try to describe the recent history of using neural networks in the domain of language and text? Well, there's been lots of history. I think I think the Elman network was a small tiny recurrent neural network applied to language back in the 80s. So the history is really, you know, fairly long at least. And the thing that started, the thing that changed the trajectory of neural networks and language is the thing that changed the trajectory of all the deep learning and that's data and compute. So suddenly you move from small language models, which learn a little bit. And with language models, in particular, you can, there's a very clear explanation for why they need to be large, to be good, because they're trying to predict the next word. So we don't, we don't know anything. You'll notice very, very broad strokes, surface level patterns, like sometimes there are characters and there is a space between those characters. You'll notice this pattern. And you'll notice that sometimes there is a coma and the next character is a capital letter. You'll notice that pattern. Eventually you may start to notice that there are certain words occur often. You may notice that spellings are a thing. You may notice syntax. And when you get really good at all these, you start to notice the semantics. You start to notice the facts. But for that to happen, the language model needs to be larger. So that's, let's linger on that. That's where you and no jobs get disagree. So you think we're actually taking incremental steps, a sort of larger network, larger compute, we'll be able to get to the semantics, to be able to understand language without what Noam likes to sort of think of as a fundamental understandings of the structure of language, like imposing your theory of language onto the learning mechanism. So you're saying the learning, you can learn from raw data, the mechanism that underlies language. Well, I think it's pretty likely. But I also want to say that I don't really know precisely what Chomsky means when he talks about him. You said something about imposing your structure and language. I'm not 100% sure what he means. But empirically, it seems that when you inspect those larger language models, they exhibit signs of understanding the semantics, where the smaller language models do not. We've seen that a few years ago when we did work on the sentiment neuron, we trained a small, you know, small, shell STM to predict the next character in Amazon reviews. And we noticed that when you increase the size of the LSTM from 500 LSTM cells to 4000 LSTM cells, then one of the neurons starts to represent the sentiment of the article of sorry of the review. Now, why is that? Sentiment is a pretty semantic attribute. It's not a syntactic attribute. And for people who might not know, I don't know if that's a standard term, but sentiment is whether it's a positive or negative review. That's right. Like, is the person happy with something or is the person unhappy with something? And so here we had very clear evidence that a small neural net does not capture sentiment while a large neural net does. And why is that? Well, our theory is that at some point, you run out of syntax to models, you start to got to focus on something else. And with size, you quickly run out of syntax to model, and then you really start to focus on the semantics is would be the idea. That's right. And so I don't want to imply that our models have complete semantic understanding because that's not true. But they definitely are showing signs of semantic understanding partial semantic understanding. But the smaller models do not show that those signs. Can you take a step back and say, what is GPT two, which is one of the big language models that was the conversation change in the past couple years?

GPT-2 (01:00:35)

