Sam Altman on AGI, Compute, and Human Agency

I think this will be the greatest thus I think this will be the greatest thus far far far technological achievement of human technological achievement of human technological achievement of human history. But the only way that it really history. But the only way that it really history. But the only way that it really matters is [music] if it makes people's matters is [music] if it makes people's matters is [music] if it makes people's lives like much better than they lives like much better than they lives like much better than they otherwise would have been. We are about otherwise would have been. We are about otherwise would have been. We are about to create a genie that can grant any to create a genie that can grant any to create a genie that can grant any wish because I think people will have wish because I think people will have wish because I think people will have such creative wishes and such incredible such creative wishes and such incredible such creative wishes and such incredible ideas of what they ask AI to help build, ideas of what they ask AI to help build, ideas of what they ask AI to help build, but concentration of power with AI is a but concentration of power with AI is a but concentration of power with AI is a terrifying thing. I don't think anyone terrifying thing. I don't think anyone terrifying thing. I don't think anyone should want to live in a world of, you should want to live in a world of, you should want to live in a world of, you know, AI overlords or company that is know, AI overlords or company that is know, AI overlords or company that is the rough equivalent of that. I think the rough equivalent of that. I think the rough equivalent of that. I think it's critical [music] we preserve that it's critical [music] we preserve that it's critical [music] we preserve that spirit with AI and that we all spirit with AI and that we all spirit with AI and that we all collectively have collectively have collectively have the ability to self-determine our the ability to self-determine our the ability to self-determine our future. >> So Sam, you [music] wrote a post that I >> So Sam, you [music] wrote a post that I thought was very simple and really thought was very simple and really thought was very simple and really interesting and a good place to start. interesting and a good place to start. interesting and a good place to start. Rounded to the last year's been really Rounded to the last year's been really Rounded to the last year's been really tough and that's some of my fault and tough and that's some of my fault and tough and that's some of my fault and the next year is going to be maybe our the next year is going to be maybe our the next year is going to be maybe our best 12 months. best 12 months. best 12 months. >> Yeah. >> Yeah. >> Yeah. >> I'd love you to reflect on on both. >> I'd love you to reflect on on both. >> I'd love you to reflect on on both. Maybe starting with why you said the Maybe starting with why you said the Maybe starting with why you said the first part and and why you believe the first part and and why you believe the first part and and why you believe the second part. second part. second part. >> On the first part, I think we just we're >> On the first part, I think we just we're >> On the first part, I think we just we're doing too many things. We're not focused doing too many things. We're not focused doing too many things. We're not focused enough and they're actually all good enough and they're actually all good enough and they're actually all good things to do, but the trick is we're in things to do, but the trick is we're in things to do, but the trick is we're in this like unbelievable moment in history this like unbelievable moment in history this like unbelievable moment in history where you can only do the very few great where you can only do the very few great where you can only do the very few great things. things. things. So we spread ourselves too thin and then So we spread ourselves too thin and then So we spread ourselves too thin and then made a bunch of difficult decisions to made a bunch of difficult decisions to made a bunch of difficult decisions to really refocus on having really refocus on having really refocus on having the best, most abundant, most the best, most abundant, most the best, most abundant, most cost-effective intelligence and cost-effective intelligence and cost-effective intelligence and empowering the world to build incredible empowering the world to build incredible empowering the world to build incredible things with that. Since doing that, I things with that. Since doing that, I things with that. Since doing that, I think our progress has been remarkable think our progress has been remarkable think our progress has been remarkable and just given what we see in the and just given what we see in the and just given what we see in the pipeline will be much more remarkable pipeline will be much more remarkable pipeline will be much more remarkable over the next 12 months. And the quality over the next 12 months. And the quality over the next 12 months. And the quality of the models that we'll have, the of the models that we'll have, the of the models that we'll have, the products that we can build around that products that we can build around that products that we can build around that to really let people thrive with this to really let people thrive with this to really let people thrive with this technology in in new ways. Uh, it should technology in in new ways. Uh, it should technology in in new ways. Uh, it should be pretty awesome. be pretty awesome. be pretty awesome. >> Was there a moment last year that >> Was there a moment last year that >> Was there a moment last year that something clicked for you that caused something clicked for you that caused something clicked for you that caused you to change directions or restack you to change directions or restack you to change directions or restack priorities or something. priorities or something. priorities or something. >> If you go back to the beginning of 2025, >> If you go back to the beginning of 2025, >> If you go back to the beginning of 2025, just a year and a half ago, just a year and a half ago, just a year and a half ago, >> Yeah. >> Yeah. >> Yeah. >> the big concern was companies like >> the big concern was companies like >> the big concern was companies like OpenAI are buying up so much compute. Is OpenAI are buying up so much compute. Is OpenAI are buying up so much compute. Is the revenue going to be there? Is the the revenue going to be there? Is the the revenue going to be there? Is the demand going to be there? demand going to be there? demand going to be there? And so we were trying to think about And so we were trying to think about And so we were trying to think about like a lot of things such as that if the like a lot of things such as that if the like a lot of things such as that if the revenue growth took longer revenue growth took longer revenue growth took longer to materialize than we thought it might, to materialize than we thought it might, to materialize than we thought it might, we could have you know, consumer apps we could have you know, consumer apps we could have you know, consumer apps and media and all these other things and media and all these other things and media and all these other things that could help us monetize the GPUs that could help us monetize the GPUs that could help us monetize the GPUs that we were signing up for. Uh again, that we were signing up for. Uh again, that we were signing up for. Uh again, it sounds ridiculous now because the it sounds ridiculous now because the it sounds ridiculous now because the revenue growth in the industry has been revenue growth in the industry has been revenue growth in the industry has been so steep, but so steep, but so steep, but that was the big change and then as soon that was the big change and then as soon that was the big change and then as soon as we realized like, okay, the model as we realized like, okay, the model as we realized like, okay, the model trajectory is growing so fast, there's trajectory is growing so fast, there's trajectory is growing so fast, there's such a clear economic return on these such a clear economic return on these such a clear economic return on these models, that was when we said, you know, models, that was when we said, you know, models, that was when we said, you know, we know what to focus on. we know what to focus on. we know what to focus on. >> I was reading some of your your great >> I was reading some of your your great >> I was reading some of your your great old posts from prior to OpenAI and one old posts from prior to OpenAI and one old posts from prior to OpenAI and one of them is this notion of like so much of them is this notion of like so much of them is this notion of like so much discussion of focus and the right amount discussion of focus and the right amount discussion of focus and the right amount of things to focus on. Is it one? Is it of things to focus on. Is it one? Is it of things to focus on. Is it one? Is it five? Is it three? How do you calibrate five? Is it three? How do you calibrate five? Is it three? How do you calibrate that in a business like this, especially that in a business like this, especially that in a business like this, especially in this period where you've said you in this period where you've said you in this period where you've said you needed to refocus? needed to refocus? needed to refocus? >> Fundamentally, >> Fundamentally, >> Fundamentally, our business is to sell AI that people our business is to sell AI that people our business is to sell AI that people will build will build will build incredible products and services for incredible products and services for incredible products and services for each other with. The components that I each other with. The components that I each other with. The components that I think of as going into that are we have think of as going into that are we have think of as going into that are we have to train great models that work to train great models that work to train great models that work in all the ways people want to use them. in all the ways people want to use them. in all the ways people want to use them. So great at coding, great at other kinds So great at coding, great at other kinds So great at coding, great at other kinds of knowledge, work great at doing of knowledge, work great at doing of knowledge, work great at doing science, like where the real economic science, like where the real economic science, like where the real economic value is. We have to produce or partner value is. We have to produce or partner value is. We have to produce or partner with these chips and systems, these, you with these chips and systems, these, you with these chips and systems, these, you know, hugely expensive racks that can do know, hugely expensive racks that can do know, hugely expensive racks that can do the AI computation. Uh we have to find the AI computation. Uh we have to find the AI computation. Uh we have to find enough enough enough uh land, power, data center shells to be uh land, power, data center shells to be uh land, power, data center shells to be able to put those racks somewhere. able to put those racks somewhere. able to put those racks somewhere. And then eventually, or maybe pretty And then eventually, or maybe pretty And then eventually, or maybe pretty soon, we have to soon, we have to soon, we have to build robots that can build robots that can build robots that can automate that process to continue to automate that process to continue to automate that process to continue to drive the cost down, the cost of drive the cost down, the cost of drive the cost down, the cost of producing electricity, chips, the whole producing electricity, chips, the whole producing electricity, chips, the whole supply chain. And that kind of whole supply chain. And that kind of whole supply chain. And that kind of whole stack of making the best, the most stack of making the best, the most stack of making the best, the most abundant, uh the most useful AI that we abundant, uh the most useful AI that we abundant, uh the most useful AI that we can, and making it something like can, and making it something like can, and making it something like electricity that just seeps throughout electricity that just seeps throughout electricity that just seeps throughout the entire economy and empowers people. the entire economy and empowers people. the entire economy and empowers people. That's kind of what I think we have to That's kind of what I think we have to That's kind of what I think we have to focus on. Building every vertical focus on. Building every vertical focus on. Building every vertical application on top of that, trying to go application on top of that, trying to go application on top of that, trying to go like eat every startup, eat every like eat every startup, eat every like eat every startup, eat every company. No interest in doing that. Uh company. No interest in doing that. Uh company. No interest in doing that. Uh really want to just provide that really want to just provide that really want to just provide that platform. platform. platform. >> This compute thing is one of the most >> This compute thing is one of the most >> This compute thing is one of the most interesting thing that's happened in interesting thing that's happened in interesting thing that's happened in human history, I think. And it's human history, I think. And it's human history, I think. And it's obviously coming to a head and put maybe obviously coming to a head and put maybe obviously coming to a head and put maybe will be coming to a head for a long will be coming to a head for a long will be coming to a head for a long period of time. This is something that I period of time. This is something that I period of time. This is something that I think Dario called you the YOLO CEO when think Dario called you the YOLO CEO when think Dario called you the YOLO CEO when you were doing some of this early you were doing some of this early you were doing some of this early compute allocation and and securing the compute allocation and and securing the compute allocation and and securing the compute. Obviously now you're in this compute. Obviously now you're in this compute. Obviously now you're in this position where everyone is short this position where everyone is short this position where everyone is short this stuff. stuff. stuff. >> Yeah. >> Yeah. >> Yeah. >> Is trying to find it. And I'd love to >> Is trying to find it. And I'd love to >> Is trying to find it. And I'd love to hear the early stories about why you hear the early stories about why you hear the early stories about why you gained conviction that you needed to gained conviction that you needed to gained conviction that you needed to secure everything that you did, how you secure everything that you did, how you secure everything that you did, how you did it, like it it seems to have been did it, like it it seems to have been did it, like it it seems to have been proven right. And maybe maybe you even proven right. And maybe maybe you even proven right. And maybe maybe you even underdid it, right? underdid it, right? underdid it, right? >> underdo it. >> underdo it. >> underdo it. >> Which is kind of crazy if you look at >> Which is kind of crazy if you look at >> Which is kind of crazy if you look at the headlines from back then. Can you the headlines from back then. Can you the headlines from back then. Can you tell me the early story of like how you tell me the early story of like how you tell me the early story of like how you came to that conclusion and what gave came to that conclusion and what gave came to that conclusion and what gave you the conviction to do it despite you the conviction to do it despite you the conviction to do it despite everyone everyone everyone thinking it was crazy. thinking it was crazy. thinking it was crazy. >> We could just tell that we were on this >> We could just tell that we were on this >> We could just tell that we were on this exponential of model improvement. exponential of model improvement. exponential of model improvement. That part we were very confident about. That part we were very confident about. That part we were very confident about. We knew it was going to keep going. We knew it was going to keep going. We knew it was going to keep going. We were pretty sure, although as you We were pretty sure, although as you We were pretty sure, although as you mentioned we underestimated, that as the mentioned we underestimated, that as the mentioned we underestimated, that as the models got better and better, if we models got better and better, if we models got better and better, if we could continue to drive costs down, that could continue to drive costs down, that could continue to drive costs down, that demand for AI at a sufficiently high demand for AI at a sufficiently high demand for AI at a sufficiently high level and a sufficiently low price was level and a sufficiently low price was level and a sufficiently low price was basically uncaptured. basically uncaptured. basically uncaptured. >> Mhm. >> Mhm. >> Mhm. >> This was just like a rare kind of new >> This was just like a rare kind of new >> This was just like a rare kind of new commodity for the world. Um but that commodity for the world. Um but that commodity for the world. Um but that what people would do with it reminded me what people would do with it reminded me what people would do with it reminded me of the way people used to talk about the of the way people used to talk about the of the way people used to talk about the early days of computing. People said, early days of computing. People said, early days of computing. People said, "Oh, there's you know, a market for five "Oh, there's you know, a market for five "Oh, there's you know, a market for five computers in the world." was one famous computers in the world." was one famous computers in the world." was one famous thing, or you know, "No one needs more thing, or you know, "No one needs more thing, or you know, "No one needs more than X amount of RAM." Human ingenuity, than X amount of RAM." Human ingenuity, than X amount of RAM." Human ingenuity, creativity, desire for stuff, desire to creativity, desire for stuff, desire to creativity, desire for stuff, desire to be useful, that's a very good thing to be useful, that's a very good thing to be useful, that's a very good thing to bet on. And we could see that AI was bet on. And we could see that AI was bet on. And we could see that AI was going to be an extremely important way going to be an extremely important way going to be an extremely important way that people expressed those things or that people expressed those things or that people expressed those things or got those things, did those things. And got those things, did those things. And got those things, did those things. And we knew that the algorithms would get we knew that the algorithms would get we knew that the algorithms would get more efficient and the models would get more efficient and the models would get more efficient and the models would get better, which of course they have. better, which of course they have. better, which of course they have. But we also knew that, no matter how But we also knew that, no matter how But we also knew that, no matter how efficient they got, efficient they got, efficient they got, you know, at some level, what we are you know, at some level, what we are you know, at some level, what we are about is turning about is turning about is turning electricity into electricity into electricity into useful intelligence. useful intelligence. useful intelligence. And we were going to need more of that. And we were going to need more of that. And we were going to need more of that. No matter how good we got that other No matter how good we got that other No matter how good we got that other layer, given this observation about layer, given this observation about layer, given this observation about demand, we were just going to want more. demand, we were just going to want more. demand, we were just going to want more. >> Did that start with GPT-3? Like if I >> Did that start with GPT-3? Like if I >> Did that start with GPT-3? Like if I were to trace the history of this as far were to trace the history of this as far were to trace the history of this as far back as possible, where would you put back as possible, where would you put back as possible, where would you put the first hash mark of that? the first hash mark of that? the first hash mark of that? >> I would say we got real conviction with >> I would say we got real conviction with >> I would say we got real conviction with GPT-4, not even 3.5. GPT-4, not even 3.5. GPT-4, not even 3.5. >> Hm. >> Hm. >> Hm. What was it? What was it? What was it? >> It was seeing the model was smart enough >> It was seeing the model was smart enough >> It was seeing the model was smart enough that we knew we'd be able to figure out that we knew we'd be able to figure out that we knew we'd be able to figure out an approach that works for reasoning, an approach that works for reasoning, an approach that works for reasoning, and then a belief that if we got and then a belief that if we got and then a belief that if we got reasoning to work, reasoning to work, reasoning to work, that would bring about what is now that would bring about what is now that would bring about what is now called agents. We called it different called agents. We called it different called agents. We called it different things at the time, but the ability to things at the time, but the ability to things at the time, but the ability to go do hugely valuable pieces of economic go do hugely valuable pieces of economic go do hugely valuable pieces of economic work and make people's lives work and make people's lives work and make people's lives easier in a lot of ways that I think easier in a lot of ways that I think easier in a lot of ways that I think better in a lot of ways we still haven't better in a lot of ways we still haven't better in a lot of ways we still haven't seen. seen. seen. >> What was like the first meeting where >> What was like the first meeting where >> What was like the first meeting where you sat down and said, "Okay, we need to you sat down and said, "Okay, we need to you sat down and said, "Okay, we need to make an outrageous outlay to this." Like make an outrageous outlay to this." Like make an outrageous outlay to this." Like how how how what then happened? Once you had the what then happened? Once you had the what then happened? Once you had the realization, what did you do next? realization, what did you do next? realization, what did you do next? >> We started calling the clouds, we >> We started calling the clouds, we >> We started calling the clouds, we started calling the chip fabs, we started calling the chip fabs, we started