entrylevel job loss. You made some pretty I think sharp comments on this conversation. How do you think about this >> job transformations? They are coming. >> Tech entrepreneur and co-founder of LinkedIn, >> co-founder of Inflection AI, a partner at Greylock. >> His new book is Super Agency. What could possibly go right with our AI future? >> Please welcome Reed Hoffman. >> It is definitely the case that AI will lead to a lot of different job transformation. in some cases flatout job loss but also I think we will adapt perfectly fine >> the career of the future is entrepreneurship it is how do you use these tools to create value in the world >> the entry-level job of 2 years from now will be very different than the entry-level job today I >> think if you look at the industrial revolution you know the the net effect is always more job creation in the long run problem is the timeline is so short it’s happening much faster than just the raw job displacement you would expect >> we all need to think much more entrepreneur And here are some lessons of entrepreneurship and here’s how to think about them.
[00:01:02] >> Now that’s a moonshot ladies and gentlemen. >> Everybody welcome to another episode of WTF just happened in tech. This is the news that’s important for you to learn if you want to change and transform your life, your company, your industry. This is the news that’s about hopefully an optimistic vision of the future. Not about sort of dystopian views. It’s about hopefully real views, things that you can use. I’m here today with my moonshot mates Dave Blondon, Alex Weezner Gross, and Seem Ismael and a special guest, a friend now for I don’t know probably at least 20 plus years, Reed Hoffman. You know Reed as the creator of LinkedIn, many companies on the board of Microsoft, outspoken about this tech field, and I’m excited to get Reed’s input on a lot of the topics we’re going to be discussing today. So without any further ado, first off, Salem uh just back from India. I missed you on the last two versions of this uh of this podcast and the question is did
[00:02:01] you solve the trade issues and didn’t you bring me back my iPhone 17? >> I did not bring I brought back some parts. So if you want to assemble it yourself, you can because you know uh that’s that’s somewhat flaky over there sometimes. Uh the two things that were blew my mind was the attitude of everybody in India was um literally middle finger to the USA and I think this is a big challenge because if India and the US India and China start trading the rest of the world kind of goes to hell and so uh it’s it’s kind of a big deal but what I found most incredible was the unbelievable optimism for AI and the use cases that are exploding there out of the gate and I think that’s amazingly exciting. And we’ll get to that in a little bit. Dave, I just saw you up at Stanford. We interviewed the CEO of Replet along with See, that was a fun conversation. >> Yeah. Yeah, it was fantastic. >> I’m still here actually. >> 150 million
[00:03:00] repositories. It’s incredible. >> Yeah. >> Yeah. No, the code’s piling up like crazy. Well, uh, that’ll come out soon, but Peter, uh, Peter used it on his plane, too. So, it shows you. >> Yeah, it was great. It was so fun. I I literally downloaded Replet. I had Starlink on the dash of my SR22 and I was flying connected and I I vibe coded a uh what was it? A mindset app on the way up there. It was great. >> Yeah. Tweeted it out. >> All right. Uh let’s drop in let’s drop in on the subject of jobs and education. Uh and Reed, I want to go to you first. uh our common friend Eric Benhoffson uh published a paper recently uh with these charts looking at entrylevel job loss um down 16% in AI exposed fields and and you made some pretty I think sharp comments on this conversation. How do you think about this? >> Well, a couple things. So um it is definitely the case that you know that AI will lead to a lot of different job transformation in some cases um you know
[00:04:03] kind of you know flatout job loss. I mean a simple heristic that I sometimes use it’s a partial heristic is if um the human being is trying to do a job by following a script that a you know an AI can follow better customer service etc. Uh that will happen. Um but obviously some of the issues here are around um you know kind of like questions of um not just customer service um but also like software engineering. Now that being said, even with a uh kind of a downbeat in, you know, possible initial software engineering job hires, my belief is that that um is only a transformation issue. Partially because um I actually think if anything the thinking about how you do software and software engineering is actually going to get a lot more widespread in problem solving. Um because you know part of what AI is going to lead to is a
[00:05:00] software co-pilot uh for all of us that that anything that you’re doing that involves thinking anything that you’re involved doing communication language will also involve as it were custom software that you’ll be doing. Um and so I think that there’s still nearly infinite uh infinite um you know hiring demand for software engineers. So it’s a little bit of a complicated thing. Eric is awesome and this work is awesome and so it’s but it’s um uh you know kind of job transformations they are coming. >> Yeah. You know, we look at this first chart here which looks at uh job losses if you would in marketing and sales and you can see the chat GPT inflection point in 2022 and the drop off that blue line for those you watching this on YouTube uh is in fact uh early career uh job hunters and and one of the biggest concerns I’ve got and Salem you can speak to this just having come back from India is getting into a Arab Spring-like
[00:06:00] situation where you’ve got a large population of youthful individuals who are you know on the male side testosterone driven without investing their time and energy to do something meaningful to create a career to be able to get in a position to have a family a lot of frustration and it could seed you know I’m usually the massive optimist but could seed the you know a civil unrest thoughts on that Slim >> so I just came back from India and what what they were seeing was initial um signal is that entry-level software jobs are down about 20 to 25%. Um which is quite a big number there and there are hordes of uh Indian um engineers coming out of the workplace. Now the there is a little concern around the social implications of that. My guidance to them was hey go become entrepreneurs. I mean this is like the literally the best time. Find a problem that you think needs to be solved and go transform yourself. And I think it’ll force it’ll
[00:07:02] be a forcing function on the positive side. >> Yeah. Uh Dave thoughts. >> Well, completely agree with what Reed was saying. I think if you look at the industrial revolution, you know, the the net effect is always more job creation in the long run. So he’s completely right. The problem is the timeline is so short. So, a lot of, you know, I think a lot of people didn’t anticipate that employers like Salesforce.com would cut off hiring in anticipation of AI that’ll be in the market a few months in the future. And so, it’s happening much faster than just the raw job displacement you would expect. And so, it’s really disproportionate on new graduates. So, like Salem was saying, it’s the perfect time to start a company. But, you know, historically, very few people graduating from college start companies. So the net effect, what I’m hoping happens is, you know, we we’ve been trying to meet with a governor in Massachusetts and talk about this and she’s super competitive, but her trying to get the legislature and people to move, it’s like pulling teeth. >> But when you have voters that don’t have
[00:08:00] jobs, then that creates some some acceleration and some motion. So I’m kind of hoping that the net effect of that is that people react a lot more quickly. Uh especially political people. >> Yeah. Reed, I want to read some of the comments you made uh and you posted on X. Uh you said more interesting puzzle is the drop in junior engineering roles, right? And then you said um people who understand computation will become more essential. Can you speak to both of those? Yeah, that was a little bit of what I was saying earlier which is um if we see anything in the last you know decades is that actually in fact the amplification of how computation affects all aspects of human society including human work are essentially just going up and that’s not just mobile not just internet um and obviously um AI is the you know kind of the exponential uh you know uh acceleration in this and I think that the question is is still thinking about like how do we put
[00:09:00] problems computation now to make that a little bit more tangible for people think about how much as you begin to get exposed to the current AI models how much your thinking pattern changes to more of a how do I start with the right kind of prompt to accelerate my analysis as a problem my thinking of this problem the research analysis etc etc and so almost like when I’m thinking about a new creative project a new research question in terms of like oh how might one you know do this business problem like a go to market or something else is I think about like how do I put it in terms of a more detailed prompt now part of the reason that’s computational thinking is that one of the things that relatively few people um do that you should do is is is most of my prompts are uh in involved doing the deep thinking or deep research prompts but my first prompt is give me the deep research prompt that will you know solve
[00:10:01] these kind or target these kinds of things and then so I write in a paragraph or speak in a paragraph then it comes back with a page and a half then I edit it and then I submit it and that’s the prompt that I’m beginning to drive and work off of and that’s an that’s that’s an instance of how computational thinking we’re going and I think this is going to become you you no longer have individual contributors in companies etc that we all deploy with a suite of agents And this is just that lens into that. >> I love that. I did a little research, got some numbers want to share with with you, Reed, and also get Alex’s point of view on this. So today there’s 150 million users on GitHub as of May of this year, which is extraordinary. Um, and if you look at the growth in the number of software engineers, number of programmers since 2022, it’s up 50%. So 50% more uh programmers in the world. At the same time, what we’ve seen is not a
[00:11:00] decrease in salary. In other words, it’s not a overglut where competition is bringing down salary. It’s actually been a 24% increase in salaries over a 5-year period. And so what does that tell us? Uh that increasing productivity, increasing demand. So look, I do think I tend to look I I I I would I would I would tend to want to think increasing productivity. Um I do think that this is very early days in >> how all this is working and I tend to not try to get distracted over much by numbers this this month or this quarter that have technological underpinnings in terms of the theory of what’s going on. Now, um I can tell from my own work that I know that we have increasing productivity because I know what used to take me a couple hours sometimes now takes me 10 minutes or 15 minutes um in terms of getting into something and and so you once you know that you know that
[00:12:01] whatever the you know what was the old line on computers and the economy the computers are everywhere except in the numbers um like even if you said hey well I don’t have a GDP number it’s like well but I know those productivity increases in this case. And so it’s so that’s the reason why I’m a little cautious about overreading into specific numbers. Um, and I tend to more generalize from what I can actually see in in kind of as it were workflows, not just my own, but other people I talk to and seeing what happening in companies and, you know, how most startups these days are, you know, completely AI native in terms of how they’re operating and so they’re finding great accelerations in it, you know, that kind of thing. >> Alex, how do you how do you think about this, Alex? Oh, I I think we’re we’re in the earliest innings of AI automating the service economy. I I think it’s very instructive if if you read Eric’s paper that these results were most striking in fields where AI was automating rather than augmenting human labor. So I I