Yes, so GPT two is a transformer with one and a half billion parameters that was trained on about 40 billion tokens of text, which were obtained from web pages that were linked to from Reddit articles with more than three uploads. And what's the transformer? The transformer, it's the most important advance in neural network architectures in recent history. What is attention maybe to? Because I think that's the interesting idea, not necessarily sort of technically speaking, but the idea of attention versus maybe what recur, recurrent neural networks represent. Yeah. So the thing is the transformer is a combination of multiple ideas simultaneously which attention is one. Do you think attention is the key? No, it's a key, but it's not the key. The transformer is successful because it is the simultaneous combination of multiple ideas. And if you were to remove either idea, it would be much less successful. So the transformer uses a lot of attention, but attention exists for a few years. So that can't be the main innovation. The transformer is designed in such a way that it runs really fast on the GPU. And that makes a huge amount of difference. This is one thing. The second thing is that transformer is not recurrent. And that is really important too because it is more shallow and therefore much easier to optimize. So in other words, it uses attention. It is a really great fit to the GPU. And it is not recurrent, so therefore less deep and easier to optimize. And the combination of those factors make it successful. So now it makes great use of your GPU. It allows you to achieve better results for the same amount of compute. And that's why it's successful. Were you surprised how well transformers worked and GPT-2 worked? So you worked on language. You've had a lot of great ideas before transformers came about in language. So you got to see the whole set of revolutions before and after. Were you surprised? Yeah, a little. A little. I mean, it's hard to remember because you adapt really quickly. But it definitely was surprising. It definitely was. In fact, I'll retract my statement. It was pretty amazing. It was just amazing to see, generate this text of this. And you know, I've got to keep in mind that we've seen at that time, we've seen all this progress in GANS. In improving the samples produced by GANS were just amazing. You have these realistic faces, but text hasn't really moved that much. And suddenly we moved from, you know, whatever GANS were in 2015 to the best, most amazing GANS in one step. And I was really stunning. Even though theory predicted, yeah, you train a big language model, of course you should get this. But then to see it with your own eyes, it's something else. And yet we adapt really quickly. And now there's sort of some cognitive scientist, right? Articles saying that GPT two models don't truly understand language. So we adapt quickly to how amazing the fact that they're able to model the language so well is. So what do you think is the bar for what for impressing us that it? I don't know. Do you think that bar will continuously be moved? Definitely. I think when you start to see really dramatic economic impact, that's when I think that's in some sense the next barrier. Because right now if you think about the working AI, it's really confusing. It's really hard to know what to make of all these advances. It's kind of like, okay, you got an advance. And now you can do more things and you got another improvement and you got another cool demo. At some point, I think people who are outside of AI, they can no longer distinguish this progress anymore. So we were talking offline about translating Russian to English and how there's a lot of brilliant work in Russian that the rest of the world doesn't know about. That's true for Chinese. It's true for a lot of scientists and just artistic work in general. Do you think translation is the place where we're going to see sort of economic big impact? I don't know. I think there is a huge number of... I mean, first of all, I want to point out that translation already today is huge. I think billions of people interact with big chunks of the internet primarily through translation. So translation is already huge and it's hugely positive too. I think self-driving is going to be hugely impactful. You know, it's unknown exactly when it happens. But again, I would not bet against deep learning. So that's deep learning in general. Deep learning for self-driving. Yes, deep learning for self-driving. But I was talking about sort of language models. I see. Just to check. I veered off a little bit. Just to check. You're not seeing a connection between driving and language. No, no. Okay. All right. Both using neural nets. That would be a poetic connection. I think there might be some... Like you said, there might be some kind of unification towards a kind of multitask transformers that can take on both language and vision tasks. That'd be an interesting unification. Let's see. What can I ask about GPT two more? It's simple. So not much to ask. It's... You take a transform, you make it bigger, give it more data. And suddenly it does all those amazing things. Yeah. One of the beautiful things is that GPT, the transformers are fundamentally simple to explain, to train. Do you think bigger will continue to show better results in language? Probably. Sort of like what are the next steps with GPT two, do you think? I mean, I think for sure seeing what larger versions can do is one direction. Also, I mean, there are many questions. There's one question which I'm curious about next following. So right now GPT two, so we feel it all is data from the internet, which means that it needs to memorize all those random facts about everything in the internet. And it would be nice if the model could somehow use its own intelligence to decide what data it wants to accept and what data it wants to reject. Just like people, people don't learn all data indiscriminately. We are super selective about what we learn. And I think this kind of active learning, I think, would be very nice to have. Yeah, listen, I love active learning. So let me ask, does the selection of data, can you just elaborate that a little bit more?

Advancements In Ai And Agi

Active learning (01:07:14)

Do you think the selection of data is like, I have this kind of sense that the optimization of how you select data, so the active learning process is going to be a place for a lot of breakthroughs, even in the near future, because there hasn't been many breakthroughs there that are public. I feel like there might be private breakthroughs that companies keep to themselves, because the fundamental problem has to be solved if you want to solve self-driving, if you want to solve a particular task. What do you think about the space in general? Yeah, so I think that for something like active learning, or in fact for any kind of capability, like active learning, the thing that it really needs is a problem. It needs a problem that requires it. It's very hard to do research about the capability if you don't have a task, because then what's going to happen is that you will come up with an artificial task, get good results, but not really convince anyone. Right, we're now past the stage where getting a result in MNIST, some clever formulation of MNIST will convince people. That's right, in fact you could quite easily come up with a simple active learning scheme on MNIST and get a 10x speed up, but then so what. I think that with active learning, active learning will naturally arise as problems that require it pop up. That's how I would, that's my take on it. There's another interesting thing that OpenAI has brought up with GPT-2, which is when you create a powerful artificial intelligence system, and it was unclear what kind of detrimental, once you release GPT-2, what kind of detrimental effect it will have.

