Summary Nat Friedman (Former GitHub CEO): Building AI-Native Products & What’s Next For AI | TransformX 2022 - YouTube (Youtube) www.youtube.com
6,448 words - YouTube video - View YouTube video
Speaker 0 We join by Nat Friedman, Former Ceo of Github and Alexander Wang, Ceo and Founder of Scale AI-Native has founded 2 startups ups.
Speaker 1 Led Github as Ceo from 20 18 to 20 22. And now invest in infrastructure, Ai and developer company. Please join me in welcoming to the stage, Not Friedman and Alexander Way.
Speaker 2 Good to see you at. Good to
Speaker 3 see you, Alex. Thanks for having me.
Speaker 2 So Nat was the Ceo of Github most recently well, I guess now runs Ai grant, which funds innovative Ai products. To to improve all our lives. Nat has made a request that we get straight into the spice and sizzle Yes, please. So so start out with, Why don't you tell us about the story of Github c copilot, which for the developers in the audience is is a pretty darn magical thing.
Speaker 3 Yeah. Great. Well, thanks for having me. It's fun to be here. So I was Ceo of Github, which I'd gotten into that position.
Speaker 3 Actually by selling a company to Microsoft. And then in my position at Microsoft then leading the acquisition of Github and then installing myself as Ceo, and
Speaker 2 it sounds hostile when you described that, but
Speaker 3 it was it was benign and intention, but Gp 3 came out. In kind of May or June of 20 20. And like a lot of people I saw it and I thought it was crazy. I couldn't believe it. And I've been playing with it and I thought, I don't know what, but we github should do something with this and code.
Speaker 3 And so we got in touch with the open Ai folks and Sat had already in his wisdom set up a partnership with Open Ai and invested and open was... Subsequently training on Microsoft servers. And so there was like there was a preexisting existing relationship in which we could work together and we started exploring what that product might be. And it was a really interesting situation because we were kind of figuring it out in this foggy environment where we had this underlying capability that was while to demoed really well. But the question is, how do you take that and make it into a useful product.
Speaker 3 And we actually... We didn't know what it was. We had a few ideas Be actually the first thing that we thought might be the product because it's this natural language model it talked it understood things was maybe a chatbot a Q and A stack overflow competitors. So instead of googling for your question, you could ask, this really smart Ai, a question and it would answer it.
Speaker 2 I was just a startup but does that. But... Yeah.
Speaker 3 And I... You know, so that was the first idea and then we thought maybe a could synthesis, can it build code to spec if you file an issue, Can it just generate the poor request or could it review my code? Could it be a code review bot? And so we started just tinkering with all of these ideas. And, you know, the the the other thing that was happening was the model was improving.
Speaker 3 Open Ai was kind of building. New and improved versions of it every week and over several weeks and so we had new to play with. And what we found out that was kind of interesting was the model was very good, you know, the the phrase I used was that it alternate between spooky and cookie. So like, sometimes it was so good that you're were like, whoa, how on Earth is it reading my mind, like, this is crazy, and sometimes it would just output nonsense. And you know was it was was actually harmful.
Speaker 3 Like, you know, that it was wrong. And so it turns out in the context of an agent that you're a dialogue with, we're, like, I'm asking you a question. If you give me the wrong answer, 70 percent of the time, like, I'm not that's not an experience of coming back for, you know, a lot. And so the question that we ended up having to answer in figuring out where the product was here as we sort of explored product space. Was how do you take a model which is frequently wrong and still make it useful.
Speaker 3 And not just useful, but kind of fun to use? And so now I think you look at c copilot and you think it's this trivial obvious product. It's kind of an auto complete for code. But finding like both in the macro sense that that was the right product and then sort of the micro sense, how do you you, how fast does it need to be, How do you know when you're going to complete a line versus a block? What are the heuristic that determine that?
Speaker 3 Should it show you multiple examples or or possible outputs and you choose between them or should just show you 1. All these sort of questions were ideas that we didn't know the answer to initially. And And so finally, we ended up with something which I think it's not just useful, but it's sort of fun. And it has the... I've described it as being like a slot machine.