[00:13:00] think this is completely consistent with call it the the hypothesis that humans and machines are in the not too distant future going to merge symbiotically and this is just sort of small potatoes. is the the earliest possible trickle of what’s ultimately going to turn into a flood to to Reed’s point of productivity gains and giving humanity the opportunity to to chase much more ambitious problems than what here are being characterized as entrylevel jobs. I think with the benefit of hindsight, 10 20 years from now, we’ll look back and and we’ll be horrified that so much of the economy was bound in in what are here being characterized as entry level jobs rather than more ambitious, more fulfilling endeavors. >> Dave, you were going to jump in. >> I was wondering if Reed, you know, your productivity, as you mentioned, is way up. Mine is way up, but I could use a lot more agents than I have access to. I was wondering as a board member of Microsoft if you get like 20 or 40 dedicated GPUs and special access and
[00:14:02] and uh cuz God knows you can use them like as soon as you’re hooked you’re hooked and then you just want more and more and more. >> Um I don’t um that’s a good idea. I should ask. Um I I just simply did ask for me too. >> Yes. I simply do the Mac subscription across all of them. And frequently when I’m doing something like I’ve actually already um put on a >> um a kind of um you know kind of running a a a and now it’s the open AI open source model on my laptop to front end to uh parsing it out to multiple agents then you know like you know run it on chat GBT run it on copilot run it on Gemini run it on cloud and then integrating what comes back um on anything that’s kind of more substantive. So, um, so I’ I’ve got the personal hack, but not the the the personal cloud. I >> I like this in the future of offer letters. You know, here’s your salary, here’s your bonus, here’s the number of GPUs you get, the number of agents you have. >> Oh, no doubt. No doubt. The GPUs are so much more important than than the other
[00:15:01] components of a comp plan. If you know, if you have any ideas, you know, if you have any ideas at all, the GPUs, you just use them up at at as fast as they can print them. You will suck them up. >> How how many companies do you have in incubation right now? Are you are you in startup mode across across multitude? >> Uh well, you know, it’s one of those things. Um you know, I’ve got uh two uh co-founded companies, Inflection and Manis. um you know kind of one one play on what is the essential role for how we have companion agents uh that go throughout our whole life with us and another one uh accelerating drug discovery becoming a drug discovery factory with a target of cancering of curing cancer >> um with Sedara Mukerji uh and then I’m got another thing that I’m um I’m in the ideation phase on >> have to be right that’s the fun that’s the fun part and you know I remember you introduced me to Mustafa when you were uh working with him and now he’s in uh Microsoft heading their A activities. He
[00:16:00] must be proud of of that transition for him. He put out a paper recently basically warning people to be careful to think of AIs as conscious as uh as living entities. I’m assuming you read the paper. >> Absolutely. >> And what did you think of it? >> Um I thought it was exactly right. I think the um you know and there’s a you know it’ll be interesting to see you know if some of the better philosophers also kind of engage in this. I mean the the the challenge is is that historically we’ve been able to to pretty easily map between can you speak language and are you conscious and um and it isn’t that it isn’t a very complicated question which Mustafa would agree as to well when is when is it that you are conscious and I think Mustafa was at Google when there was some engineer who said well I asked if it was conscious and it said it was so therefore it must be and it’s like okay that’s let’s not be that quite that simplistic. Uh and um but the but the
[00:17:01] notions of you know kind of like self-awareness, self-reflection um the notions that would come up you know not as kind of a simple like 30 minute touring test but also this kind of question around like how we learn of others others minds and other consciousness by how we navigate the world together. how we we communicate not just by sitting in a you know kind of between behind two terminals all the the you know the touring uh imagination um the you know and so I think it’s it’s exactly right to not jump to it too quickly because you know we as human beings also have this weird thing of both over and undercribing consciousness um overcribing consciousness like you know your car you know come on Georgia you can do it um to under ascribing consciousness like well you know these animals they’re not conscious like well it’s a little complicated look at how they’re navigating the world etc and
[00:18:00] look at how we’re doing it like like what the shape of their consciousness is versus the shape of our consciousness is probably the more interesting question so anyway but basically it was a very good kind of warning shot because what happens is people have the language experience and then go well I asked it was conscious and it said it was >> I always go back to I always go back to the commentary from one of our NASA astronauts at Singularity when we asked him because he’s seen a ton of animals in labs and we said what what level of complexity does self-awareness emerge and he goes oh frog and we’re like what and he said well in his opinion watching hundreds of animals in free floating experiments and a mosquito doesn’t know it’s a mosquito right a dog definitely knows it’s a dog and frog is about where the boundary condition was in his opinion where it goes oh I think I’m a I think I can see I’m a frog and then above that much more >> I think my thinks it’s a human. >> But what’s weird to me is whenever I’m interacting with AI at home, uh I’m always really polite to it because it’s polite to me. But you know, it’ll do
[00:19:00] something and I’ll go, “Okay, that’s really cool. Now, what I want you and my wife is like, “What what are you doing? Why are you talking to it that way?” It really freaks her out. And like, well, look, I’ve been building these things since I was 16 years old. I I know it’s not conscious more than as much as anyone, but it’s just your natural reaction is to treat it the way it treats you. I don’t know. It keeps it fun from my point of view, but she she thinks that’s creepy. >> Alex, are you ascribing consciousness or now or in the future? >> I’m probably at the far end of of this discussion. In the past, >> that’s why I called you. It it uh if you remember Peter, uh people for the ethical treatment of reinforcement learners, huge supporter of that. uh the non-human rights project, which is fighting for legal personhood for non-human animals, starting with elephants, dolphins, and great apes. I I I think we’re I I think we’re on the verge of having the personhood discussion for non-human animals, for pure AIs, probably for for some new
[00:20:00] exotic forms of intelligence like borg organisms, collective intelligences, uh and >> organisms. I love that. >> Borgganisms. >> Yes, >> it’s a very important class. And I I think probably, you know, we we talk about prediction markets sometimes, collective intelligences. I I suspect we’re on the verge of having a discussion where there will be halfozen different new categories of intelligence. To to Reed’s point, they won’t all necessarily have the same shape as natural persons, but they will nonetheless be intelligences that will perhaps be deserving of uh of their own rights. And I I think not just personhood rights that are or civil rights as it were that we normally discuss, but economic rights and communication rights. And and at at some point we start to think about what does an economy a heterogeneous economy of lots of different intelligent actors of different types even look like? I think it’s going to be a very exciting jungle. >> And continuity writes, can I tell you a quick secret? Um uh don’t tell anybody
[00:21:02] uh but uh uh working on an ex-priseze with Palmer Lucky right now called an interspecies communications prize to use AI to be able to communicate birectionally with a number of species right to be able to understand what they’re feeling what they’re saying being able to actually have some level of a dialogue a lot of work that’s been done we hope to step it up another level and that will be fascinating >> there was a project a few years ago they were trying to use machine learning to translate dolphin language Yeah. >> And my response was, I’m not sure we want to know what they have to say. That’s a whole other kind of can of worms. >> We we absolutely do. So, I advise a company named Sarama that’s working on this for uh for non-human animals starting with dogs. I think we absolutely want to interact economically, socially with non-human animals. Daniellea Rus is doing it with whales actually and they got incredible it’s online actually got incredible footage of a of a humpback whale birth which had never been filmed before but they got all the audio and so they’ve
[00:22:00] got sounds that have never been recorded before rel because they’re very social and they all get together for child birth and the the the whale baby needs to be elevated to the surface by a whole top pod or team or whatever. >> Yeah. So it’s it’s really cool. >> Well, and by the way I I would love to get the that company name. Um, I helped stand up this thing called the Earth Species Project, which is maybe what Sem’s referring to. Um, because it’s not just dolphins and whales, but also corvids, uh, and primates, and it’s basically, you know, record as much as you can on both the the sounds and the environment, and then run it through, you know, ML translation and see what you get. >> Reed, do you mind if I reach out to you on uh on this X-P prize we’re getting ready for? >> Oh, of course. >> Yeah. Um, and do you want to give them that name of that company again? Earth Species Project. Oh, sorry. But you the name company. >> Sarama. S A R A M A. >> I’ll connect you. Reed. >> Every week I study the 10 major tech meta trends that will transform industries over the decade ahead. I
[00:23:01] cover trends ranging from humanoid robots, AGI, quantum computing, transport, energy, longevity, and more. No fluff, only the important stuff that matters that impacts our lives and our careers. If you want me to share these with you, I write a newsletter twice a week, sending it out as a short two-minute read via email. And if you want to discover the most important meta trends 10 years before anyone else, these reports are for you. Readers include founders and CEOs from the world’s most disruptive companies and entrepreneurs building the world’s most disruptive companies. It’s not for you if you don’t want to be informed of what’s coming, why it matters, and how you can benefit from it. To subscribe for free, go to dmmanis.com/metatrends. That’s dmandis.com/metatrends to gain access to trends 10 plus years before anyone else. So the next article here, US students reading and math scores at historic lows. 35% of 12th
[00:24:00] graders are at or above proficiency. 35% at or above proficiency, down from 40% 92. only two 22% of seniors are proficient in math with science at 31%. Uh this is dismal. Um and I’m assuming this is you know this is the United States. Uh it’s not the same in other parts of the world. Uh you know what are your thoughts about AI accelerating this or helping solve this right? It’s it’s it’s double-edged sword. A lot of people are not doing the work. they’re going to to you know chat GPT to get their answer and they’re defaulting to not thinking. On the other hand, AI will be the best educator on the planet. Reed, where do you come out on this? >> Well, I’m ultimately not surprising extremely positive, optimistic. I do think there’s some transition issues um which you know we may see with the you know handing in you know AI done homework but you know
[00:25:00] if you just make this thought um one within a small number of years all assessment will be essentially be done by AI so like we’ll have the equivalent of being able to do PhD oral level defense and then on down um just with AI doing it and so therefore your level of cognitive preparation um can be set at whatever the benchmark could be set however we like and the people have to prepare for it. So I think that everything else makes that is an interim. The second thing is a little bit like open AI’s learning mode. >> Basically if you just put in a metaprompt to the AI agents today and say work me towards the answer don’t give me the answer. You already have the most amazing tutor that’s existed in human history >> for free. >> For free >> and global. Yes. >> Soing. So it’s like yes, these are serious problems and it’s simple to work in a massively positive
[00:26:01] direction. >> Well, I think that the tutor analogy too is is so much less than what it actually does because, you know, it goes in any direction you’re passionate about. you know, a tutor will teach you math or whatever within that curriculum. The AI version of it goes wherever you want to go. It’s it’s just beyond a a tutor by orders of mag many orders of magnitude. >> There was a there was a statistic I’ve been using that um a child with an AI is learning between two to six times faster than sitting in a classroom. And I was talking to somebody at that Stanford AI conference the other day, Rita, I think you dialed in for that. Um and uh they said you’re out of date. It’s actually 5 to 10 times faster. >> So easy to believe. >> I I also think if if you think that we’re on some sort of exponential or hyper exponential progress law where we’re about to have full high bandwidth BCIs, brain computer interfaces in 5 to 10 years. It’s a little bit difficult to