Staged release of AI systems (01:08:52)

Because if you have a model that can generate a pretty realistic text, you can start to imagine that it would be used by bots in some way that we can't even imagine. There's this nervousness about what it's possible to do. You did a really kind of brave and profound thing, which started a conversation about this. How do we release powerful artificial intelligence models to the public? If we do it all, how do we privately discuss with other even competitors about how we manage the use of the systems and so on. From this whole experience, you've released a report on it, but in general, are there any insights that you've gathered from just thinking about this, about how you release models like this? I think that my take on this is that the field of AI has been in a state of childhood, and now it's exiting that state and it's entering a state of maturity. What that means is that AI is very successful and also very impactful, and its impact is not only large, but it's also growing. For that reason, it seems wise to start thinking about the impact of our systems before releasing them, maybe a little bit too soon rather than a little bit too late. And with the case of GPT-2, like I mentioned earlier, the results really were stunning, and it seemed plausible. It didn't seem certain. It seemed plausible that something like GPT-2 could easily use to reduce the cost of this information. And so there was a question of what's the best way to release it and a staged release, sim logical. A small model was released, and there was time to see the... Many people use these models in lots of cool ways. There have been lots of really cool applications. There haven't been any negative application we know of, and so eventually it was released, but also other people replicated similar models. That's an interesting question though that we know of. So in your view, stage release, as at least part of the answer to the question of how do we... How... What do we do once we create a system like this? It's part of the answer, yes. Is there any other insights? Say you don't want to release the model at all, because it's useful to you for whatever the business is. While there are plenty of people who don't release models already, right, of course. But is there some moral ethical responsibility when you have a very powerful model to sort of communicate? Like, just as you said, when you had GPT-2, it was unclear how much it could be used for misinformation. It's an open question, and getting an answer to that might require that you talk to other really smart people that are outside of your particular group. Have you... Please tell me there's some optimistic pathway for people across the world to collaborate on these kinds of cases? Or is it still really difficult from one company to talk to another company? So it's definitely possible. It's definitely possible to discuss these kind of models with colleagues elsewhere and to get their take on what to do. How hard is it though? I mean, do you see that happening? I think that's a place where it's important to gradually build trust between companies, because ultimately, all the AI developers are building technology, which is becoming increasingly more powerful. And so it's... The way to think about it is that it's ultimately be only together. Yeah, it's... I tend to believe in the better angels of our nature, but I do hope that when you build a really powerful AI system in a particular domain, that you also think about the potential negative consequences of... Yeah. It's an interesting and scary possibility that there would be a race for AI development that would push people to close that development and not share ideas with others. I don't love this. I've been in a pure academic for 10 years. I really like sharing ideas and it's fun and it's exciting. What do you think it takes to... Let's talk about AGI a little bit. What do you think it takes to build a system of human-level intelligence?

How to build AGI? (01:13:41)