Speaker 3 Where it has this kind of low ongoing cost to use it. What’s kind of constantly serving you up suggestions for code that you might use. You learn yourself win to pay attention to the suggestions and when not to. So you get a sense, you in a way learn to go it, the way you learned to Google, Googling is a kind of skill, you learn learned sort of how to prompt the search engine and you get that skill with C copilot 2 and you learn when to ignore it, And then occasionally, you have this jackpot where it saves you 15 minutes or something like that and you're totally delighted. And it's this kind of randomized psychological reward that arrives periodically and you get addicted, So it's actually fun.
Speaker 3 And so that was why it works. It's a product and millions people love it and use it.
Speaker 2 Yeah. I I was talking to a developer the other day and he was he was coding on a on an airplane. And he asked he was like, why do I feel so unproductive on an airplane and he realizes because he didn't have c copilot? Yeah. Which is great name by the way, c copilot vote.
Speaker 3 Yeah. 1 of the engineers and the team came up with that. And I think it was perfect because it sort of frames you're the pilot. You know, I'm your c copilot. You're responsible, but I'm here to help.
Speaker 3 And I I think that's... The other thing that's interesting is in in the sort of user interface paradigm, dialogue was this very central idea to many Ai people, which is this concept that there's an agent. You interrogated it. You you pose questions to it answers But the sort of challenge with that is that while I'm working, I have to reform whatever I'm trying to figure out into a question, it requires some cognitive work. And the c copilot ideas instead of being on opposite sides of the table talking to each other, I'm gonna be on the same side of the table as you looking at the screen, like you're looking at it, and just trying to speed you up and help out as you go.
Speaker 3 And I think that general rubric is extremely broadly applicable. And we'll see c copilot for x, whether it's lawyers or accountants or anyone who deals in language of any kind, I'm surprised we haven't seen more. Actually when I left github how at the end of last year, I thought there'd would be... I mean, like, we had basically taken this thing that was wrong 70 percent of the time. Was working in a domain where it's very objective whether it's right or not, like the code compile or it doesn't, and we'd found out a way to make a useful product out of it.
Speaker 3 It had taken 5 or 6 months of tinkering to get there. But I thought, okay, well, the world is going to do this now, and there will be, you know, because Gp 3 out there, there will be all these other types of products as people kind of tinker around. And there's really not very many yet. So I think we have this amazing... Exciting revolution this summer with image models, the text is somehow neglected.
Speaker 3 I think it'll change next year, but I'm surprised there isn't more product privatization yet.
Speaker 2 Yeah. Well, this... I'll go straight to another topic that I know you're passionate about, which is, you know, you you built c copilot in sort of the... In this environment of a a very big company. It would within the Microsoft environment and you through obviously was this deep collaboration with Open Ai.
Speaker 2 Yeah. And I think now you're investing in startups. But Think 1 of the big questions about these art artificial intelligence products, even the new ones stay, even the ones have been successful is do... You know, is there actually a window and and an a durable advantage for startups to build products on top of these large models or we going and get to a point where the big companies called whether it be a Microsoft or Google or an Adobe or whatnot are going to integrate Ai functionality deeply into their existing, very established. Tools.
Speaker 2 Yeah. And there's going be sort of no window for start up innovation.
Speaker 3 Yeah. I think this general question of what's the market structure, value creation and capture in this Ai landscape is kind of the big question like who who is gonna benefit. I think people will benefit broadly but in terms of companies who benefits and it's not clear. The narratives keeping wrong. So the narrative has been that Ai is the central force, whoever has the most data, most dollars for hardware you know is going to and the most distribution is going to be able to build products that the startup simply can't.
Speaker 3 And so it just enhances the most of the incumbents. And I think that's probably partly true. But what we've seen so far is that the know how diffuse. Quickly. And so this idea that maybe there's 1 or 2 organizations that have the technical expertise, they keep some secrets, the secrets enable them as this priest hood to control this new you know, they're trying to summon a God using a language named after a snake.
