Date interview was first published: Mar 15 2023. https://www.youtube.com/watch?v=SjhIlw3Iffs Craig Smith 0:04 I'm Craig Smith and this is AI on AI. This week, I talked to Ilya Sutskever, a co founder and chief scientist of open AI, and one of the primary minds behind the large language model GPT. Three, and its public progeny. Chap. GPT, which I don't think it's an exaggeration to say is changing the world. This isn't the first time Ilya has changed the world. Geoff Hinton has said he was the main impetus for Alex snap the convolutional neural network, whose dramatic performance stung the scientific community in 2012. And set off the deep learning revolution. As is often the case, in these conversations, they assume a lot of knowledge on the part of listeners, primarily because I don't want to waste the limited time I have to speak to people like Delia, explaining concepts or people or events that can easily be Googled, or being the I should say, where the chat GPT can explain for you the conversation with Ilya Paul's a conversation with young Kuhn in a previous episode, so if you haven't listened to that episode, I encourage you to do so. Meanwhile, I hope you enjoy the conversation with Delia, as much as I do. Speaker 2 1:50 Yeah, it's terrific to meet you to talk to you. I've watched many of your talks online and read many of your papers. Can you start just by introducing yourself a little bit of your background? I know you were born in Russia, where you were educated, what got you interested in computer science? If that was the initial impulse, or, or brain science, neuroscience, or whatever it was? And then I'll start asking questions. Yeah, I can talk about that a little bit. Ilya Sutskever 2:26 So yeah, indeed, I was born in Russia. I grew up in Israel. And then, as a teenager, my family emigrated to Canada. My parents say I was interested in AI from a pretty early age. I also was very motivated by consciousness, I was very disturbed by it. And I was curious about things that would help me understand it better. And AI seemed like a very, like a good angle there. So I think these were some of the ways that got me started. And I actually started working with Geoff Hinton very early, when I was 17. We moved to Canada. And I immediately was able to join the University of Toronto. And I really wanted to do machine learning, because that seemed like the most important aspects of artificial intelligence that at the time was completely inaccessible. Like to give some context, the year was 2003. We take it for granted that computers can learn. But in 2003, we took it for granted that computers can't learn. The biggest achievement of AI back then was deep blue, the chess playing engine. But there it was, like, you have this game, and you have this tree search, and you have this simple way of determining if one position is better than another. And it really did not feel like that could possibly be applicable to the real world. Because there is no learning and learning was this big mystery. And so I was really, really interested in learning. And to my great luck, Geoff Hinton was a professor in the University I was in, and I was able to find him. And we began working together almost straightaway. Speaker 2 4:17 And what was your impulses? It was for Jeff, to understand how the brain worked, or was it more that you were simply interested in the idea of machines? Learning Ilya Sutskever 4:30 AI is so big. And so the motivations were just as many like it is interesting, like how does intelligence work at all? Like right now we have quite a bit of an idea, but it's a big neural net, and we know how it works to some degree, but back then, all the things in neural nets were around no one knew that all nets are good for anything. So how does intelligence work at all? How can we make computers be even Slightly intelligent. And I had a very explicit intention to make a very small, but the real contribution to AI, because there were lots of contributions to AI, which weren't real, which were, but I could tell for various reasons that they weren't real, that nothing would come out of it. And I just thought, nothing works at all. AI is a hopeless field. So in the motivation was good, I understand how intelligence work, and also make a contribution towards it. So that was my initial early motivation. That's 2003, almost exactly 20 years ago. Speaker 2 5:40 And then Alex, I've spoken to Jeff. And he said that it was really your excitement about the breakthroughs in convolutional neural networks that led you to apply for the ImageNet competition. And that the Alex had the coding skills to train the network, can you talk just a little bit about that I don't want to get bogged down in history, but it's fascinating. Ilya Sutskever 6:11 So in a nutshell, I had the realization that if you train a large neural network, on a large, sorry, large and deep, because metal and the deep part was still new, if you train a large n, a deep neural network, on a big enough data set that specifies some complicated tasks that people do, such as vision, but also others, and you just train that neural network, then you will succeed necessarily. And the logic for it was very irreducible, where we know that the human brain can solve these tasks and can solve them quickly. And the human brain is just a neural network, the slow neurons. So we know that some neural network can do it really well. So then we just need to take a smaller but related neural network, and just train it on data and the best neural network inside the computer will be will be related to the neural network that we have performs this task. So it was an argument that the neural network, the large and deep neural network canceled the task. And furthermore, we have the tools to train it, that