We talked about reasoning, we talked about long-term memory, but in general. What does it take, do you think? Well, I can't be sure, but I think the deep learning plus maybe another small idea. Do you think self-play will be involved? Sort of like you've spoken about the powerful mechanism of self-play where systems learn by sort of exploring the world in a competitive setting against other entities that are similarly skilled as them and so incrementally improving this way. Do you think self-play will be a component of building an AGI system? Yeah. So what I would say, to build AGI, I think it's going to be deep learning plus some ideas and I think self-play will be one of those ideas. I think that that is a very... Self-play has this amazing property that it can surprise us in truly novel ways. For example, I mean pretty much every self-play system, both our daughter bought, I don't know if... OpenAI had a release about multi-agent where you had two little agents who were playing hide and seek and of course also alpha zero. They were all produced surprising behaviors. They all produced behaviors that we didn't expect. They are creative solutions to problems and that seems like an important part of AGI that our systems don't exhibit routinely right now and so that's why I like this area, I like this direction because of its ability to surprises. To surprises. And AGI system would surprise us on the matter. Yes. And to be precise, not just a random surprise but to find a surprising solution to a problem that's also useful. Now a lot of the self-play mechanisms have been used in the game context or at least in the simulation context. How far along the path to AGI do you think will be done in simulation? How much faith promise do you have in simulation versus having to have a system that operates in the real world? Whether it's the real world of digital real world data or real world like actual physical world of robotics. I don't think it's an either or I think simulation is a tool and it helps. It has certain strengths and certain weaknesses and we should use it. Yeah but okay I understand that that's true but one of the criticisms of self-play, one of the criticisms of reinforcement learning is one of the its current power, its current results while amazing have been demonstrated in a simulated environments or very constrained physical environments. Do you think it's possible to escape them, escape the simulated environments and be able to learn in non-simulated environments or do you think it's possible to also just simulate in the photo realistic and physics realistic way the real world in a way that we can solve real problems with self-play in simulation? So I think that transfer from simulation to the real world is definitely possible and has been exhibited many times in by many different groups. It's been a specially successful envision. Also open AI in the summer has demonstrated a robot hand which was trained entirely in simulation in a certain way that allowed for seem to real transfer to occur. This is for the Rubik's cube? Yes that's right and so I wasn't aware that was trained in simulation. Was training simulation entirely? Really so it wasn't in the physics that the hand wasn't trained. No 100% of the training was done in simulation and the policy that was learned in simulation was trained to be very adaptive. So adaptive that when you transfer it it could very quickly adapt to the physical world. So the kind of perturbations with the giraffe or whatever the heck it was those weren't were those part of the simulation? Well the simulation was generally so the simulation was trained to be robust to many different things but not the kind of perturbations we've had in the video. So it's never been trained with the glove it's never been trained with a stuff giraffe. So in theory these are novel perturbations. Correct it's not in theory in practice and that those are novel perturbations. Well that's okay. That's a clean small scale but clean example of a transfer from the simulated world to the physical world. Yeah and now we'll also say that I expect the transfer capabilities of deep learning to increase in general and the better the transfer capabilities are the more useful simulation will become because then you could take you could experience something in simulation and then learn a moral of the story which you could then carry with you to the real world. As humans do all the time and they play computer games. So let me ask sort of an embodied question staying on AGI for a sec. Do you think AGI says that we need to have a body we need to have some of those human elements of self-awareness, consciousness, sort of fear of mortality, self-preservation in the physical space which comes with having a body. I think having a body will be useful. I don't think it's necessary but I think it's very useful to have a body for sure because you can learn a whole new you can learn things which cannot be learned without a body but at the same time I think that you can if you don't have a body you could compensate for it and still succeed. You think so? Yes well there is evidence for this for example there are many people who were born deaf and blind and they were able to compensate for the lack of modalities. I'm thinking about Helen Keller specifically. So even if you're not able to physically interact with the world and if you're not able to I mean I actually was getting at maybe let me ask on the more particular I'm not sure if it's connected to having a body or not but the idea of consciousness and a more constrained version of that is self-awareness. Do you think any GI system should have consciousness? We can't define God. Whatever the heck you think consciousness is. Yeah. Card question to answer given how hard it is to find it. Do you think it's useful to think about? I mean it's definitely interesting. It's fascinating. I think it's definitely possible that our systems will be conscious. Do you think that's an emergent thing that just comes from? Do you think consciousness could emerge from the representation that's stored within your networks? So like that it naturally just emerges when you become more and more. You're able to represent more and more over the world? Well I'd say I'd make the following argument which is humans are conscious and if you believe that artificial neural nets are sufficiently similar to the brain then there should at least exist artificial neural nets you should be conscious to. You're leaning on that existence proof pretty heavily. Okay. But that's the best sense where I can give. No I know. I know. I know. There's still an open question if there's not some magic in the brain that we're not. I mean I don't mean a non-materialistic magic but that the brain might be a lot more complicated and interesting that we give a credit for. If that's the case then it should show up and at some point we'll find out that we can't continue to make progress but I think it's unlikely. So we talk about consciousness but let me talk about another poorly defined concept of intelligence. Again we've talked about reasoning. We've talked about memory. What do you think is a good test of intelligence for you? Are you impressed by the test that Alan Turing formulated with the imitation game of natural language? Is there something in your mind that you will be deeply impressed by if a system was able to do? I mean lots of things. There's certain there's a certain frontier of capabilities today and there exist things outside of that frontier and I would be impressed by any such thing. For example I would be impressed by a deep learning system which solves a very pedestrian, pedestrian task like machine translation or computer vision task or something which never makes mistake a human wouldn't make under any circumstances. I think that is something which have not yet been demonstrated and I would find it very impressive. Yes so right now they make mistakes and they might be more accurate than human beings but they still make a difference out of mistakes. So my I would guess that a lot of the skepticism that some people have about deep learning is when they look at their mistakes and they say well those mistakes they make no sense. Like if you understood the concept you wouldn't make that mistake. And I think that changing that would be that would inspire me that would be yes this is this progress. Yeah that's a really nice way to put it but I also just don't like that human instinct to criticize a model is not intelligent. That's the same instinct as we do when we criticize any group of creatures as the other because it's very possible that GPT2 is much smarter than human beings and many things. That's definitely true. It has a lot more breadth of knowledge. Yes breadth of knowledge and even perhaps depth on certain topics. It's kind of hard to judge what depth means but there's definitely a sense in which humans don't make mistakes that these models do. The same is applied to autonomous vehicles. The same is probably going to continue being applied to a lot of artificial intelligence systems. We find this is the annoying this is the process of in the 21st century the process of analyzing the progress of AI is the search for one case where the system fails in a big way where humans would not and then many people writing articles about it and then broadly as the public generally gets convinced that the system is not intelligent and we like pacify ourselves by thinking it's not intelligent because of this one anecdotal case and this seems to continue happening. Yeah I mean there is truth to that although there is people although I'm sure that plenty of people are also extremely impressed by the systems that exist today but I think this connects to the earlier point we discussed that it's just confusing to judge progress in AI. Yeah and you know you have a new robot demonstrating something how impressed should you be and I think that people will start to be impressed once AI starts to really move the needle on the GDP. So you're one of the people that might be able to create an AGS system here not you but you and open AI if you do create an AGS system and you get the spend sort of the evening with it him her what would you talk about do you think the very first time?