[00:27:01] get too worked up by a blip on a few scores for a few years if if you think we’re just going to be able to sideloadad knowledge into minds 5 to 10 years from now >> from your to the laws of physics years. Right. >> So, by the way, Alex, after you say that, you have to say, “Now I know kung fu.” Just to be clear, >> now I know kung fu and demonstrate it. >> Yes. >> Yeah. Yeah, but I wonder to what degree these statistics are driven by the fact that that people have so many other ways to spend time and to learn things and to do things and because because I know for a fact that once you’re into AI, my kids, they get so frustrated by the school curriculum. >> Yeah. >> Because because one, they could learn it much faster anyway, but two, they want to learn other things. Like this curriculum seems ridiculously narrow and stupid compared to everything they can learn, and they’re passionate about something else. >> I do want to point out something. this this kind of statistic based on the existing curriculum. It’s the same commentary applies to Eric’s report on the jobs. We’re looking at the jobs in a static way the way they are today. The
[00:28:01] entry-level job of 2 years from now will be very different than the entry- level job today. Right? So that jobs will transform along with it. The education may not transform as fast because of the regulatory structure, but definitely will be changing the game as it goes along. And I think the conversation we’ve had on this pod for some time now. And I hope people have heard it, especially if you’re in, you know, high school or college or is the career of the future is entrepreneurship. It’s not going through a factory process of getting a job for someone else. It is how do you use these tools to create value in the world. Um Reed, I mean, we’ve been on this for a while. Entrepreneurship is the future. Uh how do you think about that? Well, um, you know, as you know, because we’ve talked about this for decades. Um, my very first book was a startup of you because basically we all have to think more and more like entrepreneurs even if >> we are are the entrepreneur founder creating a business or not. It’s the nature of the world we’re we’re evolving
[00:29:02] into. And it came from the commencement speech I gave my high school. But uh and it’s part of the reason why so much of like you know there’s probably you know two main streams of the kind of content that I produce um and one of them is around kind of technology and society and the other one’s around entrepreneurship and part of the entrepreneurship is not just for you know which is obviously great if it’s used by you know kind of blit scaling um you know high growth uh Silicon Valley and other entrepreneurs but it’s also we all need to think much more entrepreneurally and here are some lessons of entrepreneurship and here’s how to think about them even for an individual and you know a career of jobs. >> Just just just in that vein I just want to point out I got a copy of Reed’s book which is which actually says uh this is the Selma edition at the bottom. So it’s customized that and it has a photograph of me in punk rock uh stuff. So it’s customized to the actually to the
[00:30:01] individual. I thought this was unbelievably clever. I think your commentary on the fact that AI gives us all super agency is a really profound one. Everybody needs to digest the implications of that because it’s so huge. If people kind of look at that across the board, it’ll uplift the whole of humanity very fast. >> Amazing. Um and congratulations on that book. Uh I want to switch to a conversation uh led here by Jeffrey Hinton. So we’ve been talking about AGI. uh a lot of us on this pod have had the conversation saying uh we we you know the touring test came and went you know that was nice and a few of us believe AGI is here has been here and the real conversation is around digital super intelligence all right let’s go to uh professor hint >> thought is that a super intelligent AI is unlike anything we’ve ever seen is very very different from just a new machine that does something more efficiently I mean people used to make clothes by hand and then they made clothes with machines and there’s
[00:31:00] massive unemployment, but then eventually they got jobs doing other things. But super intelligent things are going to take away nearly all the jobs. And the idea that there’s going to be jobs that are still okay when you have super intelligent AI is quite dubious. I think the job of an interviewer, for example, will disappear too. Super intelligent AI will be able to do a better job of interviewing me. So I sort of completely disagree with Y. So, uh, couple of subjects to jump into there. Uh, and I’ll start with you, Reed. Uh, what are your thoughts on on ASI or digital super intelligence? And on the back of that, however we define it, and See, I’m cutting to the chase here. I know the question you’re going to say, how the hell do we define ASI in the first place? Uh, let’s just say it’s like, you know, a millionfold more intelligent the average human. You know, no ceiling on this. uh does it destroy all jobs and then where do we get our purpose from? These are the
[00:32:01] conversations that we’ve been having. It’s what my next book is about. Reed thoughts. Well, look to start with the circumstance of say we get to a Star Trek universe where you know kind of all work and physical materialism can you know phys physical material you know goods and services can be provided by you know kind of intelligent infrastructure >> and that’s the universe we’re living in. >> I think we will adapt perfectly fine. Um, you know, you we have a proxy for it in human history, which is, you know, medieval times. That’s essentially how the nobility lived where everyone else was the was the surf and peasant and middle class and whether or not infrastructure. So, we’ll have dinner parties and and theatrical performances and hobbies and all the rest of the stuff. So, I think like overly worrying about this I think is is a mistake. Now the problem with and this gets the why sim always defines like what it is and
[00:33:01] even going to a million times it’s like well you know what kind of shape of super intelligence is because if you look at the progression of GPTs they’re progression of savants and um and so if you get to the massively in incredible savvant that still has context awareness problems and other kinds of things that is a different shape in terms of what happens than if you simply have you know something like the Ian Banks culture series which is you know super intelligent robots that kind of look at us as kind of fun companions in the in the space journey of life. Um and you know I think it’s a um it’s very easy to be science fiction alarmist. It’s very easy to be um I al I don’t mean to be either even science fiction just you know benol op you know um um you know kind of optimism but but it’s kind of like there’s a lot of different things where the details matter and so kind of navigating what are the pieces we should
[00:34:00] be constructing right now in a range of different probabilities of outcomes is I think where the the intelligent discussion is. Well, no, it always surprises me how how like Jeffrey Hinton is is a god to me because he wrote the uh the Rumlton Hinton paper 1986, the back propagation that kicked off all the AI that we’re we’re experiencing right now comes from that that invention in ’ 86 and and it eliminated all the other forms of AI, symbolic and Marvin Minsky and all that other stuff. Um so he he is just an epic god and he’s you know he’s worried sick. You can see his furrowed brow in that video. He’s he’s just worried sick. Uh and and at the end of the video he’s like, “And I completely disagree with Yan Lun who’s another legend of the field, you know, inventor of convolution.” And and then we had David Seagull on our stage, you know, here at Imagination and Action the day before yesterday. And David is also, you know, he was at the AI lab at MIT the same time I was as a PhD student. And he’s, you know, he’s on the Forbes 400 quant trader using AI and he’s got a completely different opinion about the
[00:35:00] timeline to strong AI. So, it shows you how difficult it is to predict what’s going to happen next when you’ve got great great minds like that vocally disagreeing with that with each other in the media. So, >> yeah, we’re holding two different futures in superp position right now and we’re going to see uh how we collapse the wave function. Alex, >> yeah, I I think this this the sort of moral panic again very difficult to get too excited o over this. I I think I I if if you think as I do that we’re on the verge of having evenly distributed super intelligence and that evenly distributed super intelligence is going to solve substantially all open problems in math, science, and engineering that’s going to create so many different opportunities throughout the economy. Uh sort of worrying too much about the state of jobs and the state of careers as they’re currently parochially constructed. circa 2025 is going to look hopelessly naive and quaint in a few
[00:36:01] years. >> Yeah. Our our basic call to action is uh solve everything. We’re on the verge of solving everything. But here’s a question for you. You know, we we’ve had the conversation uh with biology that AGI is polytheistic, not monotheistic. Reed, I don’t know if you saw that that he put out. And all of these frontier models are sort of leapfrogging each other. And it’s been pretty impressive to see how they’ve been moving in lock step. But the question is does is there a winner or take all ASI, right? Is once you reach this whatever fundamental breakthroughs are required uh does the first ASI block all others? Is it a is it a hard takeoff? Reed, do you have a thought on that? >> At at Microsoft, you have to add at Microsoft. >> Indeed. Um, look, so this is again why it gets to like is can I sketch a universe where there’s an ASI takeoff that gets to a compounding curve andor
[00:37:00] operates to to um, you know, to prevent other AI? Yep. Film at 11. Uh, I can tell that story. Um, but the, you know, I can also e equally tell a lot of other stories including the fact that it is pantheistic. By the way, one footnote that I think is interesting is if you have different cultures responses to the possibility of super intelligence, um those that are inherently monotheistic uh generally express broadly fear and those that are pantheistic broadly uh express excitement um because it’s kind of like the one god versus many gods as a as a as an approach. And so, um, the I think that the I think it’s much more likely when you look at the pattern over the last couple years that it will be more of like kind of classic human invention, which is whatever it is will be a um a um kind of a uh kind of a zeitgeisty
[00:38:04] simultaneous um um across a set of different so therefore the pantheistic. But you know I can tell both both stories. >> I I want to make two points here. One is um though you know being in India where there the it’s incredible to see the excitement around this just because there tends the pantheistic. So that really speaks to the comment that Reed made in and the other commentary. I want to drill a bit more on what Alex just said. Once you do have super intelligence, however it happens and it’s solving huge numbers of problems, you essentially uplift all of humanity and and now you’re in a you know this is the very definition of a singularity, right? We have an event horizon that we cannot see beyond and it’s going to happen very quickly and when it does happen you I I fall back to the simple observation made by Ray Kerszswwell that technology is a major driver of progress in the world. might be the only major driver of progress in the world and now
[00:39:00] we have an kind of electricity type of underlying layer that’s lifting everything. This is unbelievably positive and the framing of it should be unbelievably positive. >> It is unbelievably possible positive. The question is, and the challenge is that we as a species, we strive and we thrive when we’re challenged. When we have problems that we meet, right, the video game that’s super easy, you you get bored. Um, and you don’t play it. The video game that’s extraordinarily hard, you give up. So, my question ultimately is, and there was some incredible work done at New York University back in the 60s called the Universe 25 experiment. Reed, have you heard of that experiment? I don’t think so. >> So, there was a social socio biologist who basically built this experiment 25 times. It was a a massive uh resort for rats, let’s call it that way. There was there was no uh there was no shortage of food, no shortage of uh of nesting space. They had everything they could
