AI Bubble, Stablecoin Boom, and Runnin' Down a Dream | BG2 w/ Bill Gurley and Brad Gerstner

00:00:00I'm applauding the innovation. I'm I'mI'm jumping on board the crypto trainand I hope I hope the incumbents aren'table to strangle this thing inWashington.

00:00:21Hey man, great to see you. Good to seeyou, Brad. What an incredible weekend ofcollege sports. You know, I have tobring this up. I mean Okay. Texas hadthat big upset of number six Oklahoma.You had to be pretty stoked about that.It was a good game. It was fun.>> People that haven't been to a neutralsite game, so like Florida does withGeorgia, Texas OU, they meet every yearin the middle of the Texas Fair. So thestadium's got several hundred thousandpeople outside of it. Oh my god. Andwhen you get inside right on the 50, onboth 50s, you know, all one team on theother and all the other fans on theother. So it's loud and it goes back andforth and it's unlike an experience youget where there's a home team and thecrowd's all just rooting for one team.Oh, that's cool. Well, my Hoosiers, myIndiana Hoosiers Bill upset the numberthree Ducks. Going to six and oh.>> I have a lot of duck fans in my in myfriend group, so I'm going toI'm going to refrain from celebratingwith you, but>> I I have to say I used to celebratingHoosier basketball, but rarely Hoosierfootball, but Kurt Zegner's done anunbelievable job turning that programaround. My 90-year-old mother waswatching that game and sending meplay-by-play. So congrats to congrats tothe Hoosiers. Well,

00:01:36as long as we'recalling out college football teams andthen we can move on.UCLA starts the season 0 and 4, thenupsets upsets Penn State at home withalmost no fans there and wins big againthis week and apparently there was acoaching change after the 0 and 4. Sothis could be the biggest turnaround inhistory of college football.Go Bruins. Pretty incredible. So there'sso much happening in the world today.We're going to unpack a few of thosethings. We're going to follow up on someof the issues Bill that I raised in theJensen pod. You know, the latest AIannouncements, all this bubble talk,circularity of revenues, quality ofrevenues, AI regulation. But we're alsogoing to do something today and coversomething we don't often talk about onthe pod and that's like life and career.Bill, you have a huge book coming out.Running down a dream, how to thrive in acareer you actually love. I'm so excitedfor this book. So we're also going totalk a little bit about that today.And and in the context of that, youknow, we're coming up on the two-yearanniversary of this pod. I can't believeit. Time has flown by, but our you know,when you and I talked about about doingthis, we said our mission really we wantto talk about markets, investing,capitalism and companies, but really

00:02:51through the eye of the investmentanalyst. You and I more than anythingelse I think are analysts. We try tofind the biggest problems,opportunities, challenges in the world,study them deeply. You know, we comparenotes non-stop. Occasionally as ananalyst that leads you to a biginvestment idea. Sometimes it leads youto a podcast, maybe writing an article,teaching a class, for you writing thisbook. And sometimes even a major policyinitiative like the Invest America Actthat actually became law. I think youwould agree with me the response overthe last two years has been amazing,more than either you or I expected. Butyou know, that also creates its ownpressure of its own to show up, todeliver those unique insights and thistakes a bunch of time. Yeah.>> So given that and I don't want to burythe lead here, you have some hugeupcoming projects you want to work on.And you're going to step back from theco-hosting the pod. I'll still talk youin on occasion, maybe to be in a guest,but you're freeing up time to work onyour big passions like this book andgoing deeper into these topics thatpeople have heard you talk about here,US-China relations, talking aboutregulatory capture, the dysfunctionalstate of US healthcare.And for those interested, the pod'smission

00:04:06remains the same. I'm going tokeep the same name. We're going to keepchopping it up with analysts I respect,sometimes with Bill and covering topicsthat matter. You know, like last week'spod with Jensen Huang or upcoming pods Ihave with Sam Altman or Sacha.This is a moment of I think reallyunique consequence. We both recognizethat. We're grateful to have theopportunity to open source theseconversations that are truly shaping thefuture.And I know I speak for you. We do it forthe love of the game. Like this keeps ussharp. It keeps us on edge.You know, and it's a privilege really toget on here and chop it up and sharesomething back with the tech ecosystemthat gives us you know, has given us somuch. So Bill, you know, you haveanything you want to say? It's a goodtwo-year run.>> Yeah, first of all, just thanks to you,Brad. Like this has been great goingback and forth. I had two primaryinitiatives coming into it. One, as youmentioned, was to stay sharp and theother one was to share and give back andI've been writing my thoughts on on thetech industry since I was a sell-sideanalyst, socoming up on 30 years and always enjoyedthinking out loud. I think it makes usbetter as analysts and helps us tounderstand. But but I also like to sharewith people and there's