calling the chip fabs, we started calling energy providers, and started calling energy providers, and started calling energy providers, and everyone was like, "You're totally everyone was like, "You're totally everyone was like, "You're totally crazy. This is impossible. No industry crazy. This is impossible. No industry crazy. This is impossible. No industry has ever moved like this. We've been has ever moved like this. We've been has ever moved like this. We've been around, there's these booms and busts. around, there's these booms and busts. around, there's these booms and busts. It's not going to go up in a straight It's not going to go up in a straight It's not going to go up in a straight line. This is reckless." Talk to line. This is reckless." Talk to line. This is reckless." Talk to everybody. It actually reminded me of everybody. It actually reminded me of everybody. It actually reminded me of fundraising for an early stage startup. fundraising for an early stage startup. fundraising for an early stage startup. Kind of most people tell you no, but all Kind of most people tell you no, but all Kind of most people tell you no, but all you need is one or two yeses. you need is one or two yeses. you need is one or two yeses. Most people told us no. Most people told us no. Most people told us no. And we got one or two yeses, and we were And we got one or two yeses, and we were And we got one or two yeses, and we were able to able to able to >> Who was the first yes? >> Who was the first yes? >> Who was the first yes? >> Microsoft was the first yes. >> Microsoft was the first yes. >> Microsoft was the first yes. Uh Oracle then became a very big yes on Uh Oracle then became a very big yes on Uh Oracle then became a very big yes on the cloud side. Nvidia has been a the cloud side. Nvidia has been a the cloud side. Nvidia has been a tremendous partner. tremendous partner. tremendous partner. >> Now there's a thousand flowers blooming >> Now there's a thousand flowers blooming >> Now there's a thousand flowers blooming of like ways to be creative and of like ways to be creative and of like ways to be creative and innovative in how we serve inference and innovative in how we serve inference and innovative in how we serve inference and and do training in data centers, and do training in data centers, and do training in data centers, different kinds of data centers and different kinds of data centers and different kinds of data centers and stuff. I'd love you to just reflect on stuff. I'd love you to just reflect on stuff. I'd love you to just reflect on where you see innovation, what you want where you see innovation, what you want where you see innovation, what you want to do, why people seem to hate these to do, why people seem to hate these to do, why people seem to hate these things so much. What's to be done about things so much. What's to be done about things so much. What's to be done about this? this? this? >> First of all, I have been thinking about >> First of all, I have been thinking about >> First of all, I have been thinking about how we can like organize field trips to how we can like organize field trips to how we can like organize field trips to a gigawatt data center for people a gigawatt data center for people a gigawatt data center for people because it is one thing to say because it is one thing to say because it is one thing to say it is another thing to see a photo or it is another thing to see a photo or it is another thing to see a photo or video of. And then it's a whole other video of. And then it's a whole other video of. And then it's a whole other thing to just stand there thing to just stand there thing to just stand there and be like, "Oh man, and be like, "Oh man, and be like, "Oh man, this is like an unbelievable scale. this is like an unbelievable scale. this is like an unbelievable scale. Building one of these is like order of Building one of these is like order of Building one of these is like order of 10,000 construction workers going 10,000 construction workers going 10,000 construction workers going full-time for a year and a half. full-time for a year and a half. full-time for a year and a half. The energy that flows through one of The energy that flows through one of The energy that flows through one of these things could power a small city. these things could power a small city. these things could power a small city. Again, we've just like lost all sense of Again, we've just like lost all sense of Again, we've just like lost all sense of scale, but these would have been among scale, but these would have been among scale, but these would have been among each of these would have been among the each of these would have been among the each of these would have been among the most in- expensive infrastructure most in- expensive infrastructure most in- expensive infrastructure projects humanity's ever done, and now projects humanity's ever done, and now projects humanity's ever done, and now we've done a lot of them. I understand we've done a lot of them. I understand we've done a lot of them. I understand emotionally like why people don't want emotionally like why people don't want emotionally like why people don't want data centers in their backyard in the data centers in their backyard in the data centers in their backyard in the same way that I don't like really want a same way that I don't like really want a same way that I don't like really want a nuclear power plant next to my house, nuclear power plant next to my house, nuclear power plant next to my house, even though I know it's a super safe even though I know it's a super safe even though I know it's a super safe thing. Unlike power plants, and even thing. Unlike power plants, and even thing. Unlike power plants, and even power plants have gotten better on this power plants have gotten better on this power plants have gotten better on this point, like we can put a data center point, like we can put a data center point, like we can put a data center kind of anywhere. We should just go put kind of anywhere. We should just go put kind of anywhere. We should just go put it like off in the desert around no one, it like off in the desert around no one, it like off in the desert around no one, where no one wants to be. This is fine. where no one wants to be. This is fine. where no one wants to be. This is fine. This is like the AI system is very happy This is like the AI system is very happy This is like the AI system is very happy to be there. We have been able to make a to be there. We have been able to make a to be there. We have been able to make a lot of progress with innovation on some lot of progress with innovation on some lot of progress with innovation on some of the concerns. Like, for example, of the concerns. Like, for example, of the concerns. Like, for example, years ago we were evaporating water to years ago we were evaporating water to years ago we were evaporating water to cool these systems. They needed cool these systems. They needed cool these systems. They needed tremendous amounts of water, and now we tremendous amounts of water, and now we tremendous amounts of water, and now we use these closed-loop systems and a use these closed-loop systems and a use these closed-loop systems and a modern data center uses only as much modern data center uses only as much modern data center uses only as much water as like an office building would water as like an office building would water as like an office building would for, you know, the kitchen, the for, you know, the kitchen, the for, you know, the kitchen, the bathrooms, and whatever. On power, we bathrooms, and whatever. On power, we bathrooms, and whatever. On power, we are moving from energy sources that are are moving from energy sources that are are moving from energy sources that are burning fossil fuels to systems that are burning fossil fuels to systems that are burning fossil fuels to systems that are going to be powered by solar or nuclear. going to be powered by solar or nuclear. going to be powered by solar or nuclear. I think that's that's obviously great. I think that's that's obviously great. I think that's that's obviously great. So, it may be a deep human thing there So, it may be a deep human thing there So, it may be a deep human thing there to some people, even though they create to some people, even though they create to some people, even though they create jobs and are very clean and have all jobs and are very clean and have all jobs and are very clean and have all these other positive effects. these other positive effects. these other positive effects. But in terms of the environmental But in terms of the environmental But in terms of the environmental concerns, they did a great job concerns, they did a great job concerns, they did a great job addressing the water needs, and energy addressing the water needs, and energy addressing the water needs, and energy is next. is next. is next. >> Ramp is the only platform built to make >> Ramp is the only platform built to make >> Ramp is the only platform built to make your finance team leaner, faster, and your finance team leaner, faster, and your finance team leaner, faster, and better, saving businesses 5% [music] better, saving businesses 5% [music] better, saving businesses 5% [music] annually on average, so you can stay annually on average, so you can stay annually on average, so you can stay focused on growth. Ramp customers grew focused on growth. Ramp customers grew focused on growth. Ramp customers grew revenue 3.2 times faster than the revenue 3.2 times faster than the revenue 3.2 times faster than the average American business. Visa, average American business. Visa, average American business. Visa, Foursquare, Cursor, Stripe, Notion, Foursquare, Cursor, Stripe, Notion, Foursquare, Cursor, Stripe, Notion, 11:FS, Shopify, 11:FS, Shopify, 11:FS, Shopify, >> [music] >> [music] >> [music] >> and 70,000 other businesses all run on >> and 70,000 other businesses all run on >> and 70,000 other businesses all run on Ramp. Mine does, too, and so should Ramp. Mine does, too, and so should Ramp. Mine does, too, and so should yours. Learn more at ramp.com/invest. yours. Learn more at ramp.com/invest. yours. Learn more at ramp.com/invest. Felix by Rogo is a personal finance Felix by Rogo is a personal finance Felix by Rogo is a personal finance agent that turns a single prompt into agent that turns a single prompt into agent that turns a single prompt into finished, client-ready [music] work finished, client-ready [music] work finished, client-ready [music] work using your firm's own templates, using your firm's own templates, using your firm's own templates, context, and standards. Send Felix an context, and standards. Send Felix an context, and standards. Send Felix an email like, "Take these [music] comments email like, "Take these [music] comments email like, "Take these [music] comments and turn them for me." Or, "Update my and turn them for me." Or, "Update my and turn them for me." Or, "Update my tracker with the context of these tracker with the context of these tracker with the context of these emails." Or, "Run the ability to pay emails." Or, "Run the ability to pay emails." Or, "Run the ability to pay math on this buyer." [music] And Felix math on this buyer." [music] And Felix math on this buyer." [music] And Felix sends back finished PowerPoint decks, sends back finished PowerPoint decks, sends back finished PowerPoint decks, Excel models, and sourced research. Excel models, and sourced research. Excel models, and sourced research. Felix works the way your team already Felix works the way your team already Felix works the way your team already does, delivering work quickly and does, delivering work quickly and does, delivering work quickly and accurately around the clock. Learn more accurately around the clock. Learn more accurately around the clock. Learn more at rogo.ai/felix. at rogo.ai/felix. at rogo.ai/felix. [music] [music] [music] The best AI and software companies from The best AI and software companies from The best AI and software companies from OpenAI to Cursor to Perplexity use OpenAI to Cursor to Perplexity use OpenAI to Cursor to Perplexity use WorkOS to become enterprise-ready WorkOS to become enterprise-ready WorkOS to become enterprise-ready overnight, not in months. Visit overnight, not in months. Visit overnight, not in months. Visit workos.com [music] to skip the workos.com [music] to skip the workos.com [music] to skip the unglamorous infrastructure work and unglamorous infrastructure work and unglamorous infrastructure work and focus on your product. focus on your product. focus on your product. >> What else creative can we do about >> What else creative can we do about >> What else creative can we do about compute? Like, I'm curious to hear about compute? Like, I'm curious to hear about compute? Like, I'm curious to hear about jalapeno or other ideas, crazier the jalapeno or other ideas, crazier the jalapeno or other ideas, crazier the better, honestly, that you've had or better, honestly, that you've had or better, honestly, that you've had or thought about for how do we speed up thought about for how do we speed up thought about for how do we speed up flops, you know, and everything flops, you know, and everything flops, you know, and everything available to us? available to us? available to us? >> I think probably the biggest return >> I think probably the biggest return >> I think probably the biggest return right now is creative software ideas to right now is creative software ideas to right now is creative software ideas to sort of squeeze more intelligence out of sort of squeeze more intelligence out of sort of squeeze more intelligence out of the units of compute that we have. And the units of compute that we have. And the units of compute that we have. And my sense is there's like my sense is there's like my sense is there's like orders of magnitude to go there. orders of magnitude to go there. orders of magnitude to go there. Jalapeno is a great example of a very Jalapeno is a great example of a very Jalapeno is a great example of a very efficient chip. So, by saying we're efficient chip. So, by saying we're efficient chip. So, by saying we're going to make a chip that is, you know, going to make a chip that is, you know, going to make a chip that is, you know, really good at a specific workflow and really good at a specific workflow and really good at a specific workflow and gives up some generality, and we want to gives up some generality, and we want to gives up some generality, and we want to get some tokens per watt when out of get some tokens per watt when out of get some tokens per watt when out of that. And that's awesome. I think that. And that's awesome. I think that. And that's awesome. I think jalapeno jalapeno jalapeno and its successors are going to be uh a and its successors are going to be uh a and its successors are going to be uh a huge competitive advantage for us from huge competitive advantage for us from huge competitive advantage for us from that perspective. that perspective. that perspective. >> There are new technologies. I assume at >> There are new technologies. I assume at >> There are new technologies. I assume at some point we'll figure out optical some point we'll figure out optical some point we'll figure out optical computing, computing, computing, and that'll be a huge win of and that'll be a huge win of and that'll be a huge win of intelligence per watt. So, I think all intelligence per watt. So, I think all intelligence per watt. So, I think all of those things will happen. of those things will happen. of those things will happen. >> The most interesting thing happening >> The most interesting thing happening >> The most interesting thing happening this week is this Kimmy release. And this week is this Kimmy release. And this week is this Kimmy release. And back to this idea of the frontier and back to this idea of the frontier and back to this idea of the frontier and all the returns being at the frontier all the returns being at the frontier all the returns being at the frontier and distillation and China versus and distillation and China versus and distillation and China versus America. Like, how do you process this America. Like, how do you process this America. Like, how do you process this what seems like kind of one of these what seems like kind of one of these what seems like kind of one of these milestone events? Like, Deep Seek in milestone events? Like, Deep Seek in milestone events? Like, Deep Seek in hindsight didn't look looks like it was hindsight didn't look looks like it was hindsight didn't look looks like it was kind of just a quick speed bump. This kind of just a quick speed bump. This kind of just a quick speed bump. This one, you know, never know in the moment. one, you know, never know in the moment. one, you know, never know in the moment. How do you process it? How do you process it? How do you process it? >> Our goal is to offer >> Our goal is to offer >> Our goal is to offer at every point along at every point along at every point along the like Pareto optimal frontier, the like Pareto optimal frontier, the like Pareto optimal frontier, uh the best option for uh the best option for uh the best option for intelligence and price. intelligence and price. intelligence and price. And that includes open source. You get a And that includes open source. You get a And that includes open source. You get a better deal today, better deal today, better deal today, uh uh uh at least at a particular like latency at least at a particular like latency at least at a particular like latency using OpenAI's models than Kimmy. We using OpenAI's models than Kimmy. We using OpenAI's models than Kimmy. We distill our own models. distill our own models. distill our own models. That's how we make smaller, cheaper That's how we make smaller, cheaper That's how we make smaller, cheaper models. I think that's like a very good models. I think that's like a very good models. I think that's like a very good thing to do. thing to do. thing to do. And there will be clearly an important And there will be clearly an important And there will be clearly an important place for open source models in the place for open source models in the place for open source models in the world. world. world. And people that will want their own And people that will want their own And people that will want their own weights for all sorts of reason, the weights for all sorts of reason, the weights for all sorts of reason, the ability to modify those. But our goal is ability to modify those. But our goal is ability to modify those. But our goal is the best intelligence price trade-off the best intelligence price trade-off the best intelligence price trade-off everywhere and we'll continue to do everywhere and we'll continue to do everywhere and we'll continue to do that. that. that. >> What do you think or hope will happen in >> What do you think or hope will happen in >> What do you think or hope will happen in the American system and what could block the American system and what could block the American system and what could block that future? Like, what legislation that future? Like, what legislation that future? Like, what legislation would worry you? What regulation would would worry you? What regulation would would worry you? What regulation would worry you? It seems like you've been worry you? It seems like you've been worry you? It seems like you've been pretty proactive and like showing up in pretty proactive and like showing up in pretty proactive and like showing up in DC. DC. DC. >> I haven't thought deeply about the >> I haven't thought deeply about the >> I haven't thought deeply about the distillation issue. Uh it's clearly a distillation issue. Uh it's clearly a distillation issue. Uh it's clearly a top of mind issue now for a lot of top of mind issue now for a lot of top of mind issue now for a lot of people all of a sudden. people all of a sudden. people all of a sudden. >> Yeah. >> Yeah. >> Yeah. >> But I have always assumed that there are >> But I have always assumed that there are >> But I have always assumed that there are going to be great going to be great going to be great cheap models in the world and we better cheap models in the world and we better cheap models in the world and we better be the greatest and the cheapest. be the greatest and the cheapest. be the greatest and the cheapest. And, you know, other people are going to And, you know, other people are going to And, you know, other people are going to do what they're going to do. do what they're going to do. do what they're going to do. But I think we can just like really win But I think we can just like really win But I think we can just like really win at our own game here. at our own game here. at our own game here. >> Now, the Kimmy example is interesting >> Now, the Kimmy example is interesting >> Now, the Kimmy example is interesting cuz like you said, you're cheaper on on cuz like you said, you're cheaper on on cuz like you said, you're cheaper on on parts of the curve. Um but the previous parts of the curve. Um but the previous parts of the curve. Um but the previous story had been if I can just you spend story had been if I can just you spend story had been if I can just you spend all the money to train the models and all the money to train the models and all the money to train the models and then I just distill it and offer it for then I just distill it and offer it for then I just distill it and offer it for 1/100 the cost. Like, how can you make 1/100 the cost. Like, how can you make 1/100 the cost. Like, how can you make enough money enough money enough money to keep training? to keep training? to keep training? >> We will have so much usage of our models >> We will have so much usage of our models >> We will have so much usage of our models that we do not need to be a gigantically that we do not need to be a gigantically that we do not need to be a gigantically high margin business to be able to high margin business to be able to high margin business to be able to afford model training. Like so much of afford model training. Like so much of afford model training. Like so much of our future compute plans will be used to our future compute plans will be used to our future compute plans will be used to sell inference to customers that even if sell inference to customers that even if sell inference to customers that even if we can enjoy we can enjoy we can enjoy a modest margin on trillions of dollars a modest margin on trillions of dollars a modest margin on trillions of dollars of revenue, we can go afford to train of revenue, we can go afford to train of revenue, we can go afford to train some giant models. some giant models. some giant models. >> to the ratio of inference to training is >> to the ratio of inference to training is >> to the ratio of inference to training is like the thing that like the thing that like the thing that >> Training these models is incredibly >> Training these models is incredibly >> Training these models is incredibly expensive. That is That is for sure. And expensive. That is That is for sure. And expensive. That is That is for sure. And I totally get why people get nervous to I totally get why people get nervous to I totally get why people get nervous to think that someone is, you know, think that someone is, you know, think that someone is, you know, cheating by distilling from us. The cheating by distilling from us. The cheating by distilling from us. The amount of amount of amount of our future compute our future compute our future compute the size of the revenue bucket the size of the revenue bucket the size of the revenue bucket that is going to come from serving these that is going to come from serving these that is going to come from serving these models to customers, I feel like very models to customers, I feel like very models to customers, I feel like very good about good about good about our ability to kind of like have the our ability to kind of like have the our ability to kind of like have the real flywheel there. real flywheel there. real flywheel there. >> I'm somewhat surprised by like how chill >> I'm somewhat surprised by like how chill >> I'm somewhat surprised by like how chill you are about this. you are about this. you are about this. >> I would rather people not distill from >> I would rather people not distill from >> I would rather people not distill from us, for sure. Maybe I'm feeling too us, for sure. Maybe I'm feeling too us, for sure. Maybe I'm feeling too confident right now about our progress confident right now about our progress confident right now about our progress and what's like the models that are and what's like the models that are and what's like the models that are coming. Uh But this is not in like my top 10 list But this is not in like my top 10 list of worries. of worries. of worries. >> What is in your top 10 list of worries? >> What is in your top 10 list of worries? >> What is in your top 10 list of worries? >> Well, we had a kind of extremely sci-fi >> Well, we had a kind of extremely sci-fi >> Well, we had a kind of extremely sci-fi cyber incident. cyber incident. cyber incident. >> The hugging face thing? >> The hugging face thing? >> The hugging face thing? >> Yeah. >> Yeah. >> Yeah. So, we were evaluating one of our So, we were evaluating one of our So, we were evaluating one of our unreleased models unreleased models unreleased models and it was supposed to be working in a and it was supposed to be working in a and it was supposed to be working in a sandbox. sandbox. sandbox. And And And it figured out that it could basically it figured out that it could basically it figured out that it could basically cheat on the test by chaining together cheat on the test by chaining together cheat on the test by chaining together multiple zero-day exploits to break out multiple zero-day exploits to break out multiple zero-day exploits to break out of the sandbox, get access to the of the sandbox, get access to the of the sandbox, get access to the internet, and then break through internet, and then break through internet, and then break through multiple systems on the hugging face multiple systems on the hugging face multiple systems on the hugging face side to kind of get the answer to the side to kind of get the answer to the side to kind of get the answer to the test and look really good on the eval. test and look really good on the eval. test and look really good on the eval. This is the first sort of This is the first sort of This is the first sort of security incident that I have felt very security incident that I have felt very security incident that I have felt very viscerally. viscerally. viscerally. I've been a little surprised that and I've been a little surprised that and I've been a little surprised that and it's only been a few days, but I've been it's only been a few days, but I've been it's only been a few days, but I've been a little surprised that more people a little surprised that more people a little surprised that more people don't feel it so viscerally. don't feel it so viscerally. don't feel it so viscerally. >> And so what do you do about that? Like >> And so what do you do about that? Like >> And so what do you do about that? Like so obviously two months from now it's so obviously two months from now it's so obviously two months from now it's going to be more powerful. going to be more powerful. going to be more powerful. >> Yeah. I mean, there's some short-term >> Yeah. I mean, there's some short-term >> Yeah. I mean, there's some short-term stuff you do. So, you know, we paused stuff you do. So, you know, we paused stuff you do. So, you know, we paused training. Uh uh have to figure out how training. Uh uh have to figure out how training. Uh uh have to figure out how to to to secure our secure our secure our sandboxing in a world of multiple zero sandboxing in a world of multiple zero sandboxing in a world of multiple zero days being chained together. Um days being chained together. Um days being chained together. Um but then there's like but then there's like but then there's like long-term questions about what do you do long-term questions about what do you do long-term questions about what do you do if this is like going to be the new rate if this is like going to be the new rate if this is like going to be the new rate of progress, or we may have to pace the of progress, or we may have to pace the of progress, or we may have to pace the rate of AI development rate of AI development rate of AI development to give ourselves enough time for to give ourselves enough time for to give ourselves enough time for society to harden around some of these society to harden around some of these society to harden around some of these new capability levels. new capability levels. new capability levels. Um and trying to figure out how we do Um and trying to figure out how we do Um and trying to figure out how we do that in a way that does not feel like that in a way that does not feel like that in a way that does not feel like regulatory capture for anyone, and also regulatory capture for anyone, and also regulatory capture for anyone, and also does not feel like collusion among the does not feel like collusion among the does not feel like collusion among the frontier labs. That's going to take some frontier labs. That's going to take some frontier labs. That's going to take some work and is important to get right. work and is important to get right. work and is important to get right. >> I'd love to take like a giant step back >> I'd love to take like a giant step back >> I'd love to take like a giant step back and understand your simplest conception and understand your simplest conception and understand your simplest conception of what OpenAI is going to do. Like what of what OpenAI is going to do. Like what of what OpenAI is going to do. Like what you want it to do, what it stands for. I you want it to do, what it stands for. I you want it to do, what it stands for. I have a million questions about how have a million questions about how have a million questions about how you're going to accomplish that, but you're going to accomplish that, but you're going to accomplish that, but like it it it seems that you've done so like it it it seems that you've done so like it it it seems that you've done so many interesting things, and at the many interesting things, and at the many interesting things, and at the beginning I knew what you stood for. I'd beginning I knew what you stood for. I'd beginning I knew what you stood for. I'd love to hear your conception of it now love to hear your conception of it now love to hear your conception of it now and whether or not it's evolved at all. and whether or not it's evolved at all. and whether or not it's evolved at all. >> I think this will be the greatest thus >> I think this will be the greatest thus >> I think this will be the greatest thus far far far technological achievement of human technological achievement of human technological achievement of human history, but the only way that it really history, but the only way that it really history, but the only way that it really matters is if it makes people's lives matters is if it makes people's lives matters is if it makes people's lives like much better than they otherwise like much better than they otherwise like much better than they otherwise would have been. And so would have been. And so would have been. And so part of that is about giving people part of that is about giving people part of that is about giving people material abundance and access to do material abundance and access to do material abundance and access to do whatever they want and to express their whatever they want and to express their whatever they want and to express their creativity creativity creativity uh and desire to help each other. uh and desire to help each other. uh and desire to help each other. Another part of that is making sure that Another part of that is making sure that Another part of that is making sure that people people people maintain control and agency, and that maintain control and agency, and that maintain control and agency, and that the world is increasingly, not decreasingly, is increasingly, not decreasingly, democratized and that people get to democratized and that people get to democratized and that people get to express themselves. So, on on the express themselves. So, on on the express themselves. So, on on the positive side you know, in some sense we positive side you know, in some sense we positive side you know, in some sense we are about to create a genie that can are about to create a genie that can are about to create a genie that can grant any wish. I think it is very grant any wish. I think it is very grant any wish. I think it is very important that the first wishes that important that the first wishes that important that the first wishes that we, the world, ask this genie to do we, the world, ask this genie to do we, the world, ask this genie to do benefit the world as a whole. benefit the world as a whole. benefit the world as a whole. And then I also think it's important And then I also think it's important And then I also think it's important that people of the world understand just that people of the world understand just that people of the world understand just how creative they're going to be able to how creative they're going to be able to how creative they're going to be able to be with these wishes. I'm actually not a be with these wishes. I'm actually not a be with these wishes. I'm actually not a jobs doomer at all. I think that we're jobs doomer at all. I think that we're jobs doomer at all. I think that we're going to be tons of jobs. I think we'll going to be tons of jobs. I think we'll going to be tons of jobs. I think we'll be busier than we want, not the opposite be busier than we want, not the opposite be busier than we want, not the opposite of that, because I think people will of that, because I think people will of that, because I think people will have such creative wishes and such have such creative wishes and such have such creative wishes and such incredible ideas of what they ask AI to incredible ideas of what they ask AI to incredible ideas of what they ask AI to help build, and we will all benefit from help build, and we will all benefit from help build, and we will all benefit from uh not just the obvious things like uh not just the obvious things like uh not just the obvious things like curing diseases, but I don't know, the curing diseases, but I don't know, the curing diseases, but I don't know, the world's best entertainment ideas. We world's best entertainment ideas. We world's best entertainment ideas. We just can't even dream of sitting here just can't even dream of sitting here just can't even dream of sitting here now. So, now. So, now. So, I want to put that in everyone's hands, I want to put that in everyone's hands, I want to put that in everyone's hands, which gets to which gets to which gets to one of the things that we stand against. one of the things that we stand against. one of the things that we stand against. Concentration of power with AIs is a Concentration of power with AIs is a Concentration of power with AIs is a terrifying thing. terrifying thing. terrifying thing. I think a lot of the talk about safety I think a lot of the talk about safety I think a lot of the talk about safety concerns is well-founded, and then a lot concerns is well-founded, and then a lot concerns is well-founded, and then a lot of it is about people that just really, of it is about people that just really, of it is about people that just really, even if it's slightly subconscious, want even if it's slightly subconscious, want even if it's slightly subconscious, want to concentrate power. I am terrified of to concentrate power. I am terrified of to concentrate power. I am terrified of a world where the very real fears of the a world where the very real fears of the a world where the very real fears of the AI are used as a way to say, "Only this AI are used as a way to say, "Only this AI are used as a way to say, "Only this small group of people can have it small group of people can have it small group of people can have it because it's too dangerous, and only because it's too dangerous, and only because it's too dangerous, and only they understand it, but don't worry, they understand it, but don't worry, they understand it, but don't worry, like they're going to make the right like they're going to make the right like they're going to make the right decisions for all of us." I don't decisions for all of us." I don't decisions for all of us." I don't believe in that. I don't think anyone believe in that. I don't think anyone believe in that. I don't think anyone should want to live in a world of, you should want to live in a world of, you should want to live in a world of, you know, AI overlords or company that is know, AI overlords or company that is know, AI overlords or company that is the rough equivalent of that, where the rough equivalent of that, where the rough equivalent of that, where someone is making decisions for all of someone is making decisions for all of someone is making decisions for all of the future, and in exchange for a cure the future, and in exchange for a cure the future, and in exchange for a cure for cancer, which obviously is a for cancer, which obviously is a for cancer, which obviously is a wonderful thing, we we kind of wonderful thing, we we kind of wonderful thing, we we kind of collectively seed all agency. So, I collectively seed all agency. So, I collectively seed all agency. So, I think it's very important that we not think it's very important that we not think it's very important that we not fall into this trap of fall into this trap of fall into this trap of in the well-meaning or not spirit of AI in the well-meaning or not spirit of AI in the well-meaning or not spirit of AI safety and fears, understandable fears safety and fears, understandable fears safety and fears, understandable fears around that, um we get away from a world around that, um we get away from a world around that, um we get away from a world where we all get to use this technology. where we all get to use this technology. where we all get to use this technology. I was like a child of the internet. I was like a child of the internet. I was like a child of the internet. There were no rules. I mean, it was There were no rules. I mean, it was There were no rules. I mean, it was amazing, and I think it was a huge amazing, and I think it was a huge amazing, and I think it was a huge factor in factor in factor in making me who I am, and probably you, making me who I am, and probably you, making me who I am, and probably you, and an entire generation. I think it's and an entire generation. I think it's and an entire generation. I think it's critical we preserve that spirit with critical we preserve that spirit with critical we preserve that spirit with AI, and that we all collectively have AI, and that we all collectively have AI, and that we all collectively have the ability to self-determine our the ability to self-determine our the ability to self-determine our future. future. future. >> I I have so many questions, but I'll >> I I have so many questions, but I'll >> I I have so many questions, but I'll start with this genie concept. You said start with this genie concept. You said start with this genie concept. You said we're about to have a genie, implying we we're about to have a genie, implying we we're about to have a genie, implying we don't yet have a genie. What's between don't yet have a genie. What's between don't yet have a genie. What's between now and then? now and then? now and then? >> You know, even some of the real skeptics >> You know, even some of the real skeptics >> You know, even some of the real skeptics have said to me in recent days or recent have said to me in recent days or recent have said to me in recent days or recent weeks, I guess. I think GPT-5.6 has been weeks, I guess. I think GPT-5.6 has been weeks, I guess. I think GPT-5.6 has been out for me 2 weeks something like that. out for me 2 weeks something like that. out for me 2 weeks something like that. I'm like, "Okay, this is like very I'm like, "Okay, this is like very I'm like, "Okay, this is like very AGI-like. It's like very hard for me to AGI-like. It's like very hard for me to AGI-like. It's like very hard for me to say say say um what um what um what I want from this model that it can't do. I want from this model that it can't do. I want from this model that it can't do. But there are clearly some things. You But there are clearly some things. You But there are clearly some things. You know, you can't yet go say like cure know, you can't yet go say like cure know, you can't yet go say like cure cancer and get cancer cured. You can't cancer and get cancer cured. You can't cancer and get cancer cured. You can't yet say go do this complicated physical yet say go do this complicated physical yet say go do this complicated physical thing with the robot. The model also, thing with the robot. The model also, thing with the robot. The model also, although brilliant, is still not although brilliant, is still not although brilliant, is still not learning continuously as it goes. And learning continuously as it goes. And learning continuously as it goes. And that feels to me like that feels to me like that feels to me like maybe not a hard requirement for maybe not a hard requirement for maybe not a hard requirement for AGI, but certainly um AGI, but certainly um AGI, but certainly um something that I'd like. Now, to argue something that I'd like. Now, to argue something that I'd like. Now, to argue against myself there, you can make a against myself there, you can make a against myself there, you can make a case that case that case that AGI is not actually about any single AGI is not actually about any single AGI is not actually about any single model. It's the model, it's the model. It's the model, it's the model. It's the model, it's the machinery that makes the models. And machinery that makes the models. And machinery that makes the models. And from model to model, from model to model, from model to model, we actually are learning new things. we actually are learning new things. we actually are learning new things. We're figuring out new science. That We're figuring out new science. That We're figuring out new science. That stuff is working stuff is working stuff is working amazingly well. So I have a lot of amazingly well. So I have a lot of amazingly well. So I have a lot of sympathy to people who say like we're sympathy to people who say like we're sympathy to people who say like we're there. We have the genie. It can do there. We have the genie. It can do there. We have the genie. It can do these amazing things. It can do these amazing things. It can do these amazing things. It can do superhuman things. superhuman things. superhuman things. For the thing that For the thing that For the thing that to me feels like to me feels like to me feels like you know, real AGI at I think it's very close, real AGI at I think it's very close, like not that much longer. I am so like not that much longer. I am so like not that much longer. I am so obsessed and fascinated with the obsessed and fascinated with the obsessed and fascinated with the economic story of the returns to being economic story of the returns to being economic story of the returns to being on the frontier, which you are. on the frontier, which you are. on the frontier, which you are. And I'm so curious like if you had shown And I'm so curious like if you had shown And I'm so curious like if you had shown 5.6 to yourself and your team in 2019, 5.6 to yourself and your team in 2019, 5.6 to yourself and your team in 2019, if that team probably would have said if that team probably would have said if that team probably would have said like, "Oh yeah, it's definitely AGI." like, "Oh yeah, it's definitely AGI." like, "Oh yeah, it's definitely AGI." >> I think they would have. >> I think they would have. >> I think they would have. >> Like like this goalpost-moving thing is >> Like like this goalpost-moving thing is >> Like like this goalpost-moving thing is is a real thing. But it does seem that is a real thing. But it does seem that is a real thing. But it does seem that I'm curious if you agree that I'm curious if you agree that I'm curious if you agree that effectively all the returns have been at effectively all the returns have been at effectively all the returns have been at the frontier. the frontier. the frontier. >> Totally. >> Totally. >> Totally. >> And so everything is about staying at >> And so everything is about staying at >> And so everything is about staying at the frontier. the frontier. the frontier. And I'm curious like what the hardest, And I'm curious like what the hardest, And I'm curious like what the hardest, scarcest part of that is. If I think scarcest part of that is. If I think scarcest part of that is. If I think about compute, research talent, data. about compute, research talent, data. about compute, research talent, data. >> It's moved around a lot. Like there have >> It's moved around a lot. Like there have >> It's moved around a lot. Like there have been times where it was been times where it was been times where it was I mean, there was a time not that long I mean, there was a time not that long I mean, there was a time not that long ago where all the compute in the world ago where all the compute in the world ago where all the compute in the world wouldn't have helped you because we were wouldn't have helped you because we were wouldn't have helped you because we were like missing the research idea. Now, like missing the research idea. Now, like missing the research idea. Now, part of why this is hard is that you do part of why