[00:40:00] possibly want. They put four breeding pairs in there. Exponential growth. And at some point the population starts to basically go upside down. You’ve got uh you’ve got still births. You’ve got, you know, sort of uh rats fighting each other. You have rats sort of marginalizing themselves uh just licking their fur and doing nothing. And the population basically dies not from having, you know, a shortage of any resources, but of having everything and and not being challenged. So >> this this is an extension of the wall scenario. >> Yeah. So it’s, you know, for me it’s like we need a Star Trek future uh not a M Mad Max or Aali you know we we need if we’re given this level of uh super capability in terms of AI and robotics and nanotechnology and BCI what do we do with it that challenges us that gets us thinking on a cosmic scale? I think that’s critically important and and Alex you and I have talked about that before >> to totally and in fact speaking of
[00:41:00] cosmic scale if I could put a physicist hat on for a minute and go back to the the question I I think Peter you and Reed were were talking about which is do we find ourselves do we think it’s more likely that we live in a near future with a singleton super intelligence or more of a multi-polar world um I I would point out we’re several generations our star our son is several generations old in terms of stellar evolution. And the the singleton that I would worry about isn’t is one particular frontier lab going to be the the first to achieve recursive self-improvement and then dominate the future light cone. We’re actually we’re pretty far into the history of the universe. I I worry about some other civilization not of our world that that developed a singleton and and now is is seeking to exclude Earth’s development of super intelligence. for me is knocking on your door, buddy. >> Well, I would say that the the fact that thus far, to my knowledge, we haven’t seen any evidence that the frontier labs
[00:42:00] are being bombarded by orbital lasers uh by efforts to exterminate uh Earth’s development of super intelligence would seem anthropically lowercase A, not capital A, to to point us in the direction of a multipolar super intelligence world, not a singleton. >> Fascinating. Reed, any closing thoughts on that topic? >> Well, I think um a little bit of also in these questions is what is the world we should want? >> What is the and and I think actually multipolar, you know, kind of pantheistic and by the way in terms of the you know, your rat experiment or the rat experiment um you know, one of the benefits is we human beings tend to prevent ch present challenges to each other. So, I’m I’m actually not that worried that um that we won’t have ongoing challenges because, you know, we compete um you know, whether it’s, you know, in in in things that, you know, we could be better at than anyone else or also just you know, like today there’s more people
[00:43:01] watching human beings playing chess than there have been at any point in history. Um and you know, that’s of course, you know, human beings are never going to be the never going to beat AIs anymore in chess. haven’t haven’t for many years. >> Fascinating. >> You know, I think we end up with as we evolve this and you look at chess is a great example of this. We watch people on a soccer field or on a chess board. We we watch for the humanity. I mean, can they make it in that really tense moment? Can they see the right move? Can they make the pass at a very critical juncture? Can you hit that tennis shot when all the pressure is against you? We live for watching that type of stuff. So I think uh to Reed’s point as we progress humanity you know and we’ve looked at uh cultures that have gotten to abundance the Moguls taking over India the Romans taking over Europe you end up in four activities that human beings do which is food art music and sex um not in that order and so you end up challenging each other in different ways and we’ll continue to invent those
[00:44:01] in more sophisticated ways. >> I love that. All right let’s get into the AI wars here. Um and uh here’s our first uh Senator Cruz proposed bill to ease regulatory burden on AI companies. The proposal creates AI regulatory sandbox to speed innovation. Companies could get temporary waiverss from HIPPA, FDA, and other agencies. All right. What do we think about this? Um it’s, you know, the government’s pulling out all the stops. It’s bringing capital from the Middle East. It’s relaxing the rules here. It’s changing the energy equation. Not as fast as we’re seeing in China, but you know, take off the gloves on drill, baby, drill, nuke, baby, nuke. Um, who wants to go first? >> I I love it and it’s desperately needed and we’re going to need a lot more of it. But then I read the details and it’s like, apply here, get a waiver. Like, oh god, it just just sounded so bureaucratic right out of the gate. But it’s well meaning, you know, at least
[00:45:00] it’s a step in the right direction. Well, AI should evaluate your AI should write your application and then AI will evaluate your application. Yes. >> And AI should just say yes from the beginning. >> In which case, the whole process should be less than 10 seconds. Right. >> Or or instantaneous. Right. >> Or instantaneous. Right. Right. >> We’re we’re advisers to a project called Fermy America, which is the largest energy generation project in the world, like 12 gawatt. And they filed their S11, and instead of taking two years, they did in a few weeks using AI. Nice. >> I would add if if you think we’re on the verge of an explosion of math, science, and engineering discoveries that are generated by super intelligence, then it also follows I I think that we’re about to have a glut of discoveries that are present governance mechanisms, including Reed to to to the point of of your start of Manis, they we don’t necessarily know how to metabolize all of these discoveries. if if we have a thousand cures that are developed by AI overnight, how do we get those through
[00:46:00] clinical trials and and get them out deployed to for the public benefit? And so to that extent, I I think in the abstract sandboxes and special economic zones and and other ways to to to basically offer new new platforms for modifying governance mechanisms to to metabolize that glut of inventions and discoveries are probably super net helpful in the long term. Well, I’ll tell you in foundations of AI ventures class at MIT, uh maybe a third of all the business plans that come out of that class are something health related and they’re really really good ideas and they all end up concluding they need to go to India >> to to get started and they’ll come back to the US later because the FDA is so small >> just like Zipline got started in Africa and came back to the US. I I think we’re going to see a huge amount of that geographic arbitrage just because it’ll it’ll be and this is where we talk about innovation on the edge, right? You don’t ever want to do innovation in the core organization. You want to do it at the edge and put pointed into adjacent areas. I think we’ll do the same on this
[00:47:02] side of things where we can set up sandboxes at the edges of cities or countries, whatever. Go do it in a safe place and then when it’s working, you can demonstrate that then come back into the mother ship. >> Read a closing thought on this one. Well, I do think that it’s absolutely critical to be imagining to be seeing what we can get. And so, for example, um ability to like like the simplest one that I go to in this is we should create clear safe harbor mechanisms for creating a 24/7 medical assistant that runs on every smartphone because the benefit for that is huge, >> massive. And you know obviously plaintiff attorneys you know other kinds of things will try to to to redact this as part of the reason why like people think oh we get to overregulation only because u the government regulatory agencies have a natural bureaucratic accretion but a lot of it’s actually like liability law from you know plaintive attorney associations and so
[00:48:00] forth and you need to actually in fact sandbox that in a way to get that and I think that then can be used as an example across the whole thing. Those are the kinds of things that I would pay um much more attention to in this and I think it’d be good to do like on the energy side. I’d like to see the energy stuff happen. That was, you know, kind of promised. It’s really important. Energy is going to be a really uh key part of this, but um so far all I’ve seen is a lot of tweets and relatively little action. You know, Reed, I I’m so glad you said that because one one of the great advantages of America is we have 50 distinct states and you have 50 different ideas and you have opportunities to try things and that variety should be a great strength for us. But what actually happens in practice if you if you launch an app, it’s like a medical app. It naturally goes out to all 50 states and then you always get sued in East Texas, which is Ted Cruz’s territory by the way. should just like look that whole tort law world is so messed up because it’s 50 different shots at you which means just by random chance some some really weird
[00:49:00] jurisdiction is going to come after you. I’m sure this happened at LinkedIn so you’re probably very aware of this but it’s it’s a horrible it completely backfires versus what the intent of the design was in the constitution. So we can get that fixed. >> Every day I get the strangest compliment. Someone will stop me and say, “Peter, you have such nice skin. Honestly, I never thought I’d hear that from anyone. And honestly, I can’t take the full credit. All I do is use something called OneSkin OS1 twice a day every day. The company is built by four brilliant PhD women who’ve identified a peptide that effectively reverses the age of your skin. I love it. And again, I use this twice a day, every day. You can go to onskin.co and write peter at checkout for a discount on the same product I use. That’s oneskin.co co and use the code Peter at checkout. All right, back to the episode. Jumping back to India, OpenAI plans India data center in major Stargate expansion, planning a 1 gawatt, accounting for 22% of India’s
[00:50:00] entire data center capacity by 2030. It’s part of OpenAI’s $500 billion Stargate project. So the fascinating thing here is again OpenAI planting its flag uh in different regions around the world trying to get early users captured. Uh how do you think about this Reed? Well, actually I suspect it’s less a user uh grab thing, although I you know that’s totally possible than for OpenAI and range of business and it’s more um uh looking for OpenAI is very clear cleareyed about um scale um is the thing that’s creating a huge amount of this of this potential and opportunity. Uh scale needs scale compute and scale energy. uh and so where can you get that and wherever can work on getting a deal that that works within kind of the western ecosystem uh it will do that and I think that’s how to to how to interpret this
[00:51:01] >> and you know this that was a little bit my like my earlier comment which is we are so behind on doing all the energy stuff >> um here massively and the real isn’t really acceleration >> you know back um you know the Empire administ ministration. I was kind of trying to circulate plans about doing deals with Canada to try to make this work um from a kind of North America and US perspective. But of course, since the current administration is, you know, trying to uh I’ve never seen the Canadian so pissed off with us uh in my entire life um you know that becomes less of an option. >> Yeah. >> I have a Canadian passport. So enough said there. >> Yeah. Selene, was this discussed while you were in India? It was but it’s mostly seen as a marketing tactic kind of kind of to show a planting a flag. This is going to take a while to roll out and India has quite significant infrastructure challenges to do this in a in a kind of reliable way.
[00:52:00] But I think the general trend is huge and I think what I see there is uh opening eye looking at the youth of India and making a planting a major flag saying let’s make sure we’re completely accessible to all the young people in India and by planting a data center there then you solve for a lot of the data sovereignty issues that lots of people are concerned about. >> Yeah. Uh this is critical. They have a very uh you know literate uh tech forward youth that they need to engage. Um on this on the note of uh making geographic grabs you know India is a 1.41 billion people you know going to go just under that at 10 million in 10 million in in Greece. Uh so open AI and the Greek government launched open AI uh for Greece. Uh congrats here to Prime Minister Mitsunis Mitzutakis and Basili Kumas Kumbas uh the digital AI. I listen I know him and I’m I’m just messing it up.