was the result of the technical work that was done in Jeff's lab. So you combine the two, we can train those neural networks, it needs to be big enough, so that if you train it, it would work well. And you need data, which can specify the solution. And with ImageNet, all the ingredients were there. Alex had these very fast convolutional kernels, image net had the large enough data and there was real opportunity to do something totally unprecedented. And it totally worked out. Yeah. Speaker 2 8:00 That was supervised learning, and convolutional neural nets. In 2017, the attention is all you need paper came out introducing self attention and transformers. At what point did the GPT project start? Was it was there some intuition about transformers? And self supervised learning? Can you talk about that? So for context, Ilya Sutskever 8:33 at open AI, from the earliest days, we were exploring the idea that predicting the next thing is all you need. We were exploring it with the much more limited neural networks of the time. But the hope was that if you have a neural network that can predict the next word, the next big, so really, it's about compression, prediction is compression. And predicting the next word is not it's let's see, when you think about the best way to explain it. Because there are there were many things going on and they were all related. Maybe I'll take a different direction. We were indeed interested in trying to understand how far predicting the next word is going to go, and whether it will solve unsupervised learning. So back before the GPT is unsupervised learning was considered to be the holy grail of machine learning. And now it's just been fully solved. The noggin even talks about it, but it was a holy grail. It was very mysterious. And so we were exploring the idea. I was really excited about it, that predicting the next word well enough, is going to give you unsupervised learning, you will learn everything about the data set. That's going to be great. But our neural networks were not up for the task. We were using recurrent neural networks. When then transformer came out. It was literally as soon as the paper came out literally the next day. It was clear to me to us that transformers address the limitations of recurrent neural networks of learning long term dependencies. It's a technical thing, but it was like the switch to transformers right away. And so the very nascent GBT effort continued, then. And then like with the transformer, it started to work better, and you make it bigger. And then we realized we need to keep making it bigger. And we did. And that's what led to eventually, GBT three, and essentially, where we are today. Speaker 2 10:36 Yeah. And I just wanted to ask, actually, I'm getting caught up in this history, but I'm so interested in it. I want to get to the problems, or the shortcomings of large language models or large models, generally. But Rich Sutton had been writing about scaling, and how that's all we need to do. We don't need new algorithms, we just need to scale. Did he have an influence on you? Or was that a parallel track of thinking? No, I would say that Ilya Sutskever 11:12 when he posted his article, then we were very pleased to see some external people thinking in similar lines, and we thought it was very eloquently articulated. But I actually think that's the bitter lesson as articulated, overstates its case, or at least I think the takeaway that people have taken from it, overstate its case, the takeaway that people have is doesn't matter what you do just scale. But that's not exactly true. You got to scale something specific, you got to have something that you'll be able to benefit from the scale. The great breakthrough of deep learning is that it provides us with the first ever way of productively using scale and getting something out of it in return. Like before that, like what would people use large computer clusters for? I guess they would do it for weather simulations, or physics simulations or something. But that's about it. Maybe movie making. But no one had any real need for compute clusters, because what do you do with them? The fact that deep neural networks when you make them larger, and you train them, and more data will work better, provided us with the first thing that is interesting to scale. But perhaps one day, we will discover it, there is some little twist on the thing that we scale, it's going to be even better to scale. Now, how big of a twist? And then of course, with the benefit of hindsight, he will say, Does it even count? It's such a simple change. But I think the true statement is that it matters what you scale. Right now, we just found like a thing to scale that gives us something in return. Speaker 2 13:09 The limitation of large language models as they exist, is their knowledge is contained in the language that they're trained on. And most human knowledge. I think everyone agrees is non linguistic, not sure, Noam Chomsky agrees. But there's a problem in large language models, as I understand that. Their objective is to satisfy the statistical consistency of the prompt. They don't have an underlying understanding of reality. That language relates to ask Chet GBT about myself. It recognize that I'm a journalist that I've worked at these various newspapers, but it went on and on about awards that I've never won and put it all read beautifully. But none of it connected to the underlying reality. Is there something that that is being done to address that? In your research going forward? Yeah. So before Ilya Sutskever 14:22 I comment on the immediate question that you ask, I want to comment about some of the earlier parts of the question. Sure. I think that it is very hard to talk about the limits, or limitations rather, of even