Speaker 3 It's like very mystical. But instead all the Ml folks, are all friends and they live in group houses and they tell everybody what's going on. And so, like, the secrets are not kept What’s half life of a secret and this field is like 3 to 6 months. And they're also the secrets are very obvious, like each paper... The papers are long and imp.
Speaker 3 But after you spend hours reading them, it's like 10 lines of code or something like that. Like, they're relatively... Once you know a secret, it's hard not to... Talk about, because it's so simple. So the secrets don't seem to be a thing.
Speaker 3 I do think incumbents will have a lot of benefits and there will be a category of bolt on features. Where you take your distribution and you add a c copilot to it or whatever, and it works great and maybe a startup that was focused, could have done something much better, but it was the pareto optimal product and you already stuffed it into your channel. And so that will happen. I think we'll also see startups. And I think the opportunity for startups to build new products is where the product doesn't fit neatly into an existing category because maybe it has a totally new workflow or user interface or...
Speaker 3 And and like that I think is something that where an incumbent might try to bolt it on to an existing product and it just like doesn't go potentially. The other 1 is the reputation or underwriting it's necessary to do some of this quote AI-Native Ai stuff, where like, What’s if it says something offensive. You know when we were shipping c copilot, there were a lot of people who were very worried about that. And, Yeah. Maybe stop there.
Speaker 3 But but we ship it anyway. So... And it probably does say offensive things. But, you know, if you're a big company with a big brand with a big business in this tiny thing that might make a hundred million dollars a year. Not going to move the needle on your business.
Speaker 3 Why would you go through all this reputation risk for it. I think some conservative incumbents won't take that bad. It's I think the flip flipside though is that's really interesting here is that in these previous technology revolutions, the new platform was a joke. So you kinda laughed the web. You said the web can't do all these things.
Speaker 3 You know, people are not gonna to move away from desktop applications. Or the smartphone, you know, we've seen it before and no one's laughing right now. And I think that's kind of interesting. Like, What’s. It might be bad news for startups.
Speaker 3 And I think the reason is that all the big companies have had 10 years of Ml and they've seen benefits from Ml and they have talent in their team that wants to work on this exciting new stuff. So I think it'll be some split between incumbents and startups and I'd pick the startups.
Speaker 2 Yeah. Yeah. I think What’s I think it's a it's a really good analogy. I think everybody's paying very deep attention to Ai right now and and and it's sort of become this sort of like... Clear thing that everyone needs to focus on.
Speaker 2 And I think there's sort of these questions, if you can what are the mo for a startup. You know, similarly, I think 1 interesting thing I'd love to get your thoughts on is is there's been this topic of like you know, is the model the moat or is the or is the sort of like, application or the workflow of the mode? Yeah. And it... This has been a really interesting 1, because They I think open Ai love their business models predicated on this thought that, like, hey, we're gonna build to your point, this very advanced algorithm, model.
Speaker 2 And we're else also gonna build on top of it.
Speaker 3 Right?
Speaker 2 And then staple diffusion came in and and sort of and and blue that What’s open and and we'll have a mod here later today to sort of talk through that. But but I'm curious your thoughts. I think it's sort of AAAA very big sort of question mark of whether or not the models have
Speaker 3 Yeah.
Speaker 2 Have endured value. Yeah.
Speaker 3 I mean, the people who are in the model is the moat camp, are retreating to smaller and smaller future scenarios where models cost like a billion dollars basically. And because right now as a, you know, Gb 3 cost, whatever 10000000 to 20000000 to train, we can do it now for a fraction of that. Moore's law may help guarantee that that price continues to decline. Techniques are improving. We're occasionally finding you low hanging fruit on the order of 3x.
Speaker 3 Training efficiency gains. And so you've got you've gotta have some argument that you need to spend a lot of money. Like, so technical secrets aren't it maybe technical secrets can unlock the dollar thing to some extent. Data engines, could be. So if you have the distribution and so you have millions of users and you've got telemetry, but how they're using your product and that allows you to kind of have some take where your products just better because so you retrain your models constantly.