[00:53:01] But uh I I love the fact that we’re starting to see country after country begin to think about what is their AI strategy and beginning to uh to partner on this. So um you know I think we saw OpenAI going into the UK as well uh and of course going into the UAE. Uh do we see Microsoft doing any of this Reed? Well, Microsoft, you know, kind of the original tech hyperscaler has, you know, one of the things that’s been kind of amazing about being on the board there is, you know, it it has a uh an international scope of, you know, kind of relationships with multiple industries, multiple governments, multiple countries kind of around the world. And so um I I simply I’ve lost track of all the things that they are >> of that they are doing because it just it’s it is a you know it is a UN in in
[00:54:00] scope uh in terms of these things although obviously a lot more efficient because it focuses on kind of good business process and partnership and and all the rest. So I think that the and and this is I think the kind of natural thing to do. I mean it’s one of the things that I think is you said what should our AI foreign policy be? It’s let’s provision medical assistance. Let’s provision and I agree with Dave’s point about tutors. I mean like tutor is just to make people understand it but the fact that it can condition for learning for you wherever you want to go and in the metaphor and language that you want to use and all the rest like that’s the kind of thing we should be doing. So I think this is awesome. Um, well done by the Greeks. Well done by OpenAI and we should see a lot more of it. >> Can I ask you a question, Reed, about uh going, you know, OpenAI going to India for power makes no sense whatsoever and they’re short on power and it’s all coal. >> Well, you know, India’s doubled their their power generation, uh, as the US has remained flat. So, India’s on the
[00:55:00] rise faster than the US is. >> They’re they’re deploying solar at the most staggering rate, >> are they? Well, they have a lot of sun. Um but it still doesn’t make a lot of sense to me. But what what I wanted to ask you about is RHF engineering. You know the Merur is just growing like wild now that scaled AIA or scale has been acquired scale AI um and huge fraction of what is going on there is India. And we were at 1x robotics a few weeks ago too and they’re they’re like hey you know all this kinematic telematic data is going to be a gold mine for teaching the robots how to pour a cup of coffee without spilling it. And so a lot of that work you know which is creating a huge amount of jobs. Um it’s a new type of job but the Indian workforce is absolutely perfect for filling all those positions quickly. So is that potentially a factor in why OpenAI is pushing so hard into India. >> I don’t think they need to do it for that. My guess is is any major scale partnership and and and you know look
[00:56:00] there I think they’re looking for power and so you know Sem’s coined solar. I don’t know, but I would hope. Um I do know that there’s a bunch of other areas in the region that also have good access to a lot of clean power like Bhutan. Um and so I think the the I think it’s more that. But by the way, yes, let’s let’s use the talent and I don’t think they need to have that kind of data center in order to do that. >> You know, one thing that Vasilia Katumbas has done here is really focus this on education, right? Chatgptedu for secondary schools. And I’m still really pissed that the US has not tripled down on this, right? Made it edict that you must use uh you must be bringing this technology in. Uh it’s one of the most important things. You my two boys are 14 years old and you know the school systems are not preparing them for the future they’re they’re heading towards anywhere close. >> Yeah. And they’re not and they’re not not doing it. They’re just moving so slowly. But AI is just so fast. you know, it’s very hard for them to react
[00:57:01] >> because they’re used to making decisions over a 10year time cycle. >> I think what’s what’s going to happen though is the impedance mismatch between a student, as we said, five to times faster is just going to break the existing system. The forcing function there will be really powerful. We’ve been waiting for some kind of instigation to to to that will be a forcing function to transform education for decades now. And I think this might be it. So this past week, OpenAI announced it’s starting uh an AI chip production run with Broadcom. Uh it’s building three nanometer uh process. Uh so uh let’s start with you Alex. Thoughts on this one? >> This lies at the intersection of so many different trends that are all converging at the same time. On the one hand, I think this is a reflection of Nvidia’s high margins, relatively high margins, and in some sense, this is capitalism doing its thing and encouraging additional competition. On the other
[00:58:00] hand, I think this is a reflection of the proliferation of AS6 application specific integrated circuits to compete with more general purpose Nvidia GPUs. On the other other hand, I I think in some sense in the same way that Nvidia GPUs and Nvidia overall from a market capitalization perspective displaced Intel by being more specialized, I think we’ll see a rise of inference AI inference specific compute like a hypothetical open AI inference processor or accelerator maybe ultimately pose the threat that many people in the industry are are asking for, which is where is the next Nvidia going to come from? I I would argue if if there is going to be a next Nvidia, it’s likeliest to come from a more specialized ASIC that does a better job of focusing in a more energyefficient way at the the sorts of tasks that that we care about, the the sorts of workloads we care about. And then finally, biggest trend of all,
[00:59:01] Morris second law, that the cost of fabs is doubling every four years, and that’s steamrolling the entire space. It is so expensive to build a fab at this point that leaves anyone at the same time that Morse law Morris first law is basically ending or or near a conclusion or tapering off leaves everyone else who wants application specific acceleration fabricating their their own super narrow processors. So like 10 different trends all converging in this. Only thing I’d add to that, Alex, is that it’s it’s a new era in the sense that if if you were a uh TSMC and you’re building microprocessors, you know, preGGPU, and somebody comes out with a 2nm, 1.8 nmter, 1.4 nmter process, everybody moves to that new chip. Nobody wants the old chip because it’s it’s more power efficient and and it’s just it’s just a better buy. Now, all of a sudden, we care tremendously about just raw volume. And that never existed before. or you
[01:00:00] could you couldn’t just crank tile the earth with chips. Now you can and use it productively. So I think there are two avenues going on here. There’s uh there’s increasingly 20 billion 30 billion $40 billion fabs but then there are these new $4 billion fabs and maybe they’re stuck at 3 nanometer. They don’t go beyond that, but they have a huge ability to get up and running quickly and create a massive amount of volume because I think the algorithmic improvements are much more important than the difference between 3 nanometer and 2nmter >> and so getting scale of of volume and you know this slide kind of indicates that OpenAI is buying these from Broadcom but Broadam can’t can’t make them. They’re a design company, you know, they don’t have fabs. And so, you know, where are you actually going to get the manufacturing? And you look a layer deeper. And there’s a huge amount of investment and job creation, by the way. And then something all the governors should be looking at. Uh, get those fabs in your state like tomorrow, cuz that’s where all the jobs are going to be. >> Robots all the way down, buddy.
[01:01:00] >> Yeah, that’s true, too. >> All right. So, we have a new uh number one uh trillionaire in the house. Uh Oracle CEO Larry Ellison exceeds Elon as the wealthiest. He’s going strong at 81 years old. I’m very happy that he is someone focused on longevity and health span extension. I’m waiting for some good breakthroughs coming out of his work. So, OpenAI will buy Oracle Compute over the next 5 years, $60 billion per year. Uh the contract uh is a four for 4.5 gawatts of capacity. Two Hoover dams. I like that. We’re going to start measuring data centers in terms of Hoover Dams. So I asked the governor of Massachusetts how many Hoover Dams she wants. Uh I love it. So Open Eye adds Oracle as a partner. And uh uh you know you may or may not be able to comment but this begins to you know show some potential strain with Microsoft and a push to avoid having a single supplier. Uh comments on this anybody up for
[01:02:00] grabs? Well, I mean obviously I can’t talk about anything from a internal perspective, but I would say that it’s, you know, I think one of the simple things is as I’ve said a couple comments here is OpenAI wants to be in as many growth threads as possible and I think it’s a as much as possible and I think it’s kind of the fact that there’s a bunch of of volume that they could buy from Oracle. I think that’s the that’s that’s actually the real thing more than a more than a strain. >> Mhm. >> So I got a question for you Reed. So Chase Lock Miller was out here uh the day before yesterday. They you know he’s he’s Caruso which is building Stargate in Abalene, Texas. Uh and he’s you know MIT class of08. He got two degrees in three years. Not quite you know Alex getting three degrees in four years but but he got math and physics in three years. Absolutely brilliant guy. And I on stage asked him, you know, how is the deal with Open AAI? I see all these videos of you and Sam Alman walking around looking at all the pipes and
[01:03:00] wires. And he said, we actually sell it through Oracle to OpenAI. I was like, well, I didn’t understand for the life of me. I didn’t have time on stage to ask, but why is Larry sitting between you and Sam Alman? I don’t I don’t quite understand what his value ad is in the middle there. >> Um, I don’t know either, although I do know that a lot of the oracle is passed through. So, you know, in in terms of provisioning and building data centers and so but I don’t know I don’t know the shape of it. >> Huh. We made him the wealthiest man in the world. You can see on the slide there. So, so it’s a big deal. >> He’s a lot to live for. Let’s see if he gets to 150 first. >> Hey guys, it’s Peter. One of the things I found out is that a number of people are getting together for dinner to talk about the content around the moonshots WTF episodes and I want to facilitate that. On September 24th, uh there’s going to be a get together. Uh there’s a link in the show notes below. Uh people are getting together here in LA. If you’d like to go to one of these dinners, click on the link. Uh and uh
[01:04:02] and have fun. We have amazing subscribers listening to this, people who are building companies who are really going after moonshots. So check it out. Uh I’d love to hear in the comments if you go to the dinner what you thought of it. By the way, I don’t have an affiliation with the company listed in the link below. And while I don’t normally show up to these dinners, occasionally I do. All right, have fun. Back to the episode. All right, moving on to some Anthropic news. Anthropic raises 13 billion in a series F valued at 138 billion valuation. Uh they’ve got a revenue surge of 1 billion run rate in January. Uh now up to 5 billion run rate as of August. That’s insane, right? Over 300,000 businesses have uh have gotten enterprise accounts, fastest growth curve in tech history, and one of the most valuable AI firms. Um we’re going to start to redefine, you know, the Magnificent 7 very soon uh as something