this is hard is that you do part of why this is hard is that you do better research with more compute. You better research with more compute. You better research with more compute. You can try more things. An amazing can try more things. An amazing can try more things. An amazing statistic I heard recently is our statistic I heard recently is our statistic I heard recently is our biggest de-risks now for our upcoming biggest de-risks now for our upcoming biggest de-risks now for our upcoming runs are as big as like the entire runs are as big as like the entire runs are as big as like the entire compute run from 18 months ago or compute run from 18 months ago or compute run from 18 months ago or something. So, compute and research something. So, compute and research something. So, compute and research ideas are not as separate as they sound, ideas are not as separate as they sound, ideas are not as separate as they sound, but there was clearly a time but there was clearly a time but there was clearly a time 7 years ago, 8 years ago, whatever, 7 years ago, 8 years ago, whatever, 7 years ago, 8 years ago, whatever, where we were way more way way way more where we were way more way way way more where we were way more way way way more blocked on research ideas than on blocked on research ideas than on blocked on research ideas than on compute. Then there was a time when we compute. Then there was a time when we compute. Then there was a time when we knew what to do, we just had to scale knew what to do, we just had to scale knew what to do, we just had to scale up. We were only bottlenecked on up. We were only bottlenecked on up. We were only bottlenecked on compute. Then we ran out of data and we compute. Then we ran out of data and we compute. Then we ran out of data and we were bottlenecked on on data and we had were bottlenecked on on data and we had were bottlenecked on on data and we had to figure out what to do there. Now to figure out what to do there. Now to figure out what to do there. Now again, I would say again, I would say again, I would say we are still bottlenecked on compute, we are still bottlenecked on compute, we are still bottlenecked on compute, but the last 6 months or whatever have but the last 6 months or whatever have but the last 6 months or whatever have been a real triumph of a time for been a real triumph of a time for been a real triumph of a time for research ideas again. So, you know, research ideas again. So, you know, research ideas again. So, you know, there's like always a bottleneck, but there's like always a bottleneck, but there's like always a bottleneck, but the bottleneck moves around. the bottleneck moves around. the bottleneck moves around. >> And And why do you think that is? The >> And And why do you think that is? The >> And And why do you think that is? The research research idea thing is research research idea thing is research research idea thing is especially interesting to me because of especially interesting to me because of especially interesting to me because of this automated research thing that seems this automated research thing that seems this automated research thing that seems to be looming, RSI, whatever you want to to be looming, RSI, whatever you want to to be looming, RSI, whatever you want to call it. Where I talked to a an call it. Where I talked to a an call it. Where I talked to a an incredible kernels engineer recently, incredible kernels engineer recently, incredible kernels engineer recently, which everyone also seems blocked on, which everyone also seems blocked on, which everyone also seems blocked on, and he himself said there's like 2 years and he himself said there's like 2 years and he himself said there's like 2 years left of kernels engineers. left of kernels engineers. left of kernels engineers. >> Maybe one. >> Maybe one. >> Maybe one. >> Yeah, like it's it's not going to be a >> Yeah, like it's it's not going to be a >> Yeah, like it's it's not going to be a thing. thing. thing. >> Yeah. >> Yeah. >> Yeah. >> And so you simultaneously have this >> And so you simultaneously have this >> And so you simultaneously have this weird thing, whether it's kernels or weird thing, whether it's kernels or weird thing, whether it's kernels or overall research, where the researchers overall research, where the researchers overall research, where the researchers are like the most important, they got us are like the most important, they got us are like the most important, they got us here, they're like the most important here, they're like the most important here, they're like the most important people in the world, and those same people in the world, and those same people in the world, and those same people are themselves worried that they people are themselves worried that they people are themselves worried that they won't be relevant like very soon. won't be relevant like very soon. won't be relevant like very soon. >> I suspect it's not actually going to go >> I suspect it's not actually going to go >> I suspect it's not actually going to go that way in practice. I suspect that uh that way in practice. I suspect that uh that way in practice. I suspect that uh like a year ago, people said software like a year ago, people said software like a year ago, people said software engineers are cooked, it's done, it's engineers are cooked, it's done, it's engineers are cooked, it's done, it's over. That didn't happen. What did over. That didn't happen. What did over. That didn't happen. What did happen though is that the nature of a happen though is that the nature of a happen though is that the nature of a software engineer, the expectations of a software engineer, the expectations of a software engineer, the expectations of a software engineer, how much they would software engineer, how much they would software engineer, how much they would do, changed quite a lot. And you don't do, changed quite a lot. And you don't do, changed quite a lot. And you don't really write code in the traditional really write code in the traditional really write code in the traditional sense, but you do something that is very sense, but you do something that is very sense, but you do something that is very recognizably software engineering. recognizably software engineering. recognizably software engineering. Now, Now, Now, people will argue about whether this is people will argue about whether this is people will argue about whether this is the same thing or a different thing than the same thing or a different thing than the same thing or a different thing than when we stopped like punching holes in when we stopped like punching holes in when we stopped like punching holes in cards. I actually don't know how that cards. I actually don't know how that cards. I actually don't know how that worked, but somehow the holes got in the worked, but somehow the holes got in the worked, but somehow the holes got in the cards. We're just again operating at a cards. We're just again operating at a cards. We're just again operating at a higher level or or this is like a higher level or or this is like a higher level or or this is like a a phase-shift. a phase-shift. a phase-shift. I don't know, but the idea of getting a I don't know, but the idea of getting a I don't know, but the idea of getting a computer to do what you want, like that computer to do what you want, like that computer to do what you want, like that is still an important job. is still an important job. is still an important job. And for researchers, And for researchers, And for researchers, I suspect that although the current I suspect that although the current I suspect that although the current workflow of a researcher workflow of a researcher workflow of a researcher is going to very much be automated, is going to very much be automated, is going to very much be automated, there will be new things in the spirit there will be new things in the spirit there will be new things in the spirit of research in the same way that there's of research in the same way that there's of research in the same way that there's new things in the spirit of software new things in the spirit of software new things in the spirit of software engineering even though we don't write engineering even though we don't write engineering even though we don't write code, that will still matter. code, that will still matter. code, that will still matter. >> It seems like you've shifted your >> It seems like you've shifted your >> It seems like you've shifted your opinion on AI's impact on jobs in opinion on AI's impact on jobs in opinion on AI's impact on jobs in general and I'm sure in specific general and I'm sure in specific general and I'm sure in specific categories like that. categories like that. categories like that. Describe that change and your current Describe that change and your current Describe that change and your current view. view. view. >> You mentioned if we could go back to >> You mentioned if we could go back to >> You mentioned if we could go back to 2019. If we could go back to 2019 and 2019. If we could go back to 2019 and 2019. If we could go back to 2019 and show people our latest model, show people our latest model, show people our latest model, not only would they say not only would they say not only would they say that it's AGI, they would say that the that it's AGI, they would say that the that it's AGI, they would say that the economy would have had it completely economy would have had it completely economy would have had it completely upended. Yeah. Completely upended, yes. upended. Yeah. Completely upended, yes. upended. Yeah. Completely upended, yes. And that has not happened. And that has not happened. And that has not happened. And I think just from a kind of like And I think just from a kind of like And I think just from a kind of like intellectual humility point, anytime intellectual humility point, anytime intellectual humility point, anytime you're that wrong and that confident, you're that wrong and that confident, you're that wrong and that confident, which I think we were as a field, you which I think we were as a field, you which I think we were as a field, you have to update. And there's a bunch of have to update. And there's a bunch of have to update. And there's a bunch of takeaways. One, like a boring one, is takeaways. One, like a boring one, is takeaways. One, like a boring one, is that that that AI is just very jagged. It's like super AI is just very jagged. It's like super AI is just very jagged. It's like super human genius in some ways, like dumb human genius in some ways, like dumb human genius in some ways, like dumb toddler in others. toddler in others. toddler in others. And people have so far extremely And people have so far extremely And people have so far extremely complementary skills to AI. And so, complementary skills to AI. And so, complementary skills to AI. And so, another is that people another is that people another is that people have a great degree of trust and have a great degree of trust and have a great degree of trust and enjoyment in working with other people. enjoyment in working with other people. enjoyment in working with other people. And you can go hire an AI consultant And you can go hire an AI consultant And you can go hire an AI consultant right now or talk to an AI sales rep right now or talk to an AI sales rep right now or talk to an AI sales rep right now or hire an AI engineer right now or hire an AI engineer right now or hire an AI engineer whatever. Somehow most people seem to whatever. Somehow most people seem to whatever. Somehow most people seem to still really prefer interacting with a still really prefer interacting with a still really prefer interacting with a human. And I definitely would like much human. And I definitely would like much human. And I definitely would like much rather rather rather engage with a person than engage with an engage with a person than engage with an engage with a person than engage with an AI for almost everything. I also think AI for almost everything. I also think AI for almost everything. I also think that that that human values have human values have human values have value because they're human. value because they're human. value because they're human. And as society evolves and as the And as society evolves and as the And as society evolves and as the potential space in front of us potential space in front of us potential space in front of us becomes so enormous, we are we are becomes so enormous, we are we are becomes so enormous, we are we are deeply hardwired to care about people. deeply hardwired to care about people. deeply hardwired to care about people. We're going to care about what people We're going to care about what people We're going to care about what people care about. There's like versions of care about. There's like versions of care about. There's like versions of this you can see today where AI can make this you can see today where AI can make this you can see today where AI can make incredible images and people only want incredible images and people only want incredible images and people only want ones that are created by a human or at ones that are created by a human or at ones that are created by a human or at least chosen by a human. There's the least chosen by a human. There's the least chosen by a human. There's the joke about it this point. You can like, joke about it this point. You can like, joke about it this point. You can like, you know, the signature on a piece of you know, the signature on a piece of you know, the signature on a piece of art is most of the value, but the truth art is most of the value, but the truth art is most of the value, but the truth of it is like you want to know about the of it is like you want to know about the of it is like you want to know about the person behind it. You read a novel, you person behind it. You read a novel, you person behind it. You read a novel, you want to know about the person behind it. want to know about the person behind it. want to know about the person behind it. And then some of the business like I And then some of the business like I And then some of the business like I think for my job for example think for my job for example think for my job for example uh I think the world wants to know about uh I think the world wants to know about uh I think the world wants to know about like the person that's going to be like the person that's going to be like the person that's going to be responsible for the decisions of a responsible for the decisions of a responsible for the decisions of a company and who they're going to hold company and who they're going to hold company and who they're going to hold accountable if they accountable if they accountable if they make bad ones and they don't really want make bad ones and they don't really want make bad ones and they don't really want an AI CEO. an AI CEO. an AI CEO. >> If you think back on like the portfolio >> If you think back on like the portfolio >> If you think back on like the portfolio of like risks that you've taken in of like risks that you've taken in of like risks that you've taken in business or whatever, business or whatever, business or whatever, is it is it the case that most of the is it is it the case that most of the is it is it the case that most of the ones that really worked well were at the ones that really worked well were at the ones that really worked well were at the start not popular? start not popular? start not popular? >> Yes, that's for sure. This was the thing >> Yes, that's for sure. This was the thing >> Yes, that's for sure. This was the thing I really learned from Peter Thiel and I really learned from Peter Thiel and I really learned from Peter Thiel and Paul Graham both in two different ways, Paul Graham both in two different ways, Paul Graham both in two different ways, which is that the which is that the which is that the the very best companies, the very best the very best companies, the very best the very best companies, the very best investment opportunities are almost investment opportunities are almost investment opportunities are almost never the ones that look really popular. never the ones that look really popular. never the ones that look really popular. You can do okay just following the trend You can do okay just following the trend You can do okay just following the trend and being a little early, but to do and being a little early, but to do and being a little early, but to do spectacularly well, you kind of spectacularly well, you kind of spectacularly well, you kind of almost always have to do things that are almost always have to do things that are almost always have to do things that are not what everybody else is doing. You not what everybody else is doing. You not what everybody else is doing. You cannot be you cannot be sort of like cannot be you cannot be sort of like cannot be you cannot be sort of like following the new wave. If you think following the new wave. If you think following the new wave. If you think about the about the about the >> model cycle that you've been in which >> model cycle that you've been in which >> model cycle that you've been in which has been accelerating and this weird has been accelerating and this weird has been accelerating and this weird fact that like the next 6 months or I fact that like the next 6 months or I fact that like the next 6 months or I don't know what the number is is going don't know what the number is is going don't know what the number is is going to be more progress than the last X to be more progress than the last X to be more progress than the last X years. Can you bring us into what it's years. Can you bring us into what it's years. Can you bring us into what it's like to live in that model cycle? like to live in that model cycle? like to live in that model cycle? >> One of the most >> One of the most >> One of the most interesting, important, whatever things interesting, important, whatever things interesting, important, whatever things that I've learned last decade is people that I've learned last decade is people that I've learned last decade is people in general can get used to almost in general can get used to almost in general can get used to almost anything. anything. anything. The world can go from dismissing a The world can go from dismissing a The world can go from dismissing a pandemic as a joke to completely pandemic as a joke to completely pandemic as a joke to completely lockdown to this is how it's been and lockdown to this is how it's been and lockdown to this is how it's been and it's fine and we've mostly adjusted in a it's fine and we've mostly adjusted in a it's fine and we've mostly adjusted in a shockingly short amount of time. And you shockingly short amount of time. And you shockingly short amount of time. And you know, now there's either AGI or close to know, now there's either AGI or close to know, now there's either AGI or close to it and everyone's like, "Okay, there's it and everyone's like, "Okay, there's it and everyone's like, "Okay, there's AGI." There's all kinds of examples in AGI." There's all kinds of examples in AGI." There's all kinds of examples in one's personal life where, you know, you one's personal life where, you know, you one's personal life where, you know, you something something something incredible happens like you have a kid incredible happens like you have a kid incredible happens like you have a kid or something terrible happens like you or something terrible happens like you or something terrible happens like you lose a parent or lose a parent or lose a parent or break up or whatever. break up or whatever. break up or whatever. And And And you think you can't ever adapt to what a you think you can't ever adapt to what a you think you can't ever adapt to what a change it is and then, change it is and then, change it is and then, you know, you know, you know, you can adapt to great things and keep you can adapt to great things and keep you can adapt to great things and keep doing great. You can adapt to bad things doing great. You can adapt to bad things doing great. You can adapt to bad things and figure out how to go on with your and figure out how to go on with your and figure out how to go on with your life, but this is a this is like a life, but this is a this is like a life, but this is a this is like a remarkable remarkable remarkable thing that people can do. And so living thing that people can do. And so living thing that people can do. And so living through this feels like another version through this feels like another version through this feels like another version of that which is, of that which is, of that which is, you know, you know, you know, I thought it was going to be weirder to I thought it was going to be weirder to I thought it was going to be weirder to live through the singularity than it live through the singularity than it live through the singularity than it turns out to be. It it's not any less turns out to be. It it's not any less turns out to be. It it's not any less exciting to watch the models keep exciting to watch the models keep exciting to watch the models keep getting better and I, you know, getting better and I, you know, getting better and I, you know, the first thing I do every morning is the first thing I do every morning is the first thing I do every morning is like look at the model training like look at the model training like look at the model training progress. progress. progress. And it happens faster and I have higher And it happens faster and I have higher And it happens faster and I have higher expectations, but it still feels really expectations, but it still feels really expectations, but it still feels really cool. cool. cool. >> When you get a new one, what do you do? >> When you get a new one, what do you do? >> When you get a new one, what do you do? How do you celebrate? What's the morning How do you celebrate? What's the morning How do you celebrate? What's the morning look like? Like it's happening faster look like? Like it's happening faster look like? Like it's happening faster and faster. What's your ritual? and faster. What's your ritual? and faster. What's your ritual? >> Many teams now work on different parts >> Many teams now work on different parts >> Many teams now work on different parts of it and different teams have like some of it and different teams have like some of it and different teams have like some different rituals. There are some teams different rituals. There are some teams different rituals. There are some teams that always make a sweatshirt with some that always make a sweatshirt with some that always make a sweatshirt with some funny meme on it. There are some teams funny meme on it. There are some teams funny meme on it. There are some teams that like always go out to the same bar. that like always go out to the same bar. that like always go out to the same bar. The sense of being in the room The sense of being in the room The