something like a language model. Because two years ago, people confidently spoke about their limitations and they were entirely different. Right? So it's important to keep this context in mind. How confident are we that these limitations that we'll see today, it will still be Review this two years from now, I am not that confident there is another comment I want to make about one part of the question, which is that these models just learn to statistical regularities, and therefore, they don't really know what the nature of the world is. And I have a view that differs from Unknown Speaker 15:20 this. In other words, Ilya Sutskever 15:24 I think that learning the statistical regularities is a far bigger deal than meets the eye. The reason we don't initially think so is because we haven't, at least most people, those who haven't really spent a lot of time with neural networks, which are on some level statistical. Like what statistical model, you just fit some parameters on what is really happening. But think there is a better interpretation to the earlier point of prediction as compression prediction is also a statistical phenomenon. Yet to predict, you eventually need to understand the true underlying process that produced the data. To predict the data well, to compress it well, you need to understand more and more about the world that produced the data. As our generative models become extraordinarily good, they will have I claim a shocking degree of understanding, a shocking degree of understanding of the world. And many of its subtleties, but it's not just the world, it is the world as seen through the lens of text, it tries to learn more and more about the world, through a projection of the world, on the space of text, as expressed by human beings on the internet. But still, this text already expresses the world. And I'll give you an example. A recent example, which I think is really telling the fascinating. Speaker 2 16:57 So we've all heard of Sydney, Ilya Sutskever 17:02 beings alter ego. And I've seen this really interesting interaction with Sydney, over Sydney became combative and aggressive. When the user told him that he thinks that Google is a better search engine than bank. Now, how can we like what is a good way to think about this phenomenon? What's a good language? What's, what does it mean? You can say, well, like it's just predicting what people will do, and people will do this, which is true. But maybe you are reaching a point where the language of psychology is starting to be appropriate to understand the behavior Unknown Speaker 17:43 of these neural networks. Ilya Sutskever 17:48 Now, let's talk about the limitations. It is indeed the case that Unknown Speaker 17:55 these neural networks are they do have a tendency Ilya Sutskever 18:00 to hallucinate. But that's because a language model is great for learning about the world. But it is a little bit less great for producing good outcomes. And there are various technical reasons for that, which I could elaborate on, if you think it's useful, but it is right now look, at the second level skip that. There are technical reasons why a language model is much better at learning about the world, learning incredible representations of ideas, of concepts of people have processes that exist. But its outputs aren't quite as good as one would hope or is or rather, as good as they could be. Which is why for example, for a system, like Chad GPT is the language model that has an additional reinforcement learning training process. We call it reinforcement learning from human feedback. But the thing to understand about that process is this. You can say that the pre training process, and you just train a language model, you want to learn everything about the world, then the reinforcement learning from human feedback. Now we care about the outputs. Now we say anytime the output isn't appropriate, don't do this again, every time the output does not make sense, don't do this again. And it learns quickly to produce good outputs. But now it is the level of the outputs, which is not the case during pre training during the language model training process. Now, on the point of hallucinations, and it has a propensity of making stuff up. Indeed, Unknown Speaker 19:37 it is true. Ilya Sutskever 19:39 Right now, these neural networks, even chargeability makes things up from time to time. And that's something that also greatly limits their usefulness. But I'm quite hopeful that by simply improving this subsequent reinforcement learning from human feedback step, we could just teach it to not hallucinate. So, now you could say, is it really going to learn? My answer is, let's find out. Speaker 2 20:05 And that feedback loop is coming from the public chat GBT interface that if it tells me that, I want to Pulitzer, which unfortunately I didn't, I can tell it that it's wrong. And will that train it, or create some punishment or reward so that the next time I asked him, it'll be more accurate. Ilya Sutskever 20:36 The way we do things today is that we hire people to teach our neural net to behave to teach everybody to behave. And right now, the manner the precise manner in which they specify the desired behavior is a little bit different. But indeed, what you described is the way in which teaching is going to like, basically be that's the correct way to teach, you just interact with it, and it sees from your reaction even first, oh, that's not what you wanted, you are not happy with its output, therefore, the output was not good, and it should do something differently next time. Unknown Speaker 21:10 So in particular, Ilya Sutskever 21:13 hallucinations come up is one of the bigger issues and we'll see, but I think there is a quite a high chance that this approach will be able to address