Speaker 3 Maybe that could be it, I think it's a classic battle though where it's about the startups getting distribution before the incumbents get product. And I think because of all this you know, all the sort of swirl around it. I think incumbents are unusually disabled right now. Like, we had a really interesting period, in the prior millennium where Clay Christ introduced sort of this disruption this idea of disruptive technology. And it was because you'd seen this pattern where incumbents were consistently disrupted by by new companies.
Speaker 3 And then the incumbents kind of learned about that. And they said, we shouldn't do that anymore, and they stopped being disrupted. And they bought competitors early enough or they were just willing to cannibal their own businesses. And now I think they'd become paralyzed again by internal cultural forces and maybe by government, some extent, like many of these companies simply aren't allowed. Acquire companies anymore.
Speaker 3 And so it's possible that there's a unique moment where you can really attack them to because they're run by more conservative people who are paid a huge amount of money and kind of don't actually want to win in these new categories. They wanna do something else.
Speaker 2 Yeah. Know, I think if if we wanna be on the side of broad innovation, I think you always wanna be on the side of the startups, but I think there's Yeah. There's a lot of... I mean, the the sort of frankly, c copilot is maybe 1 of the the the biggest points in the in the in the sort of big company column because c copilot took off... Long before any startup was able to build similar technology and we'll probably, you know, have have a lead in distribution for a very, very long time.
Speaker 3 I think What’s true, but there may be leap pro innovation. So I think we'll see next year, like new generations of models come out and you know, we have these as the models get bigger and better and they get... Soaked with more high quality data and compute. These emergent capabilities pop out where it like, couldn't do this thing at all, and now it can do it. And, you know, it's like 6 digit multiplication or something like that.
Speaker 2 So we may not be the best of them.
Speaker 3 Maybe have the best example. But it's sort sort it's sort interesting. And so... When you have emergent capabilities, you you have emerging products. You know, like, this product was not feasible before.
Speaker 3 Now it's feasible. And so I think, you know, will be there will be changes next year. My guess is Gb 3 is like, pretty good, but not great, basically. You've got a few good companies in products like Copilot and some of the copywriting and other things that have been built on it. But the next step change is where you see the real commercialization wave.
Speaker 2 Yeah. What’s 1 topic that we've also talked about before and is highly related to the Ai grant being started in the first place was that there sort of this Ai was born out of research Yep. Right. It was born as a as a research field. And for a long time, there remarkably few people in the Ai community who had...
Speaker 2 Who cared at all about building products or cared at all about building businesses and cared about sort of make getting the in the hands of of people and hiding good intuitions around it? And I think, you know, we were together the ai grant years and years ago and this sort of 1 of the initial theories is that, like, how do we How do we turn Ai just to, like, real things that actually have an impact and and now you've obviously continue with that in a big way. But I'm I'm curious to your thoughts like, how do you think about, you know, research versus products. Yeah. And and how this gonna evolve?
Speaker 3 Yeah I mean it's really interesting. Like... Okay. So what I think has happened is that researchers have done an amazing job. And they basically created this brand new set of capabilities, raw capabilities that are out there to be used to build new things.
Speaker 3 And bridging the gap between... Okay, here's a new thing we can build, and here's the thing people actually really wanna use. Requires a kind of product knows and creativity and a level of tinkering and just messing around to see what works that is an important input for progress you know, the wright brothers ran a bicycle shop. Right? They weren't doing like fundamental physical research into lyft.
Speaker 3 They were tinker. And And so what's happened is more and more people have flowed into Ai is that they've tended to chase the existing status leader boards. Like, oh, the cool thing is to publish a paper on archive that gets lots of tweets or citations. And so my encouragement is to say, you don't actually need to do that. There's a capability overhang now, which entrepreneurs and product people can fill in with products.
Speaker 3 And In a way, you don't actually even need to know how the training works. Like, there's a there's a level of that these models have. Where I think the people who tinker with them the most in a way understand them better than the people who built them because they know what they're good at and how to interact with them and that sort of thing. And so yeah, I think it's time for product people to catch up, like entrepreneurs you know, have been sleeping on it for too long. Now they're all doing text to image.