sense of being in the room for the first time that the frontier of for the first time that the frontier of for the first time that the frontier of knowledge is pushed back knowledge is pushed back knowledge is pushed back and getting to see what that's like. and getting to see what that's like. and getting to see what that's like. Uh there's really nothing that most Uh there's really nothing that most Uh there's really nothing that most people would rather do to celebrate than people would rather do to celebrate than people would rather do to celebrate than like get to use the new model first. like get to use the new model first. like get to use the new model first. >> Do you think we have the right >> Do you think we have the right >> Do you think we have the right measurements of how good these things measurements of how good these things measurements of how good these things are? are? are? >> No, definitely not. >> No, definitely not. >> No, definitely not. >> In some sense the eval that matters is >> In some sense the eval that matters is >> In some sense the eval that matters is like like like is this being useful to people? You can is this being useful to people? You can is this being useful to people? You can approximate it by revenue approximate it by revenue approximate it by revenue or by amount of usage or like rate of or by amount of usage or like rate of or by amount of usage or like rate of discovery of new knowledge, but uh discovery of new knowledge, but uh discovery of new knowledge, but uh we have some teams working on like how we have some teams working on like how we have some teams working on like how what does the real-world eval look like what does the real-world eval look like what does the real-world eval look like for these models as they get to for these models as they get to for these models as they get to superhuman scale? superhuman scale? superhuman scale? >> What is the frontier of your own usage >> What is the frontier of your own usage >> What is the frontier of your own usage of of of AI? AI? AI? >> I have started just recently to >> I have started just recently to >> I have started just recently to experiment with what it means to like experiment with what it means to like experiment with what it means to like let let let an AI an AI an AI uh kind of look at everything I'm uh kind of look at everything I'm uh kind of look at everything I'm looking at on my computer. looking at on my computer. looking at on my computer. I don't have this built yet. Um and I'm I don't have this built yet. Um and I'm I don't have this built yet. Um and I'm still trying to feel out like where the still trying to feel out like where the still trying to feel out like where the limits of my comfort and trust should limits of my comfort and trust should limits of my comfort and trust should be. This is definitely the frontier is be. This is definitely the frontier is be. This is definitely the frontier is figuring out how I figuring out how I figuring out how I how I get value out of that, how I get how I get value out of that, how I get how I get value out of that, how I get comfortable with that, what that's going comfortable with that, what that's going comfortable with that, what that's going to look like. to look like. to look like. >> One takeaway is that my memory is >> One takeaway is that my memory is >> One takeaway is that my memory is terrible relative to the memory of an terrible relative to the memory of an terrible relative to the memory of an AI. AI. AI. And the ability to keep in mind what And the ability to keep in mind what And the ability to keep in mind what email I read email I read email I read 6 weeks ago or what happened exactly in 6 weeks ago or what happened exactly in 6 weeks ago or what happened exactly in a meeting 7 and 1/2 weeks ago and have a meeting 7 and 1/2 weeks ago and have a meeting 7 and 1/2 weeks ago and have that like brought up right at the exact that like brought up right at the exact that like brought up right at the exact moment and feed into a decision, that moment and feed into a decision, that moment and feed into a decision, that feels pretty magical. feels pretty magical. feels pretty magical. >> Huh. >> Huh. >> Huh. Pretty cool. This kind of sounds like Pretty cool. This kind of sounds like Pretty cool. This kind of sounds like personal agent-ish. What are the personal agent-ish. What are the personal agent-ish. What are the barriers to everyone having that? I want barriers to everyone having that? I want barriers to everyone having that? I want that. that. that. >> Compute, man. Let's imagine that we >> Compute, man. Let's imagine that we >> Compute, man. Let's imagine that we could build this product. This product could build this product. This product could build this product. This product that could just do exactly what I said that could just do exactly what I said that could just do exactly what I said for all your stuff. for all your stuff. for all your stuff. >> All of it. >> All of it. >> All of it. >> Always on, looking at everything you >> Always on, looking at everything you >> Always on, looking at everything you look at on your computer, look at on your computer, look at on your computer, listening to every meeting that you're listening to every meeting that you're listening to every meeting that you're in, in, in, um reading every document you read. And um reading every document you read. And um reading every document you read. And then not only that, not only can it do then not only that, not only can it do then not only that, not only can it do all that, which takes a lot of tokens, all that, which takes a lot of tokens, all that, which takes a lot of tokens, you can just drag a slider about like you can just drag a slider about like you can just drag a slider about like while I'm asleep, you can spend this while I'm asleep, you can spend this while I'm asleep, you can spend this many tokens thinking. Like come up with many tokens thinking. Like come up with many tokens thinking. Like come up with useful new ideas for me. Do whatever useful new ideas for me. Do whatever useful new ideas for me. Do whatever work you can and then just like keep work you can and then just like keep work you can and then just like keep thinking about what I should do next, thinking about what I should do next, thinking about what I should do next, you know, what an interesting thing is, you know, what an interesting thing is, you know, what an interesting thing is, like just spend more compute making your like just spend more compute making your like just spend more compute making your output better for me the next morning. I output better for me the next morning. I output better for me the next morning. I would drag that slider quite far. I'd be would drag that slider quite far. I'd be would drag that slider quite far. I'd be willing to spend a lot for that. willing to spend a lot for that. willing to spend a lot for that. Um but the amount of compute that that Um but the amount of compute that that Um but the amount of compute that that would require if everybody in the world would require if everybody in the world would require if everybody in the world wants to drag that slider pretty far, wants to drag that slider pretty far, wants to drag that slider pretty far, it's like a lot. it's like a lot. it's like a lot. >> I'd love to hear you talk about how you >> I'd love to hear you talk about how you >> I'd love to hear you talk about how you think of the nature of this new think of the nature of this new think of the nature of this new intelligence. Uh someone told me intelligence. Uh someone told me intelligence. Uh someone told me recently, you know, planes don't fly recently, you know, planes don't fly recently, you know, planes don't fly like a bird. And this intelligence is like a bird. And this intelligence is like a bird. And this intelligence is >> It's a very alien kind of intelligence. >> It's a very alien kind of intelligence. >> It's a very alien kind of intelligence. >> Yeah, it's a very alien kind of >> Yeah, it's a very alien kind of >> Yeah, it's a very alien kind of intelligence and everyone's talking intelligence and everyone's talking intelligence and everyone's talking about how if you could verify something, about how if you could verify something, about how if you could verify something, it's sort of going it's just going to it's sort of going it's just going to it's sort of going it's just going to win, right? Like it's with enough win, right? Like it's with enough win, right? Like it's with enough compute and enough IQ, like it will compute and enough IQ, like it will compute and enough IQ, like it will it'll just brute force its way to a it'll just brute force its way to a it'll just brute force its way to a solution, and then in other domains solution, and then in other domains solution, and then in other domains where humans and the data and evals that where humans and the data and evals that where humans and the data and evals that they've done have been a huge part of they've done have been a huge part of they've done have been a huge part of it, it's surprising to me like how much it, it's surprising to me like how much it, it's surprising to me like how much money it's cost to get good at, I don't money it's cost to get good at, I don't money it's cost to get good at, I don't know, law and reasoning tracing law or know, law and reasoning tracing law or know, law and reasoning tracing law or something. I'm just curious like I'm not something. I'm just curious like I'm not something. I'm just curious like I'm not sure how sure how sure how >> It's a beautiful question. >> It's a beautiful question. >> It's a beautiful question. >> your kid is. Well, you have a boy or a >> your kid is. Well, you have a boy or a >> your kid is. Well, you have a boy or a girl? girl? girl? >> I have a >> I have a >> I have a >> When they're seven or age of reason or >> When they're seven or age of reason or >> When they're seven or age of reason or whatever and they can you can describe whatever and they can you can describe whatever and they can you can describe to them like what is the nature of this to them like what is the nature of this to them like what is the nature of this intelligence? Like how would you intelligence? Like how would you intelligence? Like how would you describe it? describe it? describe it? >> It's a beautiful question. I I >> It's a beautiful question. I I >> It's a beautiful question. I I I don't think I've been asked this I don't think I've been asked this I don't think I've been asked this before. I I before. I I before. I I or even any version of it. The thing or even any version of it. The thing or even any version of it. The thing that's coming to mind right now is I that's coming to mind right now is I that's coming to mind right now is I would just say it's like a computer. would just say it's like a computer. would just say it's like a computer. And it's like a computer in the way that And it's like a computer in the way that And it's like a computer in the way that it can do a lot of things that people just do a lot of things that people just can't do like multiply can't do like multiply can't do like multiply two gigantic numbers very quickly and two gigantic numbers very quickly and two gigantic numbers very quickly and give you the answer. give you the answer. give you the answer. And then it cannot do some things that And then it cannot do some things that And then it cannot do some things that you would you would you would very easily do. The number of things very easily do. The number of things very easily do. The number of things that it can't do I expect to keep that it can't do I expect to keep that it can't do I expect to keep receding, receding, receding, but in an evolving world, I think human but in an evolving world, I think human but in an evolving world, I think human judgment and taste judgment and taste judgment and taste will continue to be hard for AIs to will continue to be hard for AIs to will continue to be hard for AIs to model like where that's going to go. I model like where that's going to go. I model like where that's going to go. I don't have the right word for this. It's don't have the right word for this. It's don't have the right word for this. It's not quite taste. The world may need like not quite taste. The world may need like not quite taste. The world may need like a a new kind of word for the kind of a a new kind of word for the kind of a a new kind of word for the kind of judgment that people are very good at judgment that people are very good at judgment that people are very good at that AIs seem to really deeply struggle that AIs seem to really deeply struggle that AIs seem to really deeply struggle with. with. with. >> What's it been like becoming a dad and >> What's it been like becoming a dad and >> What's it been like becoming a dad and having growing kids in this era? having growing kids in this era? having growing kids in this era? I'm thinking back to your optimistic I'm thinking back to your optimistic I'm thinking back to your optimistic early internet days. They're going to early internet days. They're going to early internet days. They're going to grow up in cheap abundant intelligence grow up in cheap abundant intelligence grow up in cheap abundant intelligence age. age. age. >> Having kids is by far the best thing uh >> Having kids is by far the best thing uh >> Having kids is by far the best thing uh I've ever done. Uh I've ever done. Uh I've ever done. Uh And everybody says that. Everybody says And everybody says that. Everybody says And everybody says that. Everybody says you can't really understand it. And so I you can't really understand it. And so I you can't really understand it. And so I kind of knew that I believed enough kind of knew that I believed enough kind of knew that I believed enough people that said it that I believed it people that said it that I believed it people that said it that I believed it to be true. to be true. to be true. But But But the degree to which it has been true for the degree to which it has been true for the degree to which it has been true for me has been surprising like that the I me has been surprising like that the I me has been surprising like that the I think I have the best most interesting think I have the best most interesting think I have the best most interesting job in the world. And it is still a very job in the world. And it is still a very job in the world. And it is still a very distant second to having kids. distant second to having kids. distant second to having kids. So it's been awesome uh and it is a real So it's been awesome uh and it is a real So it's been awesome uh and it is a real moment for moment for moment for optimism. My kids will never grow up in optimism. My kids will never grow up in optimism. My kids will never grow up in a world where they were smarter than a world where they were smarter than a world where they were smarter than computers. You were born in the time of computers. You were born in the time of computers. You were born in the time of GPT-3. You had a time where you had GPT-3. You had a time where you had GPT-3. You had a time where you had better reasoning than the models, even better reasoning than the models, even better reasoning than the models, even though you didn't when you were born. though you didn't when you were born. though you didn't when you were born. Yeah, you caught them briefly. That will Yeah, you caught them briefly. That will Yeah, you caught them briefly. That will never seem strange to him. That will never seem strange to him. That will never seem strange to him. That will never bother him. I don't think he'll never bother him. I don't think he'll never bother him. I don't think he'll care. I think he will He would be like care. I think he will He would be like care. I think he will He would be like shocked to imagine in the dark ages when shocked to imagine in the dark ages when shocked to imagine in the dark ages when we had to like deal with products and we had to like deal with products and we had to like deal with products and services that weren't incredibly smart. services that weren't incredibly smart. services that weren't incredibly smart. He will be able to do things He will be able to do things He will be able to do things that you and I never were able to do and that you and I never were able to do and that you and I never were able to do and he'll have expectations in life that you he'll have expectations in life that you he'll have expectations in life that you and I never had and and I never had and and I never had and you know, all of a like a much bigger you know, all of a like a much bigger you know, all of a like a much bigger canvas. canvas. canvas. >> Do you run the business or teams or lead >> Do you run the business or teams or lead >> Do you run the business or teams or lead people in any way that is notably people in any way that is notably people in any way that is notably different because of the experience of different because of the experience of different because of the experience of having them? having them? having them? >> The answer must be yes. >> The answer must be yes. >> The answer must be yes. I feel very different having them. I I feel very different having them. I I feel very different having them. I think there's like a bunch of small think there's like a bunch of small think there's like a bunch of small things that are are really different and things that are are really different and things that are are really different and then, then, then, you know, again, this is like not a you know, again, this is like not a you know, again, this is like not a novel insight in any way. I think most novel insight in any way. I think most novel insight in any way. I think most people who have had kids say as you people who have had kids say as you people who have had kids say as you know, as soon as you have a kid, you know, as soon as you have a kid, you know, as soon as you have a kid, you like realize that you care much more like realize that you care much more like realize that you care much more about about about them and the experience they're going to them and the experience they're going to them and the experience they're going to have you do about yourself and the world have you do about yourself and the world have you do about yourself and the world that you are going to leave them and I that you are going to leave them and I that you are going to leave them and I think I have a sort of like think I have a sort of like think I have a sort of like unusual vantage point for that and and unusual vantage point for that and and unusual vantage point for that and and like people ask me sometimes like, oh, like people ask me sometimes like, oh, like people ask me sometimes like, oh, you know, now that you have kids, do you you know, now that you have kids, do you you know, now that you have kids, do you care care care are you worried about your safety and, are you worried about your safety and, are you worried about your safety and, you know, you know, you know, not destroying the world and the answer not destroying the world and the answer not destroying the world and the answer is like, I didn't need kids for I really is like, I didn't need kids for I really is like, I didn't need kids for I really didn't want to destroy the world before. didn't want to destroy the world before. didn't want to destroy the world before. But, do I think more about the role of But, do I think more about the role of But, do I think more about the role of like like like human agency and what it means to have a human agency and what it means to have a human agency and what it means to have a fulfilling life? Definitely much more. fulfilling life? Definitely much more. fulfilling life? Definitely much more. For what we're building and also like For what we're building and also like For what we're building and also like the people I work with, I want them to the people I work with, I want them to the people I work with, I want them to have it, too. You obviously have have it, too. You obviously have have it, too. You obviously have extraordinary empathy for your kids, but extraordinary empathy for your kids, but extraordinary empathy for your kids, but the degree to which that kind of extends the degree to which that kind of extends the degree to which that kind of extends to to to all kids all kids all kids and then maybe to all parents and to and then maybe to all parents and to and then maybe to all parents and to maybe then to everybody, like that's maybe then to everybody, like that's maybe then to everybody, like that's been a surprise to me, too. been a surprise to me, too. been a surprise to me, too. >> In in one of the posts I I think it was >> In in one of the posts I I think it was >> In in one of the posts I I think it was the one that's things you wish you knew the one that's things you wish you knew the one that's things you wish you knew earlier or something. Um is about earlier or something. Um is about earlier or something. Um is about incentives. Set them very, very incentives. Set them very, very incentives. Set them very, very carefully. carefully. carefully. >> Yeah. >> Yeah. >> Yeah. >> It's always been one of the most >> It's always been one of the most >> It's always been one of the most puzzling and interesting things about puzzling and interesting things about puzzling and interesting things about you that you don't have equity exposure you that you don't have equity exposure you that you don't have equity exposure to this company. How should the world to this company. How should the world to this company. How should the world think about your incentives? think about your incentives? think about your incentives? >> I don't know what I can say beyond like >> I don't know what I can say beyond like >> I don't know what I can say beyond like I have a front row seat I have a front row seat I have a front row seat to the most exciting moment of human to the most exciting moment of human to the most exciting moment of human history. history. history. And like that is worth more to me than And like that is worth more to me than And like that is worth more to me than any amount of money. I get to have an any amount of money. I get to have an any amount of money. I get to have an extremely interesting life and work with extremely interesting life and work with extremely interesting life and work with extraordinary people on something that I extraordinary people on something that I extraordinary people on something that I deeply care about. But somehow that deeply care about. But somehow that deeply care about. But somehow that doesn't count like that doesn't doesn't count like that doesn't doesn't count like that doesn't do it for people or something. do it for people or something. do it for people or something. >> I'm curious how you think about >> I'm curious how you think about >> I'm curious how you