them completely. Speaker 2 21:24 I wanted to talk to you about a yellow Koons work on joint embedding predictive architectures. And his idea that what's missing from large language models is this underlying world model that is non linguistic, that the language model can refer to it's not something that's bill, I wanted to hear what you thought of that and whether you've explored that at all. Ilya Sutskever 21:54 So I reviewed the on the concept proposal, and there are a number of ideas there. And they expressed it in different language. And there are some maybe small differences from the current paradigm, but to my mind, they are not very significant. And I'd like to elaborate. The first claim is that it is desirable to for a system to have multimodal understanding where it doesn't just know about the world from text. And my comment on that will be that indeed, multimodal understanding is desirable, because you learn more about the world, you learn more about people, you learn more about their condition. And so the system will be able to understand what the task that it's supposed to solve and the people and what they want better. We have done quite a bit of work on that most notably in the form of two major neural nets that we've done. One is called clip. And when it's called Dolly, north, move towards this multimodal direction. But I also want to say that I don't see the situation as a binary either or, that if you don't have vision, if you don't understand the world visually or from video, then things will not work. And I'd like to make the case for that. So I think that some things are much easier to learn from images and diagrams, and so on. But I claim that you can still learn them from text only just more slowly. And I'll give you an example. Consider the notion of color. Surely, one cannot learn the notion of color from taxonomy. And yet, when you look at the embeddings, I need to make a small detour to explain the concept of an embedding. Yeah, every neural network represents words, sentences, concepts through representations, embeddings high dimensional vectors. And one thing that we can do is that we can look at those high dimensional vectors, and we can look at what's similar to what how does the network see this concept of that concept? And so we can look at the embeddings of colors. And embeddings of colors happen to be exactly right. You know, like, it knows that purple is more similar to blue than to red, and it knows that purple is less similar to red than oranges. It knows all those things just from text, how can that be. So if you have vision, the distinctions between color just jump at you, you immediately perceive them. Whereas with text, it takes you longer. Maybe you know how to talk and you already understand syntax and words and grammars and only much later you say oh, these colors actually started understanding. So this will be my point about the necessity of multi modality, which I claim it is not necessary, but it is most definitely useful. I think it's a good direction to pursue You just don't see it in such stark either or claims. Unknown Speaker 25:03 So Ilya Sutskever 25:07 the proposal in the paper makes a claim that one of the big challenges is predicting high dimensional vectors, which have uncertainty about them. So for example, predicting an image, like the paper makes a very strong claim there that it's a major challenge. And we need to use a particular approach to address that. But one thing, which I found surprising, or at least acknowledged, in the paper, is that the current auto regressive transformers already have the property. I'll give you two examples. One is, given one page in a book predict the next page in a book, that will be so many possible pages that follow it's a very complicated high dimensional space and video visit just fine. The same applies to images, these autoregressive transformers work perfectly on images, for example, like this opening, I've done work on the IGBT, we just took the transformer and we applied it to pixels. And it worked super well. And it could generate images in a very complicated and subtle ways. It has a very beautiful unsupervised representation learning these dolly one, same thing again, you just generate, think of it as large pixels, rather than Jeric million pixels, we cluster the pixels in large pixel imaginary 1000 large pixels. I believe Google's work on image generation from earlier this year called party, I believe they will also take a similar approach. So the part where I thought that the paper made a strong comment around while the current approaches can't deal with predicting high dimensional distributions, I think they definitely can. So maybe this is another point that I would make. Speaker 2 26:45 And then what you're talking about converting pixels into vectors, it's essentially turning everything in the language of the factory is like a string of text trying to Ilya Sutskever 27:00 define language, though, you turn it into a sequence. Yeah, a sequence of what, like, you could argue that human for a human life is a sequence of bits. Now, there are other things that were that people use right now like diffusion, where they produce those bits rather than one bit at a time they produce them in parallel. But I would argue that on some level, this distinction is immaterial. I claim that at some level doesn't really matter. What matters is in like, you can get a 10x efficiency gain, which is huge in practice. But conceptually, Speaker 2 27:36 I claim it doesn't matter. On this idea of having an army of human trainers that are working with chat GPT, or a large language model, to guide and in effect with reinforcement learning. just intuitively, that doesn't sound like an efficient way of teaching a model about the underlying