00:05:21no question inmy mind that this got bigger than I everanticipated it would. I've been chaseddown in international cities recently. Iwas I started talking with someone. Theyhad no idea what I looked like, but theminute they heard my voice, they'relike, "Oh, you're the guy from thepodcast." Soumit has been popular and I I know thereare going to be people that are upsetwith me and that I can only say, youknow, I'm sorry and I apologize that I'mnot going to be doing it anymore. I cameacross this quote that was reallyinspiring to me. It said, "Life beginswhere your comfort zone ends." And therewere a number of people that helped pushme to write the book. It's taken upquite a bit of time in the last eightyears. It's been a very long project.We'll talk about it more later, but I'mfeeling a calling to go work on or atleast attempt to work on some of thesebigger issues. So I want to create aplatform for that. I want to create roomfor it and and move a bit away from thespace that I do know quite well and lovequite a bit, but pushing myself, youknow, outside of my own comfort zone andhopefully, you know, having an impact onthings that really matter. You and Italked about this throughout the entiretime I was working on Invest America. Icertainly encourage you and push you toto do

00:06:36this. I think you have an enormousamount to contribute. And listen, when Iyou know, you and I chop it up togetherevery day. I know where you stand on alot of these issues. I'll bring thoseopinions to bear for our audience andyou know, I certainly know that you'llhave the burning need to come on as aguest on occasion and and share some ofthose views. But but in the spirit ofanalysis, let's let's just dive in. ThisAI money bubble Bill and this Jensenpod. Let's start by talking about that.You know, we've had a just a flurry ofannouncements including anotherannouncement this morning between OpenAI and Broadcom where Open AI is goingto be building their own inferenceaccelerator amounting to well over atrillion dollars of incremental CapEx.That's above and beyond what we alreadyknew was going to get built out. I knowthat you have concerns about the levelof CapEx, the absolute level and I knowthat you also have meaningful concernsabout how it's being financed. Why don'tyou walk us through your major concerns?I think anybody that's been a student offinancial history hasyou know, studied different types ofactivities that that historically, let'sjust say historically have created redflags. And the reason that,

00:07:52you know,any AI you talk to would know what youmean if you said circular revenues isbecause someone has used it in the pastin a way that that wasn't good. And youknow, I had an exercise which I tweeted,we can put in the show notes, people canfind, but I just described, you know,there's not one thing. There's like sixdifferent transactions that havehappened now that I would say arenon-normal. And I just described thosethings to ChatGPT and asked it for itsanalysis both as an accountant and as afinancial investor. And the AI itself,you know, would would find its waytoward company names like Enron andWorldCom and those kind of things merelyby describing the type of transaction.And so I think that suggests if webelieve in intelligent AI, that that'sjust what historically has been come thebest practice and way to think aboutthese things. And I've I've told youbefore, I think you you have highlightedthat some of the multiples are actuallynot that high and I think this is partof the reason because there are redflags that people are looking at. If youpeel that back a little bit more, youknow, one of the things that you and Ihave talked about, the very nature, youknow, of round-tripping or circularrevenues, you know, I think

00:09:07there's thiscontinuum. On one end of the trans onone end of the continuum is a true shamtransaction. There's no underlyingdemand for the product. I send you abillion dollars, you send me the billiondollars back. Right? That's clearly asham transaction because there's nounderlying demand. On the other end, Ihave massive demand for my product. Youhave plenty of places you can go getcapital and we just happen to have a aninvestment relationship in addition tothat and I'm buying your product. Andthose things happen all over the placein our economy and you know, maybesomething to pay attention to, but it'scertainly not even close to beingillegal and it frankly doesn't evencause me a lot of concerns about thequality of revenue. And then we havethings in the middle, right? And thesethings in the middle where you can ask aquestion like, "Would this much ofrevenue or product had been purchasedbut for this investment?" Right? And Ithink that at a minimum calls therevenue, the quality of those revenuesinto question. So, when you look atthat, do you discriminate between thetypes of transactions that have beenannounced? I mean, you raised thisquestion first 18 months ago about thecredit transactions that were occurringwith a hyperscaler. So, maybe justunpack a few of the different types oftransactions. Yeah, and look, I think itstarted at the very beginning. And and Ithink that's