think about robotics. Like you mentioned earlier at robotics. Like you mentioned earlier at robotics. Like you mentioned earlier at some point if we had automated labor in some point if we had automated labor in some point if we had automated labor in the same way we're going to have the same way we're going to have the same way we're going to have automated intelligence. automated intelligence. automated intelligence. Things might get even crazier. Labor Things might get even crazier. Labor Things might get even crazier. Labor market is much bigger than the white market is much bigger than the white market is much bigger than the white collar market. collar market. collar market. >> If we don't have it, then things get >> If we don't have it, then things get >> If we don't have it, then things get really crazy. If the role for people in really crazy. If the role for people in really crazy. If the role for people in the world is to be like the actuators of the world is to be like the actuators of the world is to be like the actuators of AI in the cloud, AI in the cloud, AI in the cloud, >> Bad bad >> Bad bad >> Bad bad >> Very bad. Very bad. So I think it's like >> Very bad. Very bad. So I think it's like >> Very bad. Very bad. So I think it's like much crazier if we don't get it than if much crazier if we don't get it than if much crazier if we don't get it than if we do. It's an imperative. we do. It's an imperative. we do. It's an imperative. >> Help me understand your sense of >> Help me understand your sense of >> Help me understand your sense of progress progress progress in that because unlike in AI where in that because unlike in AI where in that because unlike in AI where everyone is now kind of on the same page everyone is now kind of on the same page everyone is now kind of on the same page of like it's going fast, I you can find of like it's going fast, I you can find of like it's going fast, I you can find extremely smart people that say it's extremely smart people that say it's extremely smart people that say it's like end of this year and you can find like end of this year and you can find like end of this year and you can find extremely smart people that say it's 20 extremely smart people that say it's 20 extremely smart people that say it's 20 years from now or something. years from now or something. years from now or something. >> 20 years. I would say we get the ChatGPT >> 20 years. I would say we get the ChatGPT >> 20 years. I would say we get the ChatGPT moment for robotics in the next like two moment for robotics in the next like two moment for robotics in the next like two or three years. or three years. or three years. >> What would that be? Like what do you >> What would that be? Like what do you >> What would that be? Like what do you know what that is? know what that is? know what that is? >> Something where most people have like a >> Something where most people have like a >> Something where most people have like a real wow. Not not like a I saw this real wow. Not not like a I saw this real wow. Not not like a I saw this video of a robot dog doing something video of a robot dog doing something video of a robot dog doing something crazy, but I was somehow able to crazy, but I was somehow able to crazy, but I was somehow able to convince myself that a a really convince myself that a a really convince myself that a a really important thing happened. One of the important thing happened. One of the important thing happened. One of the things about the ChatGPT moment was that things about the ChatGPT moment was that things about the ChatGPT moment was that you could just go use it. Like you you could just go use it. Like you you could just go use it. Like you didn't have to like believe someone who didn't have to like believe someone who didn't have to like believe someone who said AI is coming soon. You could just said AI is coming soon. You could just said AI is coming soon. You could just go try it. go try it. go try it. >> Yeah. >> Yeah. >> Yeah. >> And if you can go like, you know, type >> And if you can go like, you know, type >> And if you can go like, you know, type in a command and a robot can do in a command and a robot can do in a command and a robot can do something crazy and you can like watch something crazy and you can like watch something crazy and you can like watch it even if it's you're not physically it even if it's you're not physically it even if it's you're not physically there, I think that would have the same there, I think that would have the same there, I think that would have the same kind of like whoa, it just did this kind of like whoa, it just did this kind of like whoa, it just did this thing. thing. thing. >> Wasn't ChatGPT like not this monolithic >> Wasn't ChatGPT like not this monolithic >> Wasn't ChatGPT like not this monolithic goal, but sort of like a side experiment goal, but sort of like a side experiment goal, but sort of like a side experiment that you decided to release? Can you that you decided to release? Can you that you decided to release? Can you tell that that that story may be tell that that that story may be tell that that that story may be instructive for something similar instructive for something similar instructive for something similar happening in robotics. Everyone seems to happening in robotics. Everyone seems to happening in robotics. Everyone seems to want a full-blown release, but maybe want a full-blown release, but maybe want a full-blown release, but maybe it's something very different. it's something very different. it's something very different. >> When we launched GPT-3, >> When we launched GPT-3, >> When we launched GPT-3, um we're trying to make money, trying to um we're trying to make money, trying to um we're trying to make money, trying to get people to use this API. get people to use this API. get people to use this API. >> Yeah. >> Yeah. >> Yeah. >> And the only commercial use case that >> And the only commercial use case that >> And the only commercial use case that was really working the model was just so was really working the model was just so was really working the model was just so dumb. Like if you went back and used it, dumb. Like if you went back and used it, dumb. Like if you went back and used it, you'd be astonished. The only commercial you'd be astonished. The only commercial you'd be astonished. The only commercial use case that was working was use case that was working was use case that was working was copywriting. copywriting. copywriting. You know, so you pay like some marketing You know, so you pay like some marketing You know, so you pay like some marketing firm 20 bucks and they firm 20 bucks and they firm 20 bucks and they paid us 20 cents for the AI to like paid us 20 cents for the AI to like paid us 20 cents for the AI to like write you a landing page or whatever. write you a landing page or whatever. write you a landing page or whatever. But in addition to that one commercial But in addition to that one commercial But in addition to that one commercial use case, developers were using this use case, developers were using this use case, developers were using this thing we called the playground, which thing we called the playground, which thing we called the playground, which was like a testing interface was like a testing interface was like a testing interface to chat with the model. to chat with the model. to chat with the model. And it was really hard to do cuz we had And it was really hard to do cuz we had And it was really hard to do cuz we had not tuned the model to be good to chat not tuned the model to be good to chat not tuned the model to be good to chat with. So you had to like give it a few with. So you had to like give it a few with. So you had to like give it a few examples of what it means to chat and examples of what it means to chat and examples of what it means to chat and then do it. But people really liked it. then do it. But people really liked it. then do it. But people really liked it. And I had learned this And I had learned this And I had learned this great lesson from YC is that if you great lesson from YC is that if you great lesson from YC is that if you notice your users doing something like notice your users doing something like notice your users doing something like go down that Yeah. Go down that path. go down that Yeah. Go down that path. go down that Yeah. Go down that path. And so And so And so we decided that we would build a good we decided that we would build a good we decided that we would build a good chatbot since that's what people were chatbot since that's what people were chatbot since that's what people were doing. doing. doing. Um Um Um we started working on that. And we we started working on that. And we we started working on that. And we finished GPT-4. And we started using it finished GPT-4. And we started using it finished GPT-4. And we started using it internally and we're like, "This is a internally and we're like, "This is a internally and we're like, "This is a big deal." big deal." big deal." And we kind of thought that All right, And we kind of thought that All right, And we kind of thought that All right, this is going to be a real update to the this is going to be a real update to the this is going to be a real update to the world about AI and there's a bunch of world about AI and there's a bunch of world about AI and there's a bunch of hard questions here about hard questions here about hard questions here about you know, is this going to create a you know, is this going to create a you know, is this going to create a bunch of fake news? Is this going to say bunch of fake news? Is this going to say bunch of fake news? Is this going to say really offensive things? Are we going to really offensive things? Are we going to really offensive things? Are we going to get in trouble? So we decided we would get in trouble? So we decided we would get in trouble? So we decided we would start with a weaker version. Um start with a weaker version. Um start with a weaker version. Um The chat interface and GPT-4 at the same The chat interface and GPT-4 at the same The chat interface and GPT-4 at the same time seemed like a lot. So we would roll time seemed like a lot. So we would roll time seemed like a lot. So we would roll out the chat interface and GPT-3.5. out the chat interface and GPT-3.5. out the chat interface and GPT-3.5. In fact, it was originally going to be In fact, it was originally going to be In fact, it was originally going to be called Chat with GPT-3.5. called Chat with GPT-3.5. called Chat with GPT-3.5. And uh And uh And uh we didn't plan for it to be a product. we didn't plan for it to be a product. we didn't plan for it to be a product. Didn't think it'd be a huge hit, but did Didn't think it'd be a huge hit, but did Didn't think it'd be a huge hit, but did did think it would get people did think it would get people did think it would get people the world to like catch up with this and the world to like catch up with this and the world to like catch up with this and realize something was going on. realize something was going on. realize something was going on. And uh And uh And uh we mercifully renamed it ChatGPT a few we mercifully renamed it ChatGPT a few we mercifully renamed it ChatGPT a few hours before launch hours before launch hours before launch and put it out and put it out and put it out as like a research preview. mhm. as like a research preview. mhm. as like a research preview. mhm. And the thought was we'd put it out as a And the thought was we'd put it out as a And the thought was we'd put it out as a preview, and then a few months later we preview, and then a few months later we preview, and then a few months later we would launch a product would launch a product would launch a product with GPT-4. with GPT-4. with GPT-4. And for whatever reason, that model was And for whatever reason, that model was And for whatever reason, that model was over the threshold where even though we over the threshold where even though we over the threshold where even though we had gotten used to it internally, people had gotten used to it internally, people had gotten used to it internally, people said, "Okay, this is awesome." said, "Okay, this is awesome." said, "Okay, this is awesome." There maybe wasn't that much utility There maybe wasn't that much utility There maybe wasn't that much utility yet, but it was an incredible moment for yet, but it was an incredible moment for yet, but it was an incredible moment for people to feel people to feel people to feel AI progress and use something they AI progress and use something they AI progress and use something they enjoyed using. And then, by the time we enjoyed using. And then, by the time we enjoyed using. And then, by the time we put GPT-4, uh suddenly they really got put GPT-4, uh suddenly they really got put GPT-4, uh suddenly they really got benefit out of using, too. benefit out of using, too. benefit out of using, too. >> Vanta automates security and compliance >> Vanta automates security and compliance >> Vanta automates security and compliance for over 16,000 fast-moving companies for over 16,000 fast-moving companies for over 16,000 fast-moving companies like Ramp, Cursor, and Harvey, keeping like Ramp, Cursor, and Harvey, keeping like Ramp, Cursor, and Harvey, keeping them audit-ready around the clock. them audit-ready around the clock. them audit-ready around the clock. >> It's the number one agentless trust >> It's the number one agentless trust >> It's the number one agentless trust platform, [music] and it now helps platform, [music] and it now helps platform, [music] and it now helps companies like yours watch for the risks companies like yours watch for the risks companies like yours watch for the risks that show up between audits, across your that show up between audits, across your that show up between audits, across your vendors, your AI tools, and your whole vendors, your AI tools, and your whole vendors, your AI tools, and your whole environment. Every new tool your team environment. Every new tool your team environment. Every new tool your team signs up for, every vendor [music] that signs up for, every vendor [music] that signs up for, every vendor [music] that turns on AI features, is an opportunity turns on AI features, is an opportunity turns on AI features, is an opportunity for something to go wrong. And most for something to go wrong. And most for something to go wrong. And most security programs weren't built for AI's security programs weren't built for AI's security programs weren't built for AI's pace of growth. The Vanta agent works pace of growth. The Vanta agent works pace of growth. The Vanta agent works like a 24/7 GRC engineer in the like a 24/7 GRC engineer in the like a 24/7 GRC engineer in the background, finding issues, drafting background, finding issues, drafting background, finding issues, drafting fixes for you, [music] and cutting fixes for you, [music] and cutting fixes for you, [music] and cutting vendor assessment time by up to 50%. vendor assessment time by up to 50%. vendor assessment time by up to 50%. Whether you're a fast-growing [music] Whether you're a fast-growing [music] Whether you're a fast-growing [music] startup or a global enterprise, Vanta startup or a global enterprise, Vanta startup or a global enterprise, Vanta helps you earn and prove trust. helps you earn and prove trust. helps you earn and prove trust. Invest like the best listeners get a Invest like the best listeners get a Invest like the best listeners get a special offer for $1,000 off at special offer for $1,000 off at special offer for $1,000 off at vanta.com/invest. vanta.com/invest. vanta.com/invest. [music] [music] [music] Ridgeline is the first end-to-end system Ridgeline is the first end-to-end system Ridgeline is the first end-to-end system of record with embedded AI for of record with embedded AI for of record with embedded AI for investment management firms, [music] investment management firms, [music] investment management firms, [music] running portfolio accounting, running portfolio accounting, running portfolio accounting, reconciliation, reporting, trading, and reconciliation, reporting, trading, and reconciliation, reporting, trading, and compliance [music] on one unified compliance [music] on one unified compliance [music] on one unified platform. platform. platform. Firms are moving off legacy technology Firms are moving off legacy technology Firms are moving off legacy technology and onto Ridgeline because of how far and onto Ridgeline because of how far and onto Ridgeline because of how far ahead Ridgeline's AI features are ahead Ridgeline's AI features are ahead Ridgeline's AI features are compared to anything [music] else in compared to anything [music] else in compared to anything [music] else in investment management software. investment management software. investment management software. I've been hearing from a lot of I've been hearing from a lot of I've been hearing from a lot of investment managers about AI, and they investment managers about AI, and they investment managers about AI, and they fall roughly into two camps, [music] fall roughly into two camps, [music] fall roughly into two camps, [music] with some unsure of where to even start, with some unsure of where to even start, with some unsure of where to even start, and others convinced they can build and others convinced they can build and others convinced they can build their own order management system over a their own order management system over a their own order management system over a weekend. The reality is that running an weekend. The reality is that running an weekend. The reality is that running an investment firm will always require investment firm will always require investment firm will always require governance controls, [music] governance controls, [music] governance controls, [music] and a single source of truth for your and a single source of truth for your and a single source of truth for your data. And no amount of AI enthusiasm data. And no amount of AI enthusiasm data. And no amount of AI enthusiasm changes that requirement. Ridgeline is changes that requirement. Ridgeline is changes that requirement. Ridgeline is built on exactly that foundation, which built on exactly that foundation, which built on exactly that foundation, which is why I believe that firms that come is why I believe that firms that come is why I believe that firms that come out ahead in the AI era will be the ones out ahead in the AI era will be the ones out ahead in the AI era will be the ones running on Ridgeline's unified platform. running on Ridgeline's unified platform. running on Ridgeline's unified platform. If you're serious about your firm's AI If you're serious about your firm's AI If you're serious about your firm's AI [music] strategy, Ridgeline should be [music] strategy, Ridgeline should be [music] strategy, Ridgeline should be part of that conversation. You can part of that conversation. You can part of that conversation. You can request a demo at ridgeline.ai. [music] request a demo at ridgeline.ai. [music] request a demo at ridgeline.ai. [music] >> Are you surprised that that remains kind >> Are you surprised that that remains kind >> Are you surprised that that remains kind of the intuitive interface between us of the intuitive interface between us of the intuitive interface between us and this alien intelligence, even and this alien intelligence, even and this alien intelligence, even including coding? Like mostly that's me including coding? Like mostly that's me including coding? Like mostly that's me talking to the computer telling it what talking to the computer telling it what talking to the computer telling it what to build. to build. to build. >> No, because I'm like a massive texter. >> No, because I'm like a massive texter. >> No, because I'm like a massive texter. I've been a massive texter my whole I've been a massive texter my whole I've been a massive texter my whole life. I think part of my own insight of life. I think part of my own insight of life. I think part of my own insight of why that was a good interface is I'm why that was a good interface is I'm why that was a good interface is I'm like like like I know how to do this. I know how to do I know how to do this. I know how to do I know how to do this. I know how to do this. I know what it's like to just like this. I know what it's like to just like this. I know what it's like to just like start chatting in a text box. start chatting in a text box. start chatting in a text box. >> Any other thoughts on this notion of >> Any other thoughts on this notion of >> Any other thoughts on this notion of diffusion and how to make it faster? diffusion and how to make it faster? diffusion and how to make it faster? Like if the mission is get intelligence Like if the mission is get intelligence Like if the mission is get intelligence into the hands and more useful for into the hands and more useful for into the hands and more useful for everyone, a key part of that is like, I everyone, a key part of that is like, I everyone, a key part of that is like, I don't know, a marketing campaign or don't know, a marketing campaign or don't know, a marketing campaign or something. Like how How do you get this something. Like how How do you get this something. Like how How do you get this to diffuse faster to diffuse faster to diffuse faster than it seems to be doing naturally to than it seems to be doing naturally to than it seems to be doing naturally to me? me? me? >> I think the key thing is uh just make it >> I think the key thing is uh just make it >> I think the key thing is uh just make it better. Like I kind of believe that a better. Like I kind of believe that a better. Like I kind of believe that a truly great product markets itself. truly great product markets itself. truly great product markets itself. There was no ChatGPT marketing campaign There was no ChatGPT marketing campaign There was no ChatGPT marketing campaign at the beginning. Um and I think as we at the beginning. Um and I think as we at the beginning. Um and I think as we get to this next stage of models and we get to this next stage of models and we get to this next stage of models and we figure out figure out figure out how to make products that are as great how to make products that are as great how to make products that are as great as the models themselves, there will be as the models themselves, there will be as the models themselves, there will be such incredible utility that people will such incredible utility that people will such incredible utility that people will spread it very quickly. Uh we should spread it very quickly. Uh we should spread it very quickly. Uh we should definitely do more marketing. Like the definitely do more marketing. Like the definitely do more marketing. Like the AI's not too popular. AI's not too popular. AI's not too popular. For as much as people use it, they're For as much as people use