reality of its language. Isn't there a way Unknown Speaker 28:17 of automating that? And to Younes? Credit, I think that's what he's talking about, is coming up with an algorithmic means of teaching a model the underlying reality without a human having to intervene. Ilya Sutskever 28:43 Yeah, so I have two comments on that. I think. So the first place, so I have a different view on the question. So I wouldn't agree with the phrasing of the question. I claim that our pre trained models already know everything they need to know about the underlying reality. They already have this knowledge of language, and also a great deal of knowledge about the processes that exist in the world that produce this language. And maybe I should reiterate this point, it's a small tangent, but I think it's so important. The thing that large generative models learn about their data and in this case, large language models about text data Unknown Speaker 29:33 are some Ilya Sutskever 29:36 compressed representations of the real world processes that produce this data, which means not only people and something about their thoughts, something about their feelings, but also something about the condition that people are in and the interactions that exist between them, the different situations a person can be. All of these are part of that compressed process that is represented by neural net to produce the text has a better language model, the better the generative model, the higher the fidelity, the more the better this, the better it captures this process. So that's the first comment that we'll make. And so in particular, I will say, the models already have the knowledge. Now, the army of teachers, as you phrase it, indeed, you know, when you want to build a system that performs as well as possible, you just say, okay, like, if this thing works, do more of that. But of course, those teachers are also using AI assistants, those teachers aren't on their own, they are working with our tools together, they are very efficient. It's like the tools are doing the majority of the work. But you do need to have, you need to have oversight, you need to have people reviewing the behavior, because you want to have it to eventually to achieve a very high level of reliability. But overall, I'll say that we are at the same time, this second step, after we take the finished, pre trained model. And then we apply the reinforcement learning on it. There is indeed a lot of motivation to make it as efficient and as precise as possible. So that the resulting language model it will be as well behaved as possible. So yeah, there is Unknown Speaker 31:25 these human teachers Ilya Sutskever 31:29 who are teaching them how to model the desired behavior, they are also using AI systems. And the manner in which they use AI systems is constantly increasing. So their own efficiency keeps increasing. So maybe this will be one way to answer this question. Yeah. Speaker 2 31:49 And so what you're saying is, through this process of eventually, the model will become more and more discerning more and more accurate in its outputs. Yes, and it's, Ilya Sutskever 32:04 that's right. There is an analogy here, which is it already knows all kinds of things. And now I just want to really say, No, this is not what we want. Don't do this here, you made a mistake here in the output. And of course, it's exactly as you say, with as much AI in the loop as possible. So that the teachers who are providing the final correction to the system, their work is amplified. They're working as efficiently as possible. So it's not unlike an education process, how to act well, in the world way. We need to do additional training, just to make sure that the model knows that hallucination is not okay ever. And then once it knows that, now you are Speaker 2 33:00 in business, outsourcing. And it's that reinforcement learning, human teacher loop that will teach it human Ilya Sutskever 33:09 teacher loop or some other variant. But there is definitely an argument to be made that something here should work. And we'll find out pretty soon. Speaker 2 33:20 I assume that's one of the questions. Where is this going? What research are you focused on right now? Ilya Sutskever 33:28 I can't talk in detail about the specific research that I'm working on. But I can mention a little bit. I can mention some of the research in broad strokes. And it would be something like I'm very interested in making those models more reliable, more controllable, make them learn faster from less data, less instructions, make them so that, indeed, they don't hallucinate. And I think that all this cluster of questions which I mentioned, they're all connected. Unknown Speaker 34:03 And there's also question Ilya Sutskever 34:06 of, how far in the future are we talking about in this question? And what I commented here on is the perhaps near future, Speaker 2 34:15 you talk about the similarities between the brain and neural nets is a very interesting observation that Geoff Hinton made to me. I'm sure it's not new to other people. But that large models are large language models, in particular, hold a tremendous amount of data with a modest number of parameters compared to the human brain, which has trillions and trillions of parameters, but a relatively small amount of data. Have you thought of it in those terms? And can you talk about what's missing and large models to pay Have more parameters to handle the data? Is that a hardware problem? Or a training problem? Ilya Sutskever 35:09 This comment which you made is related to one of the problems that I mentioned in the earlier questions of learning from less data. Indeed, the current structure of the technology does like a lot of data, especially early in training. Now later in training, it becomes