Question to AGI (01:25:00)

First time? Well the first time would just ask all kinds of questions and try to make it to get it to make a mistake and that would be amazed that it doesn't make mistakes and just keep keep asking broad. Okay what kind of questions do you think would they be factual or would they be personal emotional psychological what do you think all of their Bob would you ask for advice? Definitely. I mean why would I limit myself talking to a system like this? Now again let me emphasize the fact that you truly are one of the people that might be in the room where this happens. So let me ask sort of a profound question about I just talked to Stalin historian. I've been talking to a lot of people who are studying power. Abraham Lincoln said nearly all men can stand adversity but if you want to test a man's character give him power. I would say the power of the 21st century maybe the 22nd but hopefully the 21st would be the creation of an AGI system and the people who have control direct possession and control of the AGI system. So what do you think after spending that evening having a discussion with the AGI system what do you think you would do? Well the ideal world would like to imagine is one where humanity alike the board members of a company where the AGI is the CEO. So it would be I would like the picture which I would imagine is you have some kind of different entities different countries or cities and the people that leave their vote for what the AGI that represents them should do and the AGI that represents them goes and does it. I think a picture like that I find very appealing and you could have multiple aid you would have an AGI for a city, for a country and it would be trying to in effect take the democratic process to the next level and the board can always fire the CEO essentially press the reset button say. Press the reset. We randomize the parameters. But let me sort of that's actually okay that's a beautiful vision I think as long as it's possible to press the reset button. Do you think it will always be possible to press the reset button? So I think that it's definitely really possible to build. So the question that I really understand from you is will humans or humans people have control over the AI systems of the build? Yes. And my answer is it's definitely possible to build AI systems which will want to be controlled by their humans. Wow that's part of their so it's not that just they can't help but be controlled but that's that's the they exist. One of the objectives of their existence is to be controlled in the same way that human parents generally want to help their children they want their children to succeed. It's not a burden for them they are excited to help the children and to feed them and to dress them and to take care of them and I believe with high conviction that the same will be possible for an AGI. It will be possible to program an AGI to design it in such a way that it will have a similar deep drive that it will be delighted to fulfill and the drive will be to help humans flourish. But let me take a step back to that moment where you create the AGI system. I think this is a really crucial moment and between that moment and the the Democratic board members with the AGI at the head there has to be a relinquishing power. So it's George Washington despite all the bad things he did one of the big things he did is you relinquished power. First of all didn't want to be president and even when he became president he gave he didn't keep just serving as most dictators do for indefinitely. Do you see yourself being able to relinquish control over an AGI system given how much power you can have over the world at first financial just make a lot of money and then control by having possession as AGI system? I'd find it trivial to do that. I'd find it trivial to relinquish this kind of part. I mean the kind of scenario you are describing sounds terrifying to me. That's all. I would absolutely not want to be in that position. Do you think you represent the majority or the minority of people in the AGI community? Well I mean the open question an important one. Our most people good is another way to ask it. So I don't know if most people are good but I think that when it really counts people can be better than we think. That's beautifully put. Are there specific mechanisms you can think of aligning AGI values to human values? Do you think about these problems of continued alignment as we develop the AI systems? Yeah definitely. In some sense the kind of question which you are asking is so if you have to translate the question to today's terms it would be a question about how to get an RL agent that's optimizing a value function which itself is learned. If you look at humans humans are like that because they reward function the value function of humans is not external it is internal. There are definite ideas of how to train a value function. Basically an objective and as objective as possible perception system that will be trained separately to recognize to internalize human judgments on different situations and then that component wouldn't be integrated as the base value function for some more capable RL system. You could imagine a process like this. I'm not saying this is the process I'm saying this is an example of the kind of thing you could do. So on that topic of the objective functions of human existence what do you think is the objective function that's implicit in human existence?