Speaker 3 Which is cool. I just like more people to do language stuff, because I think language is sort of an Ag complete problem if you solve language, you could probably solve reasoning. There's a huge amount of value to be unlocked there.
Speaker 2 Yeah. Totally. I mean 1 of the things I'm kind of curious your thoughts on is like... These it What’s it's very weird, but we've just released just reached this point where these models are platforms. And they actually are like, you know, before they were always so brittle.
Speaker 2 Yeah. But you actually really had to understand the technology to really use them in the first. It really wasn't the case that you could just take it out the box and then play with it. Now you could take it out the box Yes. And and play with it.
Speaker 2 And Mean I'm curious, you know, you you probably interact with a lot of tinker? Like, you know, What’s what do you think makes... What are the qualities that make the best tinker curves?
Speaker 3 That's a good question. Well, okay. So what you have to do when you build a new product is you have to kind of find the intersection... Between what's a new thing we can build that's never been built before and what's the thing that people are gonna wanna to use like, almost every day that they're gonna love. Then if you're trying to build a startup, you have to, okay, what's the thing I can get distribution for too.
Speaker 3 So you're sort of trying to intersect those sets. So I think it requires, like a kind of... Truth seeking, curiosity and an understanding of what intuition, some intuition for what people actually want. So you need to, like, understand people a little bit, and d interested in tinkering with new things. There's a lot of people who can do this.
Speaker 3 So yeah. I think we're gonna see a huge amount of it. Yeah.
Speaker 2 You going hopping from the spicy topics. I think of the 1 of the really big questions right now is it's entirely relevant to C copilot, also the image generation models is this question of intellectual property. Right? And I think that 1 of the very tricky things about these about these models is that you know, let's talk about the image generation use case. If you look at most of the prompts that create really cool images with...
Speaker 2 For image generation, They have some artist name in it. Right. Oftentimes multiple artist name.
Speaker 3 Yeah.
Speaker 2 And you're just explicitly asking the algorithm to rip off their style. And it's it's really weird and the artist isn't getting compensated for that. And you could, you know, just as a as a as an example of something that could happen. Someone could make some Ai art using some artist name, sell that piece of Ai art for a million dollars Mh. And the original artist could still be putting stuff up on Devi art (Former for pennies.
Speaker 2 Right? Know. And so it's it's this really weird moment where the sort of there's like a new form of pla. I mean, Ai pla if you will. The same is true in code.
Speaker 2 You know, there's the same paradigm code and I'm sure there there's lots of examples where c copilot has an an allowed individuals to basically rip off code from some open source repository that they're don't have the license for. And so how do you think this... What’s what do you think we need to do? I mean, this is a really big problem.
Speaker 3 Yeah. Well, I mean, first, I think people have an intellectual property, you know, to the extent that there's intellectual property protections. I think it's important to sympathize with that artist? Who, you know, see someone literally invoking their name to produce something that as if they had produced it or if you're programmer and you're using, you a code generator and you it like seems to produce code that you wrote. Like that's like a visceral moment of feeling cheated somehow But I think there's also kind of like the basic question of what's happening here.
Speaker 3 And I don't know. I think 1 thing I remember well and you just had a Eric up here, but is when search engines got big. There were actually a huge draft of Ip lawsuits against the big search engines from people who felt that the act of creating a search engine, indexing my content storing it on your server, you know, allowing people to search it, snippet it, even caching the whole page and just having a cash link there, that felt like theft to some people. And their, you know, Google ended up funding what I'm sure was a very expensive set of lawsuits and defense, arguing that what they were doing was fair use and 1. And I think they won on the letter of the law.