00:10:23one of the things that'scausing this is it's become part of thecompetitive landscape and thecompetitive dynamic. So, yeah, I thinkthere are many boards and many CFOs whohave been put in a position where theysay, "Well, if we don't do it, everyoneelse is doing it. You might fallbehind." But it But it started with theoriginal, from my perspective, with theoriginal Microsoft OpenAI deal, wherecredits go in as a in-kind investment,and then those credits are used backagainst, you know, Azure and andMicrosoft cloud services. And in the inthat case, you know, and I said it backthen, I'll say it again now, that's arevenue That's a cashless transaction.Like, there's no cash, but it becomes anincome statement uh revenue item forMicrosoft. And I don't think that'sideal from an economic standpoint. Andthat practice is now, I think, happenedat Amazon and happened at Google. Ithink they've made investments in otherAI startups with the same kind of thing.And at the very least, it drives usageof their product versus someone else.Um and in the worst case, you know, itcreates revenue that might not haveexisted had it had

00:11:38it not been for thatdeal, or at least not on those terms. Uhbut anyway, it started there. It'sbecome quite competitive now. There's aninteresting podcast on Plain Simple,which I I don't Plain English, which isa a uhin the Bill Simmons family, with PaulKad Kadrosky. And he He makes theargument that part of the reason thesetransactions are taking place I think Ithink this is a credible argument hemakes is because some of these playershave already put so much CapEx, so muchdebt on themselves that they don't wantto take the next step. And so, in thatcase, you know, you have reached somelevel where the company's saying, "Oops,you know, I feel uncomfortable goingfurther than this." The transaction thatcomes to my mind when I think of that isthere was one where MicrosoftI mean, Meta agreed to pay for thefailure of debt on a facility where theydon't own the debt, you know? And to me,that's classic off-balance sheetfinancing. If they own the risk of it,just because they don't own the paper,you know, it I don't see the difference,really. But this is Like I said, this ishappening in a in a lot of differentplaces.You know, one of the things that youknow, again, I've I've talked

00:12:53about alittle bit on All-In and other places.If I look at the Nvidia deal as anexample, Bill, Nvidia has theopportunity to invest, though not theobligation to invest. OpenAI has theopportunity to use their chips, thoughnot the obligation to use their chips,as evidenced by the fact they justannounced their own chip this morning,and they just cut up a huge deal withAMD. And, you know, in the case ofNvidia, you're not talking about ahighly levered business. It's a companythat's going to generate $450 billion offree cash and is taking a small fractionof that over the next 3 years andinvesting in companies that it it itthinks are good returning investments. Imean, Google and Google Capital havebeen doing this for, you know, foryears, etc. So, again, I think in thosecases, you can say for certain thatmaybe uh you know, more of their productis being consumed than would haveotherwise been consumed, right? Uh youknow, we saw this announcement last weekwhere they're investing in this xAI withrespect to their new round. Most ofthese companies that they're investingin, I think, have the economicwherewithal to raise the capital inother places. Elon could certainly raiseit in other places. OpenAI was welloversubscribed, so they could haveraised it in other places. But here'swhat I think people should be on thelookout for. Okay, so where would I havemore

00:14:08concern?Okay? Now, imagine there's a a a chipthat there's only one customer for. So,there's not a lot of demand for thechip.And that that chip manufacturer gives acustomer $10 billion, and that customerturns around and buys that chip,right? So, there's no other potentialcustomers, and that the the the buyerwould not have had a the ability to buyit but for that capital. That to meraises big red flags. And I do think inthis overall ecosystem, the reason I'mhappy you're bringing it up, I think oneof the things we need to do to keep thewall of worry there, to keep theexcesses from emerging, is to call themout. I'm not concerned as a shareholderin Nvidia with what I'm seeing Nvidia dotoday. I like how they're deployingtheir cash on their balance sheet. But Ido think that as you go further andfurther out the risk curve, right?Further and further to these startup neoclouds or further and further to startupchips, you know, etc., where people, toyour point, are a little bit moredesperate for capital, don't have thebalance sheets, don't have the marketleadership position, I would not besurprised at all in this moment to seemore of those yellow flags emerge.There's a couple things, you