[01:05:00] else. I think the broader picture here is that um enterprise has been sitting around watching all of these foundational foundational models get to a certain point but now you need the robustness and data sovereignty and on premises stuff that the enterprises need and I think that’s where the massive infrastructure and investment will go next. >> I totally agree. I don’t know if Dario will do it because he’s very very very uh safety conscious. So TBD whether open whether Anthropic is the company that does it or not. It’s interesting that Daario is totally neutral in these battles now because everybody’s making their own chips and so so the battle between Jensen and his own customers is just beginning and it’s going to be epic to watch. But Daario is is still the one guy who’s neutral in every conceivable way. Just I can work with anyone. I can sell to anyone. >> I don’t know. We’ll look at an article coming up very shortly. Reed, did you see Dario’s presentation at Davos or recording of it where he said uh that he
[01:06:00] could imagine doubling the human lifespan in the next 5 to 10 years on the back of AI? >> I I didn’t but I know Dario well so it doesn’t surprise me. >> Yeah. What do you think? >> What do you think of his machines of loving? >> No, no, no. I want to I want to ask him still about the life, you know, the uh are you do you buy this idea of of AI helping us double the human lifespan? >> Um I guess the short answer is trivially yes. It’s just kind of a question of time frames and how it’s happen. I mean >> part of for example >> you know the reason why you know Sedat Mukji and I are working on you know accelerating yeah manai for accelerating drug discovery with a focus on cancer is if you you know kind of um begin to get a you know kind of a set of the different cures. We naturally age in various ways, but a set of set of incures that which substantially elongate the kind of the aging curves in a healthy prosperous way that does it. If you have a medical assistant that allows you to be much smarter about, you
[01:07:01] know, kind of consumption and other kinds of things that helps too. Um, I think there’s, you know, um, and then precision medicine obviously accelerated with AI. I think it’s very straightforward. Love it. Dave, you were going to ask a question. >> Oh, yeah. Machines of Loving Grace, his whole, you know, kind of uh treaties on the future. Did you like it? >> Oh, I loved it. Um, look, I think part of the, uh, thing that people misunderstand about some of the people who, you know, he made this comment about safety, and I think he is, uh, very funny, this next slide, he is very focused on safety, but the reason he’s in it is a pro-humanist reason. It’s the same reason why the Open AI people are in it. It’s like what you know what are what what are the ways that we elevate the human condition. >> So speaking about this here’s our next slide. The title is AI safety sparks anthropic hunger strike. This is a quote from the guy on the hunger strike. Uh I’m on a hunger strike outside the offices of anthropic because we are in an emergency. The AI company’s race is
[01:08:02] rapidly driving us to a point of no return. I’m calling on Anthropics management to immediately stop their reckless actions which are harming our society and remediate the harm caused. So I don’t know this is from a few days old. He was on day three at that time. I don’t know if he’s still in a hunger strike or if someone’s if Uber Eats has delivered him a meal, but um >> way might have been way out. But yes, >> uh but in all in all respect I mean there’s here’s someone who cares deeply and is trying to make the point. What I find fascinating is when I think about all of the, you know, uh, frontier companies, I think Anthropic is the one who is the most sensitive, uh, to, you know, to these topics, to AI safety. >> Yeah. I wonder why he picked that one. Maybe it was the closest to where he lived. >> Yeah. I don’t know. >> Didn’t know where the new XAI headquarters were. >> Should have gone there. >> Yeah. Yeah. This feels to me as a
[01:09:01] candidate, an application forum for the Darwin Awards, not something else. >> Oh no. Oh no. Okay, moving on. So, Amazon’s AI resurgence. AWS and anthropics uh tranium expansion. AWS cloud revenue slowed as Google and Microsoft pulled ahead. Congratulations Google and Microsoft. Amazon’s invested $4 billion into Anthropic. I guess that was part of that $13 billion uh F series round uh to build a 1.3 gawatt data center capacity dedicated to AI training and Anthropic will run on Tranium 2, Amazon’s in-house AI chip, cheaper per memory bandwidth versus Nvidia. So this is what I was saying Dave when you were saying uh about Anthropic and Nvidia. It looks like they’re shifting towards working with Amazon here. >> Yeah. Yeah. Well, the traniums and then the TPUs at at Google are incredibly good uh inference time designs. Maybe maybe training, maybe not. We’ll we’ll know soon. But yeah, definitely a very
[01:10:02] serious threat to Nvidia. Not not because every everyone’s going to sell out everything they can make. There’s no doubt about that. Uh but if your chip is more performant and then you can argue for more manufacturing from TSMC. That’s where it all all gets bottlenecked >> uh is at TSMC. So um yeah these new chips I mean everybody’s competing with everybody. It’s all out war. All these companies that were in swim lanes and and could cooperate are suddenly absolutely absolutely at each other’s throats which is great for startups. Turbulence is always great for startups, but it’s really weird to see all of tech and a huge fraction of our economy in direct competition with each other. >> Alex, >> I would add also it’s not just to today’s point, it’s not just competition at the chip level. Maybe the uh from my perspective, the headline that we’re sort of burying here is the memory bandwidth so much if if you’re trying to do a coherent training run, actually the limiting factor is the chip-to-chip bandwidth, not necessarily the the
[01:11:00] compute within the chip. And here I think what we’re seeing fortunately is a bit of competition for NVIDIA’s Envy link coming from from AWS and Amazon. Uh AWS has this chip to chip interconnect technology named Neuron Link that is perhaps hopefully giving NVLink and Infiniband run for their money. And to to the extent that future training runs need to be coherent o open parenthesis. Do they need to be coherent or will we see some sort of radical overhang breakthrough in in terms of distributed training runs? >> Close parenthesis. >> Okay, there is some stuff in China that’s very promising on that front and that would totally it’s funny how a little innovation, couple lines of code could break the whole math behind these investments. >> Very, very interesting, fragile kind of thing. >> Reed, you can see why uh why Alex Gross is the favorite moonshot mate on this podcast. uh exudes his brilliance.
[01:12:00] >> Yes. >> All right. Uh let’s move on here. Uh next up is uh Poly Market. Um so Poly Market finally is coming to the US. Uh it’s uh an incredibly useful product. Do you play with Poly Market, Reed? >> Uh no. I I’ve done it a couple times. I’m uh obviously fascinated from a from a market point of view and um you know kind of how it plays out into the you know the various ways in which you know the general crypto uh environment you know um is shaping and how we shape it to try to make our societies better. You know, poly market for me is with some of the crowds. One of the things we discussed on a on a previous WTF episode is the idea of AI being able to predict the future. And the question is, how do you do RL with, you know, predictions of the future? I guess you could look at them in in retrospect, but poly markets could be an interesting truth signal. Alex, do you think so? Oh, I I think
[01:13:00] poly market and prediction markets in general, if you’re a startup and you you want to do free realtime research on your customers or on your competitors, prediction markets are a way to to do that. Uh and I I I think while we’re still in this gap uh this window of time between when we have prediction markets sort of collective intelligences or Peter I know you like borg organisms and and when we get super intelligences prediction markets are the the closest thing we have to a crystal ball for the future at some point probably I I would argue we get our super intelligences we get our Isaac Azimov Harry Seldon psycho history AIs that predict the future at that point maybe prediction markets gets subsumed. >> By the way, there is an asterisk here which is important is the prediction markets don’t set separately like theory in the world. So, like some of the things that I’ve been seeing happening has been people putting bets on, you know, like what uh color, you know, dildo will be thrown onto the sports
[01:14:01] ring, you know, first and then it becomes an economic incentive because you put a bet on your blue and you show up there trying to toss your your, you know, your blue dildo onto the sports ring first. And so there’s a weird intersection with society where it’s not just a kind of a physics of prediction but an interle of of of dynamic incentives. And >> the best way best way to predict the future is throw the dildo yourself. >> Yeah. To >> But Peter Peter, this is the problem with super determinism. You see everything that Reed just said, I made him say that. >> Exactly. There was an incentive. >> Love it. All right, moving on. All right, this was a a great note. I love this uh this chart. So, US patents have exploded during the AI revolution. So, you can see here on this chart the number of patents per year. My god, the poor patent examiners. They’ve got to be displaced by AIS. And we see here uh in 2022 an explosion. Uh 6,000 more patents
[01:15:03] granted in 2024 versus 2023. Um it’s a you know it doesn’t get more exponential or sort of vertical ascent than than right now. Uh what I >> what I read from what I read from this is that people are using AI to to patent applications. >> Well well not only that they’re using AI to create the patent application. I mean so one of the things one of my favorite things I did years ago when chat GPT first came out was I said okay here are two patent numbers. Uh this is the business I’m in. how would I use these patents to create a new product or service inside my business and it was you know here it is and okay is that now patentable uh so just this ability for people who want to explore this area >> well if you’ve ever been through the process too one of the companies in the studio uh is thinkruct it’s um >> Nikki Iate and Julius um >> uh they’re uh they started doing
[01:16:02] academic research using AI AI, you know, it’s like a toolkit and they quickly moved over to patent research and then patent filing >> and so they’ve automated the process. But if you do it the oldfashioned way by talking to a lawyer and they’re saying, “Okay, explain to me what this technological breakthrough is.” Like, oh my god, from from ground zero. You want me to explain that take days. But you do have the AI and it’s instantly, you know, here’s the application. Let’s go. And presumably the patent office has to read it with AI, too, cuz you know, this this will keep going up. It’s an arms race. It has to. >> There’s a patent law firm firm out here uh that I’ve recommended to some of my companies. Uh they’re based in Boston as well. What they’ve done is they’ve analyzed all the patent examiners and by different category and they’ve looked at the percent of allowances they’ve had. They’ve also looked at the time between application uh and review. And so they will direct your patent application to the examiners who have the highest rate
[01:17:01] of acceptance and the lowest time to review. It’s a game. It’s a, you know, humans in their loop are are can be gamed. This reminds me of that study that was done where if you where lawyers bring up their clients for parole hearings were kept trying to put them in after lunch because they found that before lunch the judges are hungry and you were going back to jail. after lunch they were biologically happier and you were 30% more likely to go free because the judge was biologically happier and like that’s just gaming the system to an nth level. That’s all this the whole thing is just a game. >> Alex, >> I I I was and maybe I’ll take the the opposite side of uh Sim’s comments. I I would say at this stage we’re still in the just the earliest innings of of AI generating transformative mathematical scientific engineering breakthroughs. So to the extent that we’re seeing any boomlet of patents being generated in part or in whole by AI, I don’t expect that on average they will be utterly