Speaker 3 But I think they also won because people recognize recognized that this was the future, and it was valuable. And that in fact, Google by indexing these things was helping a lot of people, including, by the way, probably those content authors, themselves, you, who may have had some kind of reaction to it. A And so I do think kind of ultimately society decides these things based on whether it likes them or not. And if these Ai things you know, allow me to express myself which I've never been able to do or make me a lot more productive as a programmer and, I think that will inform the sort of overall Democratic governance, policy decisions that people make. The other thing I would say though is that the people who are reacting to this and haven't necessarily thought through all the implications of the positions they're taking.
Speaker 3 So if the position you take is, I put some... Code out there publicly and you statistically trained on it and your model became better as a result, you shouldn't be allowed to do that. Well, the inputs for these models are public data and Gpus. If you can't train on public data because people are saying that that's that should not be allowed, then who will be able to train these models well large companies will just spend a billion dollars and get private data. And so only large companies will be able to train models like this, is not really a great future you're arguing for.
Speaker 3 It's kind of a pro big company future. It's like an anti little guy kind of future. And so I think a lot of the reaction, you know, at least on the coding side has a lot to do with just big companies are, you know, rich and we don't like entities that are rich. As this diffuse and more people benefit from it, I think people will start to take new positions, On the just pure legal question, I think fair use is the backbone of this whole topic and In the U. S.
Speaker 3 You know machine learning statistical learning across datasets is various.
Speaker 2 Yeah. But to press on that, For example, like, if that let you know, obviously Out of the bottle. But let's say, Dolly 2 was the only way that people ever created you know, Ai generated imagery, then being you know, Open on their own could basically in some form solve this problem. They could sort of ensure that royalties were given to the artist. They could, you know, and enable artists to opt out of their name being used and prompts, all the stuff.
Speaker 2 And so in some sense, you know, maybe from a little guy from a company perspective, not the little guy from an. Hardest perspective, they actually... We could have had a a better outcome. Now now I think it's impossible, but I think that the are tough questions.
Speaker 3 Yeah. I think it's, you know, it it's going to be a public debate, I think it's going to come down to the overall benefits to society and the sense of fairness that emerges after this stuff's been available for a couple of years. And Yeah. You 1 thing I think about there is, like, 19 98, the Us passed, the digital millennium Copyright Act, and it was... In retrospect this unbelievably foresight enlightened moment of governance in the country where the Internet was starting to take off, And there is this idea that there would be platforms that would have user generated content on them and there should be some kind of legal framework and Congress and some forward thinking people in congress debated you know, as elected representatives and past the.
Speaker 3 And and basically, the extent that we have U platforms today it really rests on the provisions of the Dmc to a large extent. Ideally something like that would start to happen now. Where, you know, enlightened people that when Congress would have a debate and there'd be a law. I think the alternative is the judges, it just gets kicked to judges under existing legal frameworks, and judges or regulators. Yeah.
Speaker 3 Yeah. And I think What’s be more unpredictable.
Speaker 2 Yeah. And and I mean, you you can see the precedent and other... New technologies like at crypto. Right. Legislation is lagged and therefore, regulators are just sort of are just acting Without the without grounding a...
Speaker 2 Of legislation early.
Speaker 3 Yeah. When c copilot came out, we we basically we launched it and like May of last year and we've stupid whatever didn't have enough Gpus. So we only have, like, 10000 users. Was not a good move on my part. And the people who were using it loved it, but the people who were not using it, we where like, this looks like it writes bugs.
Speaker 3 Sometimes the code insecure, like, you know, whatever it doesn't look good. So all the sort of fu was coming from the people who didn't use it, but we knew, like, if you just used it, more often than not, you came away saying this is amazing. It's indispensable. And, like, actually, those things are not as big issues as they might appear at first. Didn't And so I was like we just have to get more people to use this.
Speaker 3 And so I went to the Azure team and I said I need more Gpus. And I said, well, we have this 1 block of Gpus, but it's a lot, and you have to... It's all or nothing and you have to decide today. Like 4008 1 hundreds or something. Tens of millions of dollars a year and I was like, okay.
Speaker 3 Right, we'll take them. Then we just opened the flood What’s, and it sort of worked, like, then the people who'd used it said, I actually use it every day, and they started to sort of be reply guys to the people who were complaining. So I think, you know, the interesting thing is Google was really broadly used. And I'm not sure how broadly is the stuff will be. Yeah.