a bit less data hungry, which is why, in the end, it can learn very, not as fast as people yet, but it can learn quite quickly. So already, that means that in some sense, do we even care that we need all this data to get to this point. But indeed, more generally, I think we will be possible to learn more from this data. Unknown Speaker 35:53 I think it's just, Ilya Sutskever 35:55 I think it's requires some creative ideas. But I think it is possible. And I think learning more from this data will unlock a lot of different possibilities, it will allow us to teach our eyes the skills that is missing and to convey to it our desires and preferences exactly how we want it to behave more recently. So I would say that the faster learning is indeed very nice. And although already after language models are trained, they can learn quite quickly, I think there was opportunities to do more there. Speaker 2 36:28 Kurdi maker commented that we need faster processors to be able to scale further. And it appears that the scaling of models that there's no end in sight, but the power required to train these models, we're reaching the limit, at least the socially accepted limit. Ilya Sutskever 36:59 So I just want to make one comment, which is, I don't remember the exact comment that I made that you're referring to. But you always want faster processors, of course, always want more of them. Of course, power keeps going up, generally speaking, the cost is going up. And the question that I would ask is not whether the cost is large, but whether the thing that we get out of being this cost outweighs the cost. Maybe by all this cost, you get nothing, then yeah, that's not worth it. But if you get something very useful, something very valuable, something that you can solve a lot of problems that we have, which we really want solved, then the cost can be justified. But in terms of the processes, faster processors, yeah. Speaker 2 37:44 Any day. Are you involved at all in hardware question? Work with cerebrus? For example, the wafer scale chips. Ilya Sutskever 37:58 Now all our hardware comes from Azure Speaker 2 38:02 to be used, they procure. Yeah, yeah. You did talk at one point I saw about democracy, and about the impact that that AI can have on democracy. People have talked to me about that, if you had enough data, and a large enough model, you could train the model on the data, and it could come up with an optimal solution that would satisfy everybody. Do you have any aspiration? Or do you think about where this might lead in terms of helping humans? Manage society? Yeah, Ilya Sutskever 38:47 let's see, it's such a big question. Because it's a much more future looking question. Like, I think that there is still many ways in which models will become far more capable than they are right now. There's no question. In particular, the way we train them, and use them and so on, there's going to be a few changes here and there. They might not be immediately obvious today. But I think in hindsight, it will be extremely obvious. That will indeed allow it to have the ability to come up with solutions to problems of this kind. It's unpredictable exactly how governments will use this technology as a source of getting advice of various kinds. I think that to the question of democracy, one thing which I think could happen in the future, is that because you have these neural nets and they're going to be so pervasive, and they're going to be so impactful in society, we will find that it is desirable to have some kind of a democratic process where this let's say the citizens of a country provide some information to the neural net about how they'd like things to be, how they'd like it to behave or something along these lines, I could imagine that happening, that can be a very, like a high bandwidth form of democracy, perhaps, where you get a lot more information out of each citizen, and you aggregate it to specify how exactly one such distance to act. Now, it opens a whole lot of questions. But that's one thing that could happen in the future. Yeah, Speaker 2 40:26 I can see in the democracy example, you give that that individuals would have the opportunity to input data, but it's sort of goes to the world model. Question. Do you think AI systems will eventually be large enough that they can understand a situation and analyze all of the variables? But you would need a model that does more than absorb language? I would think, Ilya Sutskever 41:02 what does it mean to analyze all the variables, eventually, there will be a choice, you need to make ReeseI. These variables similarly important, I want to go deep. Because a person can read the book, I can read 100 books, or I can read one book very slowly and carefully and get more out of it. So there will be some elements of that. Also, I think it's probably fundamentally impossible to understand everything in some sense. Anytime there is any kind of complicated situation in society, even in a company within a midsize company. It's already beyond the comprehension of any single individual. And I think that if we build our AI systems to go to the right way, I think AI could be incredibly helpful in pretty much any situation. Speaker 1 42:01 That's it for this episode. I want to thank Elio for his time. I also want to thank le George for helping arrange the interview. If you want to read a transcript of this conversation, you can find one on our website I on ay ay, that's EY e hyphen, o n.ai. We'd love to hear from listeners. So feel free to email me at Craig CR a IG at EY e hyphen, o n dot A i i get a lot of emails so put listener in the subject line so I don't miss it. We have listeners in 170 countries and territories. Remember, the singularity may not be near but AI is Changing your world. So pay attention