Philosophical Discussions

Meaning of life (01:32:07)

What's the meaning of life? I think the question is wrong in some way. I think that the question implies that the reason objective answer which is an external answer you know your meaning of life is x. I think what's going on is that we exist and that's amazing and we should try to make the most of it and try to maximize our own value and enjoyment of our very short time while we do exist. It's funny because action does require an objective function it's definitely theirs in some form but it's difficult to make it explicit and maybe impossible to make it explicit. I guess is what you're getting at and that's an interesting fact of an RL environment. Well what I was making is slightly different point is that humans want things and they're wants create the drives that cause them to you know our wants are our objective functions our individual objective functions. We can later decide that we want to change that what we wanted before is no longer good and we want something else. Yet where there's so dynamic there's got to be some underlying sort of Freud there's a thing there's like sexual stuff there's people think it's the fear of death and there's also the desire for knowledge and you know all these kinds of things of procreation the sort of all the evolutionary arguments it seems to be there might be some kind of fundamental objective function from from which everything else emerges but it seems that because that's very different. I mean I think I think that probably is an evolutionary objective function which is to survive and procreate and make you make your students succeed that would be my guess but it doesn't give an answer to the question of what's the meaning of life. I think you can see how humans are part of this big process this ancient process we are we exist on a small planet and that's it so given that we exist try to make the most of it and try to enjoy more and suffer less as much as we can. Let me ask two silly questions about life one do you have regrets moments that if you went back you would do differently and two are there moments that you're especially proud of that made you truly happy. So I can answer that I can answer both questions of course there are there's a huge number of choices and decisions that have made that with the benefit of hindsight I wouldn't have made them and I do experience some regret but you know I try to take solace in the knowledge that at the time I did the best I could and in terms of things that I'm proud of there are I'm very fortunate to have done things I'm proud of and they made me happy from some for some time but I don't think that that is the source of happiness. So your academic accomplishments all the papers you're one of the most cited people in the world all the breakthroughs I mentioned in computer vision and language and so on is what is the source of happiness and pride for you. I mean all those things are a source of pride for sure I'm very grateful for having done all those things and it was very fun to do them but happiness comes but you know you can happiness well my current view is that happiness comes from our to a lot to a very large degree from the way we look at things you know you can have a simple meal and be quite happy as a result or you can talk to someone and be happy as a result as well or conversely you can have a meal and be disappointed that the meal wasn't a better meal so I think a lot of happiness comes from that but I'm not sure I don't want to be too confident I being humble on the face of the uncertainty seems to be also part of this whole happiness thing well I don't think there's a better way to end it than meaning of life and discussions of happiness so Iliya thank you so much you've given me a few incredible ideas you've given the world many incredible ideas I really appreciate it and thanks for talking today yeah thanks for stopping by I really enjoyed it thanks for listening to this conversation with Iliya Siskar and thank you to our presenting sponsor Cash App please consider supporting the podcast by downloading Cash App and using code LEXPOTCAST if you enjoy this podcast subscribe on youtube review it with five stars and apple podcast support on patreon or simply connect with me on twitter @LEXFREEDMAN and now let me leave you with some words from Alan Turing on machine learning instead of trying to produce a program to simulate the adult mind why not rather try to produce one which simulates the child if this were then subjected to an appropriate course of education one would obtain the adult brain thank you for listening and hope to see you next time you