Speaker 3 But I think, like, the faster there are great products and the more people those benefit. And the more people just have a tactile sense for the value and the contours of this. The better the debate will be and probably the better outcome we have.
Speaker 2 Yeah. As my final question to make fun of the venture capital community a little bit. And I'll actually exclude you because we literally were working on Ai grand back and I think 20 18 was the 20 17. 20 17. 20 17 was the first year And so you've obviously been passionate about this for a while.
Speaker 2 But we're obviously now every venture capitalist has taped into Ai as their top Second theme. And and I'm just curious if gal this. I mean the last time this happened was web 3 and and Yeah. I you know, I won't say how What’s necessarily worked out, but Yeah.
Speaker 3 I said I said the other Tweeted like a stable diffusion is Ellis Island for rep refugees from crypto seeking a better life. And I kinda of believe that. I mean, like, yeah, it's fun to make software that does stuff. You know. And I I don't have anything against Crypto.
Speaker 3 It seems to make some people happy, so I'm glad, but I you know, Thank god. You know, it's sort of like, okay, we had the web with the Internet and the web and then we had mobile... And we had... Cloud in there somewhere. And then there's been this question for a decade, which is like what's the next new platform wave.
Speaker 3 And there were these sort of (Former if you are at Ar, kind we had this massive sort of side show of crypto going on. It was very noisy. And kind of stole a lot of iq points from the rest of tech for a long time. And they owe us. They gotta give...
Speaker 3 We want that back. So I'm glad we have a new thing and it's an exciting thing, and I think it's gonna show up in people's lives in a big way.
Speaker 2 Do you think we're getting over hyped?
Speaker 3 Definitely. Yeah. I think we're definitely we're sort of it depends on your time scale. It's hard for the thing that no 1 is making fun of not to be over hype. Like, I would feel much more comfortable if people were making fun of Ai more and like, dividing it and saying it's a joke and underestimating it, and they just don't see that enough.
Speaker 3 So I think we're definitely over hypo it in the short run. But in the long run, like I think we're gonna solve reasoning. Gonna have computers that think. And you're gonna be able to type, you know, David Do gave me this 1, but it's like a well known formula me left for a room. Temperature semi super is, submit and like, get get the answer at some point.
Speaker 3 And temperature so I think it's gonna be absolutely huge, but I'm not sure quite what the timescale scale is of that. And in the short term, people are gonna lose a lot of money. Like, look, there's a lot of people who lost a lot of many in the Internet. They were right about the Internet, but they just invested in the wrong companies along the way. And that's probably gonna happen here too.
Speaker 2 Yeah. It's the the classic channel tunnel problem where there were tons of investors who invest in the channel tunnel. They lost a lot of money. Right.
Speaker 3 But bubbles are good because bubbles, they create this big bubble and it's this distributed search function where all of society in the free market is searching for good things. And and they're finding a lot of bad things along the way. But they're gonna find good things too, then the bubble collapses, but it leaves behind What’s sediment layer of progress that the next generation builds on. I think about this a lot. Like, the dot com bubble created web band, which was like, you know, sort of an Insta card before its time, and it failed because the Internet wasn't big enough.
Speaker 3 But then the executives who ran web van, some of them went often and started this robotic warehouse company they became Key systems that Amazon bought and then Amazon built tens or hundreds of thousands of these robots. And so in a way, the sediment entry layer left behind from Web van Powers, Amazon's Warehouse. And then some of those executives ended up running Amazon Fresh and subsequently bought Whole Foods. So Web van lost as an equity vehicle But for society, it produced all this progress in the form of this sort of... And that's gonna to happen here too.
Speaker 3 A lot of things that don't work will be tried. People who are sophisticated investors will lose money, doing that, but we'll all learn something and move forward.
Speaker 2 Yeah. Well, Great conversation. You heard it here first. Bubbles are good, and and thank you so much, That. Thanks Howard good.