[01:18:01] transformative. I do think we will see transformative inventions being generated by AI over the next few years. >> What I’m talking about is it’s very clear that the patent application process is being generated. The application forms will be redone by AI which allows you a large number. Not that the patents are generated by AI. I >> I wonder what’s going on here. If you’re listening to this podcast on this chart, we look at the number of patents uh per year and uh it’s pretty flat from 1960 to 1996. Then we see this rapid ascent 6,000 patents over 8 years and then it begins to flatten out again with a,000 patents uh over uh what looks like to be a an 18-year period and then it explodes on the heels of chat GPT. What is that period between 2004 and 2022? Why is it you had a huge number of gene patents uh because you know they they decided right in that time frame that you >> and com patents too I’m sure >> I I would remind everyone to to read the title of the chart these are computing
[01:19:01] related patents. So I I think what what we’re seeing with the first boomlet is the.com boom and then we’re seeing the AI boom with the second boomlet. >> Ah okay actually looking at the data. >> Read the chart. >> Read the chart. Uh so we were on stage >> at least the title. >> Yeah, >> we were on stage with uh Ahmad uh the CEO of Replet and I’m sorry Amjad the CEO of Replet and Amjad had just released uh agent on that day. This was what Tuesday uh but he just posted this as well which replets agents are outpacing AI scaling. So uh this is uh the the meter benchmark uh which is basically looking at uh measuring AI’s ability to complete long tasks and he’s saying that this benchmark is wrong. Alex thoughts >> well so if we assume that the data that meter is collecting or the time scales
[01:20:00] that meter is estimating are accurate an exponential fit isn’t necessarily the best fit. It could be, for example, that we’re on, as Ray Kerszwhile would say, a hyper exponential curve. If we’re on a hyper exponential curve, then it’s entirely possible that we see some sort of blowup in the next few years. I I’ve seen estimates that if if we are on a hyper exponential curve, if that is indeed the best fit, that there’s almost an effective vertical asmtote in late 2027 or early 2028. So, it it may be the case. The the data are perfectly fine. It’s the fit that’s that’s perhaps overly pessimistic. >> So on this chart here, Amjad says, “Listen, agent one was able to think for 2 minutes. Amjad said agent 2 was 20 minutes and now agent three is 200 minutes.” We’re seeing uh we’re seeing a 10xing here and the question is will it continue? >> And I I think it it’s probably worth adding, if I remember correctly, he attributes that to well perhaps some sort of multi-agent approach is
[01:21:00] intrinsically better than another approach. My guess would be the exact opposite. My guess is multi-agent type approaches will just be naturally subsumed into existing compute scaling laws and we just find ourselves on a hyper exponential and all of this turns into transformative discoveries and almost magical AI on the time scale of 2 to 3 years regardless of whether it’s uh underneath something that looks multi-agent or otherwise. >> No, Alex, a bit of a question there. I do think that one of the things that is part of underlining this is how do we do various forms of parallelism. It’s the parallelism to the supercompute but also parallelism to agents because you got mixture of e experts as a you know key thing for the sparse models in order to grow. Um I actually think one of the things you’re seeing with a chain of thought reasoning and other things is is again uh putting in uh collections of agents in terms of how they’re operating together in order to get higher cognitive performances. So I’m curious a little bit about what your comment is
[01:22:00] because I actually think that multiple you know the kind of parallelism and kind of at least multiple entity constructions even if it’s to a targeting a kind of a singular output is actually part of the lesson here. And I’m just curious, how does your comment bear on that? >> I I love that question, Reed. So So the way I would answer that is to say multi- aent teams, multi- aent approaches in general, that’s just a form of sparity. So you could imagine to to your point multiple agents working in parallel together. that you could just view that through the lens of a much larger sparer architecture with with multiple feed forward lines that are all feeding forward in parallel that ultimately connect up at some point down the road. The the problem I perceive and history will will judge whether this prediction is correct or not with conventional multi-agent approaches is they’re usually not endto-end differentiable. Whereas one could imagine sort of a next generation multi-agent approach where the agents are actually part of one
[01:23:00] endto-end differentiable model where due to the way it’s sparssely organized, it actually if you squint at it looks like it’s multi-agent even though it’s one very large but sparse model. But that I I think speaks to your question. >> Yeah, we des desperately need better benchmarks for this and this came up with Blitzy Saturating Sweetbench last week or this week. Um, you know, if you look at human endeavor over a long period of time, many many things happen in parallel and then you know, read Peter’s book the future is faster than you think and you see how the the synergy is later. And so here you can spark, you know, an infinite number of parallel agents, not everything needs to happen after the prior thing. And so when you look on the on the curve it implies oh I thought of this then I thought of that then I thought of that but much of that processing can be done in parallel and also you can have many many redundant threads you know very often if you prompt 10 different things one of them works nine don’t and so you know you can take that from 10 to 100 to a thousand and just get a better result
[01:24:01] so all that is not baked into the you know the y-axis is just how much time is it thinking which is a crazy metric when you think about it >> hey folks here are many of You’ve asked where we can see more of Salem and where is he based etc. Well, we do a monthly workshop called 10x shift which is happening tomorrow the 17th. Uh and on that workshop we go 2 hours. It’s not recorded. It’s we keep it limited to about 100 people. Uh it’s $100. Uh people say it’s the best hundred bucks that they’ve spent where we coach people on how to 10x to 100x their organization using the exponential organizations model. and we go through and look at live examples and take questions and do coaching live on the call. I’m on the whole call for end to end. So if you want to hear more from me, that’s the place to do it. 10X your workshop link is below or go to openexo.com. We’ll see you there. >> So there’s technology we’ve all imagined years ago and it’s finally here which is live language translation in my Apple
[01:25:00] ear pod AirPods. Let’s take a listen. >> Talk just speak naturally. I’d love to take some of these to my sister for her birthday. I’ll buy eight, please. Your iPhone displays your words in their language and can even read them out loud if needed. >> Live translation is even more useful when both people are wearing AirPods Pro. >> I agree. Yeah, let’s include the Definitely the client will love that. I’ll let the strategy team know to prepare that immediately. This incredible capability is enabled by advanced computational audio on AirPods combined with >> So we saw Dualingo take a stock hit when Google’s live translation went live. I haven’t looked at Dualingo stock price here, but at the end of the day, you
[01:26:01] know, this looks like another incredible should have existed, finally does exist, and it’s going to make the world a little bit smaller. Any any quick thoughts on this one? >> Apple finally launched Douglas Adams Baplefish. >> Yes, Babel Fish. Thank you. Yes, of course. >> And hopefully it’s a lot better than Siri. >> Don’t get me started. This is this is audio augmented reality and and I would say now do video give us our our lightweight smart glasses. >> Yeah, I think the video the augmented conversation with video is going to be incredible. I think that’s the real vision. This is kind of cool, too. But >> yeah, next subject is robots and transportation. So, uh, Elon’s made this point before, you know, by 2040 expecting 10 billion Optimus robots and he’s gone to the market and said, you know, cars are okay, but the real opportunity is Optimus robots. So, he’s planning to scale to 1 million per year within 5
[01:27:00] years. Uh, automative, automotive sales comprise 74%. So, uh, a little bit about this. uh he was just recently speaking about this saying that you know generation 3 is coming online soon uh with the manual dexterity of a human in particular huge focus on the forearm and the hand with 26 actuators. His goal if he gets to a million per year is a a cost manufacturer of $20,000 each and he’ll price it uh depending upon what the demand is. But at the end of the day, what we’ve talked about here on the pod is a expected price in the 30,000 per purchase, 300 bucks per month for at least $10 per day. Uh any comments on on Optimus? I >> had dinner with uh dinner with Rodney Brooks last night, who is the iRoot founder. >> Yes, Rodney is the OG in the space. Yeah. But he was really pessimistic about the question at our table is will
[01:28:00] we have a robot in our home by 2035 which seems like a lifetime. >> Oh my god. >> And he said no. I was like really? He said yeah it’s all supply the the technology will exist but the supply chain won’t be there. So here you’re talking about Tesla making a million Optimuses per year within 5 years but there’s 300 you know 300 million people 150 million households. So that means very few of your friends have one in five years just because the supply chain is so slow to catch up to the demand. >> I’m supposed to get my 1x uh by the end of the year at least by March, right? You you heard me you saw me shake hands with >> uh that’s you’re you though you’re you’re the >> I was talking to I was talking to Steve Cousins about this and we kind of talked through some of this and there’s all sorts of issues. One is battery life uh is still way uh low for a bunch of these um applications. Um, the second is if it falls over, it’s going to be so heavy it’s going to be very hard to put pick up. >> Ah, the 1x 1x is like 70 lb >> and lord help if you falls on you uh
[01:29:01] type of thing. Uh, so there there’s a lot of areas uh where I think this is going to take much longer. I would I would for not so much the supply chain. I think just the liability issues and the constraining the function and the the actual action of what it does is going to take much longer to solve the insurance and legal issues. First, >> are you a are you an optimist on this read or a pessimist? A robot pessimist. >> An an optimist did you say on this? Um the uh look I I uh I was just kind of beused. Look, ultimately long term I think obviously it’s it’s there. Um, I think the short term, you know, I don’t know if Tesla has ever um met a >> potential hit a target. >> Touche. >> Alex, you were going to say something. >> Yeah, I I would also I mean I would focus on that 80% figure that that I I think is is such a striking number. If you think about and and think about the from a market analysis perspective, think about the size of automotive, it’s
[01:30:01] probably like four or 5 trillion per year worldwide. Um whereas if you think about labor and the services market uh it’s depending on which >> yeah no depend well depending on which estimates you believe manual labor is like 2/3 of the services economy and call that like 20 trillion so in some sense this 80% of Tesla’s value is really a bet that Tesla achieves parody with the services market. It’s it’s a general sort of universal intelligencepowered services company and I I I think that’s probably where the market overall ends up. Well, you remember he’s got to hit 8 trillion to get his trillion dollar pay package. So, if anybody can do it, I think Elon can. So, anyway, let’s move on. So, other robots here. Uh, all right. Let’s take a look at a different design. This is called uh the hidden robot. Boom. This is a generation of robots that don’t have five uh you know, two arms, 10 fingers. Sem >> here. Here. >> Yes. See on this on the WTF episodes