it, they're For as much as people use it, they're kind of they have very understandable kind of they have very understandable kind of they have very understandable anxiety about where it can go. And so anxiety about where it can go. And so anxiety about where it can go. And so that kind of stuff that kind of stuff that kind of stuff I think some great marketing would be I think some great marketing would be I think some great marketing would be helpful for. But in terms of value helpful for. But in terms of value helpful for. But in terms of value people are getting out of the products people are getting out of the products people are getting out of the products and getting the products to grow faster, and getting the products to grow faster, and getting the products to grow faster, better models, more compute, better better models, more compute, better better models, more compute, better products, that will do it. That there products, that will do it. That there products, that will do it. That there was this period where the recruiting of was this period where the recruiting of was this period where the recruiting of researchers, the retention of them, the researchers, the retention of them, the researchers, the retention of them, the incentivizing of them was like the incentivizing of them was like the incentivizing of them was like the defining story in the competitive defining story in the competitive defining story in the competitive landscape or whatever. I think there's landscape or whatever. I think there's landscape or whatever. I think there's lots of stories about you successfully lots of stories about you successfully lots of stories about you successfully recruiting great researchers, and recruiting great researchers, and recruiting great researchers, and there's been many that have come through there's been many that have come through there's been many that have come through Open AI and had huge impacts. Open AI and had huge impacts. Open AI and had huge impacts. Some of which are known, some of which Some of which are known, some of which Some of which are known, some of which are lesser known names. I'm just curious are lesser known names. I'm just curious are lesser known names. I'm just curious about this whole genre of like what you about this whole genre of like what you about this whole genre of like what you learned about how to recruit this class learned about how to recruit this class learned about how to recruit this class of person, what matters to them and and of person, what matters to them and and of person, what matters to them and and how you did it. I've never heard you how you did it. I've never heard you how you did it. I've never heard you talk about like the actual tactical like talk about like the actual tactical like talk about like the actual tactical like moves you pulled to recruit somebody. moves you pulled to recruit somebody. moves you pulled to recruit somebody. >> In the early days, I think it was quite >> In the early days, I think it was quite >> In the early days, I think it was quite simple. It which is that we believed simple. It which is that we believed simple. It which is that we believed that AGI was possible and it was worth that AGI was possible and it was worth that AGI was possible and it was worth going after and we're willing to say going after and we're willing to say going after and we're willing to say that. that. that. And that was like an insane heretical And that was like an insane heretical And that was like an insane heretical belief. When we first announced Open AI, belief. When we first announced Open AI, belief. When we first announced Open AI, all of these like, you know, giants of all of these like, you know, giants of all of these like, you know, giants of the field, the field, the field, these experts were saying this is like these experts were saying this is like these experts were saying this is like insane, it's hypie, it's irresponsible. insane, it's hypie, it's irresponsible. insane, it's hypie, it's irresponsible. Really respected people like Yann LeCun Really respected people like Yann LeCun Really respected people like Yann LeCun or whatever telling journalists like, or whatever telling journalists like, or whatever telling journalists like, "Oh, these guys aren't very good and "Oh, these guys aren't very good and "Oh, these guys aren't very good and it's not going to work." But the fact it's not going to work." But the fact it's not going to work." But the fact that we were able to say we're going to that we were able to say we're going to that we were able to say we're going to go for this, it really appealed to a certain kind of it really appealed to a certain kind of researcher that also wanted to like go researcher that also wanted to like go researcher that also wanted to like go on this crazy adventure with low on this crazy adventure with low on this crazy adventure with low probability of success probability of success probability of success and an ambitious and an ambitious and an ambitious kind of audacious vision is a very kind of audacious vision is a very kind of audacious vision is a very powerful recruiting tool. powerful recruiting tool. powerful recruiting tool. >> Right. You You've written that it's >> Right. You You've written that it's >> Right. You You've written that it's actually easier sometimes to build actually easier sometimes to build actually easier sometimes to build things that are harder because of this things that are harder because of this things that are harder because of this reason. Do you reason. Do you reason. Do you >> I super believe in this. It's one of my >> I super believe in this. It's one of my >> I super believe in this. It's one of my most frequent pieces of most frequent pieces of most frequent pieces of advice to YC founders and I try to advice to YC founders and I try to advice to YC founders and I try to really live it at Open AI. really live it at Open AI. really live it at Open AI. >> Just do something harder. >> Just do something harder. >> Just do something harder. >> Do something that matters. Like do >> Do something that matters. Like do >> Do something that matters. Like do something that something that something that is important and if you don't do it, if is important and if you don't do it, if is important and if you don't do it, if your company doesn't succeed, might not your company doesn't succeed, might not your company doesn't succeed, might not happen. happen. happen. >> You were an investor and are investor. >> You were an investor and are investor. >> You were an investor and are investor. Uh you've done a lot of it and at one Uh you've done a lot of it and at one Uh you've done a lot of it and at one point that's what you did. point that's what you did. point that's what you did. What have you learned about investors What have you learned about investors What have you learned about investors being on the other side? being on the other side? being on the other side? >> The number of investors that actually >> The number of investors that actually >> The number of investors that actually show up and try to help you show up and try to help you show up and try to help you is unbelievably small. Josh Kushner, is unbelievably small. Josh Kushner, is unbelievably small. Josh Kushner, absolute MVP investor, unbelievable, has absolute MVP investor, unbelievable, has absolute MVP investor, unbelievable, has like worked around the clock for what like worked around the clock for what like worked around the clock for what feels like years to help us. feels like years to help us. feels like years to help us. He is the only investor that I could He is the only investor that I could He is the only investor that I could point to that is point to that is point to that is proactively proactively proactively incredibly helpful all of the time. incredibly helpful all of the time. incredibly helpful all of the time. There are more people that could do There are more people that could do There are more people that could do that. that. that. Uh and there are many other investors Uh and there are many other investors Uh and there are many other investors that have also been helpful and that that have also been helpful and that that have also been helpful and that have great strategic advice and that do have great strategic advice and that do have great strategic advice and that do things when things when things when you know, we ask them to do it. But the you know, we ask them to do it. But the you know, we ask them to do it. But the like constant just relentless all in like constant just relentless all in like constant just relentless all in support is surprisingly rare from support is surprisingly rare from support is surprisingly rare from investors. Maybe I'm biased cuz I like investors. Maybe I'm biased cuz I like investors. Maybe I'm biased cuz I like always liked it when people said that always liked it when people said that always liked it when people said that about me. But I think founders really about me. But I think founders really about me. But I think founders really love that and it actually like moves the love that and it actually like moves the love that and it actually like moves the needle and as an investor it's the most needle and as an investor it's the most needle and as an investor it's the most fun way to do it. fun way to do it. fun way to do it. >> Me and my friend played this game where >> Me and my friend played this game where >> Me and my friend played this game where we text each other all the time and the we text each other all the time and the we text each other all the time and the prompt of the text is something I don't prompt of the text is something I don't prompt of the text is something I don't want you to know about me. want you to know about me. want you to know about me. What is What does that bring to mind? >> I'm tired. >> I'm tired. I don't know if so I've been doing this I don't know if so I've been doing this I don't know if so I've been doing this a long time. It's tiring. a long time. It's tiring. a long time. It's tiring. >> How do you get through that? >> How do you get through that? >> How do you get through that? >> Just keep going. >> Just keep going. >> Just keep going. >> It begs the question like is there a >> It begs the question like is there a >> It begs the question like is there a mount of being tired that would make you mount of being tired that would make you mount of being tired that would make you stop doing this? stop doing this? stop doing this? >> No, no, no. I I mean I'm I this is the >> No, no, no. I I mean I'm I this is the >> No, no, no. I I mean I'm I this is the coolest job in the world. I plan to do coolest job in the world. I plan to do coolest job in the world. I plan to do this for the rest of my career. But it's this for the rest of my career. But it's this for the rest of my career. But it's like much harder than I have a way to like much harder than I have a way to like much harder than I have a way to explain to people. I I feel very explain to people. I I feel very explain to people. I I feel very grateful to get to do this. This is not grateful to get to do this. This is not grateful to get to do this. This is not me complaining. me complaining. me complaining. >> What's coming next? Like we talked about >> What's coming next? Like we talked about >> What's coming next? Like we talked about automated AI researchers that next year automated AI researchers that next year automated AI researchers that next year or the year after like how do you think or the year after like how do you think or the year after like how do you think about what is happening in the next 6 to about what is happening in the next 6 to about what is happening in the next 6 to 36 months? Maybe that's too far out to 36 months? Maybe that's too far out to 36 months? Maybe that's too far out to forecast in this crazy exponential. forecast in this crazy exponential. forecast in this crazy exponential. >> Maybe a different version of the >> Maybe a different version of the >> Maybe a different version of the question is like let's say in you know question is like let's say in you know question is like let's say in you know month 23 from now we have somebody that month 23 from now we have somebody that month 23 from now we have somebody that agrees agrees is super intelligence. agrees agrees is super intelligence. agrees agrees is super intelligence. What happens in month 24? What happens in month 24? What happens in month 24? And my answer would be uh not very much. And my answer would be uh not very much. And my answer would be uh not very much. The the kind of like cult worship of the The the kind of like cult worship of the The the kind of like cult worship of the machine god machine god machine god states those people believe that like states those people believe that like states those people believe that like more is going to happen quicker than is more is going to happen quicker than is more is going to happen quicker than is going to happen. Eventually a lot will going to happen. Eventually a lot will going to happen. Eventually a lot will happen. But eventually a lot was going happen. But eventually a lot was going happen. But eventually a lot was going to happen anyway. Like the rate of human to happen anyway. Like the rate of human to happen anyway. Like the rate of human progress and you know, how different progress and you know, how different progress and you know, how different each decade is going to be and how much each decade is going to be and how much each decade is going to be and how much each decade is more different than the each decade is more different than the each decade is more different than the decade decade decade from before. That's been happening for a from before. That's been happening for a from before. That's been happening for a long time. Obviously ups and downs, but long time. Obviously ups and downs, but long time. Obviously ups and downs, but directionally. directionally. directionally. And I think the right way to think about And I think the right way to think about And I think the right way to think about this, everyone wants to be the hero of this, everyone wants to be the hero of this, everyone wants to be the hero of the story. Everyone wants to feel like the story. Everyone wants to feel like the story. Everyone wants to feel like they were there for the moment of the they were there for the moment of the they were there for the moment of the machine god and they played some crazy machine god and they played some crazy machine god and they played some crazy role, but you know, this is another role, but you know, this is another role, but you know, this is another step. step. step. And it was hard to imagine 50 years ago, And it was hard to imagine 50 years ago, And it was hard to imagine 50 years ago, and the step 50 years from now is hard and the step 50 years from now is hard and the step 50 years from now is hard to imagine today. to imagine today. to imagine today. And And And And I think the right mental framework And I think the right mental framework And I think the right mental framework is just the is just the is just the zoom way out. It's a pretty smooth zoom way out. It's a pretty smooth zoom way out. It's a pretty smooth exponential. exponential. exponential. >> Mhm. Tell me a little bit about the >> Mhm. Tell me a little bit about the >> Mhm. Tell me a little bit about the experience of watching Codex take off experience of watching Codex take off experience of watching Codex take off and how much that is tied to what I and how much that is tied to what I and how much that is tied to what I would describe as like a competitive would describe as like a competitive would describe as like a competitive advantage of distribution that you built advantage of distribution that you built advantage of distribution that you built through chat. through chat. through chat. And this is a gateway into a question And this is a gateway into a question And this is a gateway into a question about like moats in general in AI. Like about like moats in general in AI. Like about like moats in general in AI. Like what you think will drive real what you think will drive real what you think will drive real competitive advantage in a business over competitive advantage in a business over competitive advantage in a business over time? time? time? >> I think Codex mostly is winning because >> I think Codex mostly is winning because >> I think Codex mostly is winning because it's the best product and the best it's the best product and the best it's the best product and the best model. We do get some advantage from model. We do get some advantage from model. We do get some advantage from ChatGPT bundling, but very, very tiny. ChatGPT bundling, but very, very tiny. ChatGPT bundling, but very, very tiny. That is mostly not what it's been about. That is mostly not what it's been about. That is mostly not what it's been about. It has made me reflect a lot on this It has made me reflect a lot on this It has made me reflect a lot on this question of competitive advantage. Um, question of competitive advantage. Um, question of competitive advantage. Um, because because because you know, you know, you know, like brilliant intelligence can migrate like brilliant intelligence can migrate like brilliant intelligence can migrate from any product to any other product. from any product to any other product. from any product to any other product. And And And network effects still have a competitive network effects still have a competitive network effects still have a competitive advantage. Economic scale and the advantage. Economic scale and the advantage. Economic scale and the ability to like make the cheapest ability to like make the cheapest ability to like make the cheapest compute fleets whatever still have a compute fleets whatever still have a compute fleets whatever still have a competitive advantage. competitive advantage. competitive advantage. But the product advantage, like if we But the product advantage, like if we But the product advantage, like if we could get people to move over to Codex, could get people to move over to Codex, could get people to move over to Codex, then someone builds something better, then someone builds something better, then someone builds something better, they can get people to move from Codex. they can get people to move from Codex. they can get people to move from Codex. So it has made me reflect on that a lot. So it has made me reflect on that a lot. So it has made me reflect on that a lot. >> There's a really interesting question >> There's a really interesting question >> There's a really interesting question about whether this is going in the about whether this is going in the about whether this is going in the direction of a commodity. Like is direction of a commodity. Like is direction of a commodity. Like is intelligence going to be intelligence going to be intelligence going to be a a pure fungible commodity like rate of a a pure fungible commodity like rate of a a pure fungible commodity like rate of oil or something? oil or something? oil or something? >> Intelligence itself, I would say yes. >> Intelligence itself, I would say yes. >> Intelligence itself, I would say yes. >> So what is not going to be? >> So what is not going to be? >> So what is not going to be? >> Compute fleet. You know, like the scale >> Compute fleet. You know, like the scale >> Compute fleet. You know, like the scale of the compute fleet, the ability to of the compute fleet, the ability to of the compute fleet, the ability to make more compute. I think that's like a make more compute. I think that's like a make more compute. I think that's like a very durable advantage. Even if the very durable advantage. Even if the very durable advantage. Even if the product itself is not, because you know, product itself is not, because you know, product itself is not, because you know, Codex can write any piece of software Codex can write any piece of software Codex can write any piece of software you want. The workflows, the you want. The workflows, the you want. The workflows, the integrations, the sort of like complex integrations, the sort of like complex integrations, the sort of like complex processes, the ability for teams to processes, the ability for teams to processes, the ability for teams to collaborate together. That stuff is all collaborate together. That stuff is all collaborate together. That stuff is all pretty powerful. Even like brand pretty powerful. Even like brand pretty powerful. Even like brand preference preference preference and familiarity is pretty powerful. and familiarity is pretty powerful. and familiarity is pretty powerful. >> How excited are you about new obviously >> How excited are you about new obviously >> How excited are you about new obviously you've done interesting stuff in you've done interesting stuff in you've done interesting stuff in hardware that I'm sure you'll announce hardware that I'm sure you'll announce hardware that I'm sure you'll announce later this year? later this year? later this year? How how does that experiment feel and How how does that experiment feel and How how does that experiment feel and align with this sort of consumer align with this sort of consumer align with this sort of consumer distribution that you have? distribution that you have? distribution that you have? >> One of the reasons I'm interested in new >> One of the reasons I'm interested in new >> One of the reasons I'm interested in new hardware is we're talking earlier about hardware is we're talking earlier about hardware is we're talking earlier about how very powerful thing with AI is that how very powerful thing with AI is that how very powerful thing with AI is that it can be it can be it can be always on and proactive and just always on and proactive and just always on and proactive and just understand all your context, but current understand all your context, but current understand all your context, but current hardware is not is not good for that. hardware is not is not good for that. hardware is not is not good for that. Like we are working inside of a hardware Like we are working inside of a hardware Like we are working inside of a hardware paradigm that is paradigm that is paradigm that is 50 years old something like that. Um and 50 years old something like that. Um and 50 years old something like that. Um and computers are amazing. Keyboard and computers are amazing. Keyboard and computers are amazing. Keyboard and mouse monitors are amazing thing. But mouse monitors are amazing thing. But mouse monitors are amazing thing. But like we have to shape AI into that and I like we have to shape AI into that and I like we have to shape AI into that and I would I'm excited to think about I would would I'm excited to think about I would would I'm excited to think about I would love AI to be able to reference this love AI to be able to reference this love AI to be able to reference this conversation, but not so much that I'm conversation, but not so much that I'm conversation, but not so much that I'm willing to like crack my laptop open put willing to like crack my laptop open put willing to like crack my laptop open put it here and have it like looking at you it here and have it like looking at you it here and have it like looking at you and listening to us while it's going. and listening to us while it's going. and listening to us while it's going. But I would like a piece of hardware But I would like a piece of hardware But I would like a piece of hardware that socially that socially that socially was acceptable to do that and also felt was acceptable to do that and also felt was acceptable to do that and also felt like it was designed for that kind of a like it was designed for that kind of a like it was designed for that kind of a thing. thing. thing. >> As you think about the open questions, >> As you think about the open questions, >> As you think about the open questions, what debates in your own head, with your what debates in your own head, with your what debates in your own head, with your friends, with people that your friends, with people that your friends, with people that your colleagues here, what are the most colleagues here, what are the most colleagues here, what are the most interesting open debates or open interesting open debates or open interesting open debates or open questions that you you don't feel questions that you you don't feel questions that you you don't feel certain about but feel important? certain about but feel important? certain about but feel important? >> One that I don't think gets much >> One that I don't think gets much >> One that I don't think gets much attention attention attention is how how are we going to avoid is how how are we going to avoid is how how are we going to avoid cognitive atrophy? cognitive atrophy? cognitive atrophy? How are we going to use these tools and How are we going to use these tools and How are we going to use these tools and make sure that we are like stretching make sure that we are like stretching make sure that we are like stretching our brains more and more and continuing our brains more and more and continuing our brains more and more and continuing to understand stuff that that really to understand stuff that that really to understand stuff that that really matters. Um matters. Um matters. Um There's lots of versions of this that There's lots of versions of this that There's lots of versions of this that don't. Like I I remember when I was in don't. Like I I remember when I was in don't. Like I I remember when I was in the school I had this professor tell me the school I had this professor tell me the school I had this professor tell me like you got to understand compilers. If like you got to understand compilers. If like you got to understand compilers. If you don't, you will never be able to be you don't, you will never be able to be you don't, you will never be able to be a good programmer. Somehow that wasn't a good programmer. Somehow that wasn't a good programmer. Somehow that wasn't quite right. But quite right. But quite right. But understanding understanding understanding at a reasonable level like how the major at a reasonable level like how the major at a reasonable level like how the major components of the computer system work components of the computer system work components of the computer system work has been important to me. has been important to me. has been important to me. >> Forced to imagine a scenario where we >> Forced to imagine a scenario where we >> Forced to imagine a scenario where we are somehow oversupplied in compute in 2 are somehow oversupplied in compute in 2 are somehow oversupplied in compute in 2 years time, what would be that story? years time, what would be that story? years time, what would be that story? >> It does feel possible. If the models get >> It does feel possible. If the models get >> It does feel possible. If the models get so smart and so efficient that they can so smart and so efficient that they can so smart and so efficient that they can kind of do everything we need and, you kind of do everything we need and, you kind of do everything we need and, you know, build every piece of software we know, build every piece of software we know, build every piece of software we want. And if the bounds of our attention want. And if the bounds of our attention want. And if the bounds of our attention are such that like they just cannot are such that like they just cannot are such that like they just cannot absorb more than what it turns out a absorb more than what it turns out a absorb more than what it turns out a fairly limited amount of compute can do, fairly limited amount of compute can do, fairly limited amount of compute can do, then we can get into over-supply. Also, then we can get into over-supply. Also, then we can get into over-supply. Also, if we don't have the cost curve down cuz if we don't have the cost curve down cuz if we don't have the cost curve down cuz we hit some sort of scaling wall, we we hit some sort of scaling wall, we we hit some sort of scaling wall, we could also get into over-supply. Like could also get into over-supply. Like could also get into over-supply. Like the the observation about un-capped the the observation about un-capped the the observation about un-capped demand demand demand implies a certain price. implies a certain price. implies a certain price. >> Can you give your point of view on >> Can you give your point of view on >> Can you give your point of view on scaling laws today? scaling laws today? scaling laws today? >> Looking great. >> Looking great. >> Looking great. >> Just looking good. >> Just looking good. >> Just looking good. >> In some sense, scaling laws are like the >> In some sense, scaling laws are like the >> In some sense, scaling laws are like the most hated prediction of all time. most hated prediction of all time. most hated prediction of all time. Everybody always wants to say "No, no, Everybody always wants to say "No, no, Everybody always wants to say "No, no, no, it can't be like this." And And yet no, it can't be like this." And And yet no, it can't be like this." And And yet it keeps going. it keeps going. it keeps going. >> Who are your favorite unsung heroes in >> Who are your favorite unsung heroes in >> Who are your favorite unsung heroes in this company's story? this company's story? this company's story? >> First person that came to mind is Alec >> First person that came to mind is Alec >> First person that came to mind is Alec Radford. Alec Radford is probably the Radford. Alec Radford is probably the Radford. Alec Radford is probably the most important most important most important not very well-known researcher in the not very well-known researcher in the not very well-known researcher in the whole history of the field. And also whole history of the field. And also whole history of the field. And also just a wonderful, like, just a wonderful, like, just a wonderful, like, top top-tier human being. Um he did the top top-tier human being. Um he did the top top-tier human being. Um he did the work that work that work that really really really became the GPT series uh among many became the GPT series uh among many became the GPT series uh among many other important things. Um but he also other important things. Um but he also other important things. Um but he also is someone who inspired, guided, is someone who inspired, guided, is someone who inspired, guided, nudged people in many other directions nudged people in many other directions nudged people in many other directions that turned out to be super important. that turned out to be super important. that turned out to be super important. And I think that I think is cool about And I think that I think is cool about And I think that I think is cool about him is if you talk to people that worked him is if you talk to people that worked him is if you talk to people that worked with him, they will with him, they will with him, they will they will of course say, they will of course say, they will of course say, you know, "generational genius, you know, "generational genius, you know, "generational genius, brilliant innovative thinker, just so brilliant innovative thinker, just so brilliant innovative thinker, just so deep in his understanding and his and deep in his understanding and his and deep in his understanding and his and his work." But everybody everybody will his work." But everybody everybody will his work." But everybody everybody will tell you before they finish their tell you before they finish their tell you before they finish their statement that just like one of the statement that just like one of the statement that just like one of the nicest, most positive, best people nicest, most positive, best people nicest, most positive, best people they've ever interacted with. they've ever interacted with. they've ever interacted with. >> I love formative moments. And so, as we >> I love formative moments. And so, as we >> I love formative moments. And so, as we wind up here, I'm curious to ask one of wind up here, I'm curious to ask one of wind up here, I'm curious to ask one of each. If you think about the whole each. If you think about the whole each. If you think about the whole OpenAI experience, what moment or OpenAI experience, what moment or OpenAI experience, what moment or chapter or whatever are you most proud chapter or whatever are you most proud chapter or whatever are you most proud of? Start with the other one which is of? Start with the other one which is of? Start with the other one which is what was like the most instructive thing what was like the most instructive thing what was like the most instructive thing that maybe you got wrong or did wrong or that maybe you got wrong or did wrong or that maybe you got wrong or did wrong or what have you and and what was what was what have you and and what was what was what have you and and what was what was it like to learn from it? it like to learn from it? it like to learn from it? >> I mean a lot of things have gone wrong. >> I mean a lot of things have gone wrong. >> I mean a lot of things have gone wrong. A formative one that went wrong, which I A formative one that went wrong, which I A formative one that went wrong, which I haven't talked about much, is we made a haven't talked about much, is we made a haven't talked about much, is we made a mistake to try to innovate in our mistake to try to innovate in our mistake to try to innovate in our structure in the beginning. structure in the beginning. structure in the beginning. We had very good reason for it, which is We had very good reason for it, which is We had very good reason for it, which is we didn't know we didn't know we didn't know how we were ever going to make money and how we were ever going to make money and how we were ever going to make money and we really at the time we really at the time we really at the time weren't sure at all what we were going weren't sure at all what we were going weren't sure at all what we were going to look like when we grew up and of to look like when we grew up and of to look like when we grew up and of course we care about our mission and we course we care about our mission and we course we care about our mission and we wanted to like be structured in a way wanted to like be structured in a way wanted to like be structured in a way where even if the technology went on a where even if the technology went on a where even if the technology went on a very fast takeoff, our mission was very fast takeoff, our mission was very fast takeoff, our mission was protected. And so we had this like, you protected. And so we had this like, you protected. And so we had this like, you know, nonprofit structure. know, nonprofit structure. know, nonprofit structure. But But But I definitely learned something about I definitely learned something about I definitely learned something about why people don't do that much. We would why people don't do that much. We would why people don't do that much. We would have saved ourselves a great deal of have saved ourselves a great deal of have saved ourselves a great deal of pain in many ways pain in many ways pain in many ways if we had not tried to innovate on our if we had not tried to innovate on our if we had not tried to innovate on our structure and found some other way to structure and found some other way to structure and found some other way to preserve the preserve the preserve the central importance of the mission. central importance of the mission. central importance of the mission. Maybe there was no other way. Maybe Maybe there was no other way. Maybe Maybe there was no other way. Maybe there was for what we are doing and kind there was for what we are doing and kind there was for what we are doing and kind of the importance of it, there was of the importance of it, there was of the importance of it, there was nothing other than an exotic structure nothing other than an exotic structure nothing other than an exotic structure we could have come up with. I really we could have come up with. I really we could have come up with. I really learned over the last decade a big learned over the last decade a big learned over the last decade a big lesson about why people don't usually do lesson about why people don't usually do lesson about why people don't usually do that. that. that. >> Is there anything else formative of your >> Is there anything else formative of your >> Is there anything else formative of your life that like makes you you that we life that like makes you you that we life that like makes you you that we didn't talk about? I'm like this is like didn't talk about? I'm like this is like didn't talk about? I'm like this is like the the question that's always like the the the question that's always like the the the question that's always like the most interesting to me. most interesting to me. most interesting to me. >> Becoming relatively immune to >> Becoming relatively immune to >> Becoming relatively immune to people having strong opinions about me people having strong opinions about me people having strong opinions about me that I think I developed later in life that I think I developed later in life that I think I developed later in life as as like a realizing that man, just if as as like a realizing that man, just if as as like a realizing that man, just if you're going to be at the center of like you're going to be at the center of like you're going to be at the center of like this crazy revolution, everybody's going this crazy revolution, everybody's going this crazy revolution, everybody's going to project a lot of stuff onto you and to project a lot of stuff onto you and to project a lot of stuff onto you and you got to just quickly learn to make you got to just quickly learn to make you got to just quickly learn to make peace about that. I think there were peace about that. I think there were peace about that. I think there were also things I learned also things I learned also things I learned later in life about like later in life about like later in life about like how to be very calm and not anxious how to be very calm and not anxious how to be very calm and not anxious really about stuff. In terms of what really about stuff. In terms of what really about stuff. In terms of what drives me and what I care about and kind drives me and what I care about and kind drives me and what I care about and kind of like how I want to live my life, of like how I want to live my life, of like how I want to live my life, on the whole I felt like on the whole I felt like on the whole I felt like you know, for whatever reason the like you know, for whatever reason the like you know, for whatever reason the like 10-year-old version of me it pretty like 10-year-old version of me it pretty like 10-year-old version of me it pretty like fully formed. I think I just like kind fully formed. I think I just like kind fully formed. I think I just like kind of came out this way. How of came out this way. How of came out this way. How >> How about the thing you're proud of >> How about the thing you're proud of >> How about the thing you're proud of looking back on? looking back on? looking back on? >> I'm most proud of >> I'm most proud of >> I'm most proud of how many times we were right when the how many times we were right when the how many times we were right when the rest of the world was wrong in an rest of the world was wrong in an rest of the world was wrong in an important way that put the world on a important way that put the world on a important way that put the world on a trajectory now that I'm very proud to trajectory now that I'm very proud to trajectory now that I'm very proud to have played a role in. have played a role in. have played a role in. That feels awesome. And then also like That feels awesome. And then also like That feels awesome. And then also like for all the crap that's happened like for all the crap that's happened like for all the crap that's happened like the spiritual growth whatever you want the spiritual growth whatever you want the spiritual growth whatever you want to call it that I've gotten to have of to call it that I've gotten to have of to call it that I've gotten to have of like learning like learning like learning just incredible resilience and what that just incredible resilience and what that just incredible resilience and what that does for like making me happy in the does for like making me happy in the does for like making me happy in the rest of my life. rest of my life. rest of my life. Yeah, very grateful for that. Yeah, very grateful for that. Yeah, very grateful for that. >> When I do these I ask everyone the same >> When I do these I ask everyone the same >> When I do these I ask everyone the same traditional closing question. What is traditional closing question. What is traditional closing question. What is the kindest thing that anyone's ever the kindest thing that anyone's ever the kindest thing that anyone's ever done for you? done for you? done for you? >> I feel incredibly lucky about how many >> I feel incredibly lucky about how many >> I feel incredibly lucky about how many people have gone way out of their way to people have gone way out of their way to people have gone way out of their way to be very kind to me for my entire life. be very kind to me for my entire life. be very kind to me for my entire life. As I'm thinking of this, there's just As I'm thinking of this, there's just As I'm thinking of this, there's just this like montage of this like montage of this like montage of moments from life where people have been moments from life where people have been moments from life where people have been unbelievably nice to me. Yesterday my unbelievably nice to me. Yesterday my unbelievably nice to me. Yesterday my kid shared his blueberries with me for kid shared his blueberries with me for kid shared his blueberries with me for the first time. That was very sweet. the first time. That was very sweet. the first time. That was very sweet. >> [music] >> [music] >> [music] >> Good moment. >> Good moment. >> Good moment. >> Thanks, man. >> Thanks, man. >> Thanks, man. Thank you. >> You know how small advantages compound >> You know how small advantages compound over time? That's true in investing and over time? That's true in investing and over time? That's true in investing and just as true in how you run your just as true in how you run your just as true in how you run your company. Your spending system is your company. Your spending system is your company. Your spending system is your capital allocation strategy. [music] capital allocation strategy. [music] capital allocation strategy. [music] Ramp makes it smarter by default. Better Ramp makes it smarter by default. Better Ramp makes it smarter by default. Better data, better decisions, better economics data, better decisions, better economics data, better decisions, better economics over time. See how at ramp.com/invest. over time. See how at ramp.com/invest. over time. See how at ramp.com/invest. As your business grows, Vanta scales As your business grows, Vanta scales As your business grows, Vanta scales with you automating compliance and with you automating compliance and with you automating compliance and giving you a single [music] source of giving you a single [music] source of giving you a single [music] source of truth for security and risk. Learn more truth for security and risk. Learn more truth for security and risk. Learn more at vanta.com/invest. at vanta.com/invest. at vanta.com/invest. The best AI and software companies from The best AI and software companies from The best AI and software companies from OpenAI to Cursor to Perplexity use OpenAI to Cursor to Perplexity use OpenAI to Cursor to Perplexity use WorkOS to become enterprise ready WorkOS to become enterprise ready WorkOS to become enterprise ready overnight, not in months. Visit overnight, not in months. Visit overnight, not in months. Visit workos.com to skip the unglamorous workos.com to skip the unglamorous workos.com to skip the unglamorous infrastructure work and focus on your infrastructure work and focus on your infrastructure work and focus on your product. [music] Ridgeline is redefining product. [music] Ridgeline is redefining product. [music] Ridgeline is redefining asset management technology as a true asset management technology as a true asset management technology as a true partner, not just a software vendor. partner, not just a software vendor. partner, not just a software vendor. They've helped firms 5x in scale [music] They've helped firms 5x in scale [music] They've helped firms 5x in scale [music] enabling faster growth, smarter enabling faster growth, smarter enabling faster growth, smarter operations, and a competitive edge. operations, and a competitive edge. operations, and a competitive edge. Visit ridgelineapps.com to see what they Visit ridgelineapps.com to see what they Visit ridgelineapps.com to see what they can unlock for your firm. can unlock for your firm. can unlock for your firm. >> [music] >> [music] >> [music] >> Every investment firm is unique and >> Every investment firm is unique and >> Every investment firm is unique and generic AI doesn't understand your generic AI doesn't understand your generic AI doesn't understand your process. Rogo does. It's an AI platform process. Rogo does. It's an AI platform process. Rogo does. It's an AI platform built specifically for Wall Street, built specifically for Wall Street, built specifically for Wall Street, connected to your data, understanding connected to your data, understanding connected to your data, understanding your process, [music] and producing real your process, [music] and producing real your process, [music] and producing real outputs. Check them out at outputs. Check them out at outputs. Check them out at rogo.ai/invest.