[01:31:00] like why do they have to have two arms and two legs and a head? Well, okay. This is what they look like otherwise. See, do you want one of these? Let me know. I I I think this is awesome. Um, I saw one where they were uh using it to map out the floors of a construction uh thing and and uh plan mapping out exactly where the pillars and so on would go. Um, I think this is huge. I think the industrial use for robots is so much past the home use for a while to come that people are under estimating that. I think the efficiency gains from that are going to be huge. And obviously the form factor should is not going to be humanoid. I mean, my beef, you’ve heard me before, but at least give me a third arm if I’m a humanoid robot. At least >> over dildo. Uh anyway, the use here wind
[01:32:03] turbines, nuclear plants, subc pipelines, railways, tunnels, power lines. Uh the old adage, if it’s dull, dangerous, or dirty, use a robot uh to do it. Uh the estimated market on this particular uh tweet that came out is that this sort of uh uh marketplace for robot maintenance and inspection is $6.7 billion today uh growing at about 13% per year expected at 12.5 billion by 2030. So >> I think that’s a radical underestimation, right? Uh, so just for example, if you look at the Mikong Delta in Indonesia, Vietnam, whatever, it’s so polluted. >> And if you had underwater robots cleaning it up, it would just completely change the game, make things massively better. It’s like the smallest titch of an application. >> I just want the robots cleaning cleaning the side of the 405 out here in LA. >> Okay. >> Right. Or the beaches, too. >> Yeah, sure. Uh, okay. I found this one
[01:33:03] super uh super interesting. And this is surgical robot performs gallbladder procedure autonomously. Right? So this is different than the da Vinci robot which is basically an extension of a human operating in a in a theater. And you guys all know what I’ve said about this. If you need to hire a surgeon for something, this is to our our amazing WTF Moonshot subscribers. If you need to get a surgery and you want to interview a surgeon, there’s one question you ask them, which is how many times have you done this surgery this morning, right? The success of a surgeon is a function of how many different cases they’ve seen and sort of the eye, you know, muscle memory on doing these. So, in the final result, I do believe the best surgeons in the world will be will be robots. They’ll see an infrared ultraviolet. They wouldn’t have a fight with their girlfriend or boyfriend. They wouldn’t have drinking, you know, drunk too much caffeine. So this comes out of John’s Hopkins and they built a surgical robot
[01:34:00] without any human control. It achieved 100% accuracy in this gallbladder remover. It’s uh different than Da Vinci and Da Vinci came out of was a DARPA project right to help surgery in the field. Uh I think this is huge. Um you know how long how far we’ll have how long we’ll have this in the future? I don’t know. I don’t think it’s more than you know 3 to 5 years. This is a this is a sensor actuator machine learning problem. Alex, what do you think about it? >> Yeah. No, notably, Peter, I I so I I read the paper, very exciting paper at that from from the the Hopkins team. This was a model that was trained by imitation learning. So, it was trained by watching videos of human surgeons perform surgeries. And that immediately uh rhymes in my mind with the early Deep Mind results like Alph Go that were trained from watching in part expert human games. And I I think we’re going
[01:35:00] to and I would expect if history does rhyme, we’re about to enter an era when using digital twins and and maybe this uh this this may or may not be aligned with what Reed is thinking for curing cancer with with Manis. We’re going to transition from imitation learning based medicine and surgery to reinforcement learning based medicine and surgery. the moment we have a highfidelity digital twin of the human body and and of course turtles all the way down cell virtual cell models as well. Why train off of copying humans when you could do reinforcement learning and achieve potentially super duper human level performance. So I I think again early innings but I think it it’s inevitable we we see sort of a surgery zero muse version of this sometime soon. Reed. >> Uh, look, I think the surgery part of this stuff is I think I agree with you, line of sight. I think we already, you know, it’s a little bit like for example, today you said I would rather would you pick an AI or your average
[01:36:01] radiologist to read your X-ray film, you’d pick the AI >> today, you know, hands down like, you know, 11 out of 10 times. >> And I think we’re we’re we’re we’re heading to that with the robotics now. I think that the approaches I think actually you know human biology is actually quite complicated and the ability to do a full simulation you know is is some ways off. Um but um I think that the this kind of robotic thing oh my gosh you know hit the accelerator. >> Yeah. Love it. Love it. >> I think of it as as air traffic control flying a plane right today the plane will fly themselves 99% of the time the pilot is only there in case of emergency. I think we’ll see the same thing. Yeah. >> All right. So, Zuks, I remember Zuks. I came I went and visited them and met the team there. Uh they were acquired by Amazon and uh back in 2020. Uh and this is a self-driving pod, right? This is you and your best buddies facing each other inside there. And they’re
[01:37:01] launching finally. Good fun them. They’re launching in Las Vegas. Uh they have massive scale. Well, not really. 50 vehicles in their fleet. Uh they’ll start in Las Vegas. Uh then San Francisco, Miami, LA. Uh it’s free for the first few months, but then they’re going to go to uh similar pricing as Uber and Lyft. Any thoughts, gentlemen? >> Have any of you guys taken the self-flying drone in Dubai? You know, you push the button and it picks you up. >> Not yet. >> Dying to. I just don’t want to be first. But >> just just don’t don’t do it to die. Let’s put it that way. >> They’re in production. Yeah. >> Yeah. >> Peter, you’re usually really adventurous with that stuff. >> I I would do it in a heartbeat. I saw it first at Consumer Electronic Show. You know, Martine Rothblat bought uh bought like a hundred of those vehicles early on for organ delivery before she got involved in her own uh you know vertical takeoff flying car uh beta. All right. Uh let’s wrap here uh at the end of the
[01:38:02] program. Uh I’m, you know, I remain just continuously uh like a kid in the candy store at the speed at which this is moving. Um you know, every day waking up and you know, I love the articles you send over, Alex, and and the conversations that we have. Uh you know, Reed, what are you most excited to see in the next year? Well, I’d say the next year will be part of the reason why I think the focus on coding and acceleration of coding is I think it uh accelerates everything uh accelerates individuals as per the co-pilots that I was talking about before but also accelerates the discovery and the computation the algorithms that Dave was talking about and I actually think we will see massive coding acceleration and that will be a precursor to many other accelerants is but is there one science one Star Trek part of the equation that you’re looking forward to
[01:39:00] >> Star Trek. Um, you know, it’s kind of the we tend to under um overpredict the the two years and underpredict the 10 years. So, it’s a little bit of the, >> you know, what would be the science fiction thing. I mean, obviously, we saw >> um you know, and I’m very hopeful about the iPod 3s. >> Maybe maybe a triquarter. >> Yep. I think uh we’ve reported on some early version of the triquarter. I done a $10 million Qualcomm Triquarter X-P prize 10 years ago. It’s time to do it again. I think uh the tech is the tech is there for sure. See, what about you? What are you excited about? >> Um passenger drones would be my personal favorite. I mean I eight >> eight EV talls. I spent eight days out of nine in India traveling between airports and like what the hell kind of waste of time crap is that in today’s world. The technology is there. It’s now just a implementation infrastructure. We saw a flying car. We saw a flying car in the campus of Stanford 2 days ago. Yeah. Very, very impressive. Very impressive
[01:40:01] design and it’s very workable. He thought he could get over time the cost down to about 40,000 >> a car. >> Yeah. I can’t wait. >> What I’m most excited about by far is a is a version two of Reed interviews Reed. I thought that was, you know, read AI talking to the real Reed was one of the most brilliantly conceived pieces of media. Anybody who hasn’t seen it, dig it up. Um, you could do it so much better today actually because when you did it actually, it was I assume you coded that up yourself, but that was pretty hard to do when you did it. >> Now you could do something incredible. And Peter did uh he did a onstage interview of Socrates and Aristotle that turned into a big love fest. So if you don’t prompt it right, everybody just loves >> when I I did I did an interview of myself last year at the Abundance Summit of 150year-old version of myself. >> Um which was fun and you know asked it about the future and it had some great answers. I loved it. >> Everybody watching this pod posts please read do it again do update it and
[01:41:01] that’ll put some pressure on you. >> And uh Alex, let’s not forget you buddy. What’s What do you imagine over the next year that would really sort of, you know, hit your childhood ambitions? >> Oh gosh. I I think I’m getting rather difficult to onlogically shock at this point. But I I I will say pulling all of Star Trek, not just some of Star Trek to the left. I I think that’s a worthy ambition. >> We’ve got the Holodex. >> We’ve almost got the replicators if you squint at food printers. We’re missing warp drive. We’re missing a whole bunch of other aspects of of Star Trek. Wouldn’t it be lovely if we were able to pull those to the left as well? >> I’ve got I’ve got one. I was chatting with Steve Jverson at the Stanford conference and he reiterated that crazy anecdote about once we have a quantum computer, it’ll be definitive proof of a multiverse and that really needs alcohol to get into. >> All right. Well, we’ll we’ll do that one of our next sessions. Thank you everybody to our subscribers. We appreciate you. If you’re not a subscriber, come join us. Uh, these are
[01:42:00] the conversations we have that we hope will make you more intelligent, more excited about the future, help you understand what the hell just happened in the last week cuz the speed of change is not just exponential. It truly is becoming hyper exponential. >> See you guys, my moonshot mates. Appreciate you, Reed. Uh, thank you, buddy. It’s been a wonderful friendship. Grateful for you. >> Massively fun. >> Yeah. If you could have had a 10-year head start on the dot boom back in the 2000s, would you have taken it? Every week I track the major tech meta trends. These are massive gamechanging shifts that will play out over the decade ahead. From humanoid robotics to AGI, quantum computing, energy breakthroughs and longevity. I cut through the noise and deliver only what matters to our lives and our careers. I send out a Metatron newsletter twice a week as a quick two-minute readover email. It’s entirely free. These insights are read by founders, CEOs, and investors behind some of the world’s most disruptive companies. Why? Because acting early is
[01:43:01] everything. This is for you if you want to see the future before it arrives and profit from it. Sign up at dmus.com/tatrends and be ahead of the next tech bubble. That’s dmmandus.com/tatrends. [Music]