00:15:23know, thatI would say in response. One, there's areason we know about a lot of thesethings, and that's because some auditorsomewhere made them disclose them. Like,they felt that it wasumabnormal enough to require disclosure.Second, I heard I listened to you andand the All-In team talk about thisissue. I I do think investment isriskier or more risk-seeking thancustomer loans, which it was comparedto. And Cisco got in trouble just withthe customer loans, cuz they were givingloans to startups who really didn't havethe wherewithal to pay them back. Butwhen That's really the issue for me,though, when you switch from a loan toan equity, you no longer have to pay itback. So, in some ways, it's uh it'seasier on the purchaser than if you hada loan themselves. But But here's myHere's my bottom line.I think what this does, this thisoverall situation, is first of all, Ithink it's driven by competition at thispoint. And like the first the first stepinto the gray zone was way back at thebeginning. And so, now I think we'rewe're fairly pregnant with it. So, Ithink it's a competitive dynamic. Ithink it increases the chance that we goover the top, that that we end upover-provisioning. And I I kind of feltlike that was unavoidable anyway. Butnow I think it's higher.

00:16:39And But I thinkit maybe pushes out when we find outthat happens, because you've you've justcreated more virtual leverage on thewhole system, and you might be hidingsome of the signs that would tell youthings are slowing down. I'll give you agreat example. One of the more peculiarof all the deals is, and this wasdisclosed in a CoreWeave filing, wasNvidia has promised to buy any ofCoreWeave'sservice availability that they can'tsell to anyone else. That is veryunusual. That's not the same as makingan investment. That could easily helpCoreWeave with their debtors in ingetting more debt financing. But it alsomeans, as a investor, we we don't knowwhat's going on with real demand forCoreWeave, because we probably won't betold if they start moving into the worldwhere they'reoffloading to Nvidia or not. And ifyou'd have said to me, "What would youlook for to see if we've, you know, kindof reached a point where things areslowing a little bit?" You'd say, "Well,let's look at one of the pure plays."Um And so, now that's money. I could seehow it it could be, but I I would expectthat every analyst on every CoreWeavecall for the next, you know, eightquarters, maybe thanks to you you just

00:17:54raising the flag, is going to be askingthe question, right? Do you see anyslowing? Do are you having to send anyof your uh of your demand to Nvidia, youknow, as a result of this? I mean, oneof the things I like about this as well,right? These are public companies, bothCoreWeave and Nvidia. It is a disclosedtransaction. It's not like this stuff'soccurring in the dark of night. Peoplecan ask the questions of this you know,this with regard to demand. And I willtell you,there is the the amount of money that isbeing spent to track every single partof this supply chain, from Taiwan to theUnited States. I mean, look at DylanPatel's business at SemiAnalysis. Thething has exploded. The amount of moneypeople are spending just to stay on topof this. And the second they seesomething that smacks of any, you know,leakage in demand, boom, docs fall andand and and warnings go up. So, I thinkit's a good point. But I think Let metransition, because I do want to talkabout this question of demand. So, onthe one hand, there's this questionabout quality of revenues. On the otherquestion is, are we overbuilding? So,let's show this chart again. This isbasically the $3 trillion of build-outexpected over the next 5 years. This isthe CapEx chart that we've

00:19:09shown herebefore. To put that in perspective,Bill,that's about 60 gigs, right? Because wewe now are normalizing everything's togigawatts of data center. So, that'sabout 60 gigs. It's not all incremental.A lot of that is is is replacement orupgrade. Keep that in mind. The secondis this chart of Nvidia revenues. Thisis the Nvidia sell-side forecast, okay?Forecast this year is for about 200billion in revenues, growing to about350 billion in revenues over the next 5years.So, this year that means that they'reselling about four to five gigs worth ofcompute. And again, most of that'sincremental, but it's not allincremental.And that would grow to nine gigs ofcompute, nine, in 2029-2030.So, that's 350 or 400 billion. That'sthe NVIDIA consensus revenue forecast,right? And I asked Jensen on the podabout this, and I said, "What is thechance that we get into a glut over thecourse of, you know, the next four orfive years?" And we'll play the youknow, the piece, but he basically saidthere's zero chance over the next two tothree years because all the build-outwill go to the biggest hyperscalers withthe biggest balance sheets

00:20:24in the world,and they're building it to to run theircore businesses. We haven't even gotinto the fullamount with respect to these newgenerative AI workloads. So, what is thepercentage probability that you thinkwe'll have a glut, we'll run into a glutin the next three or four or five years?Until wefullyconvertallgeneral purpose computing to acceleratedcomputing and AI. Until we do that.>> Yes.I think the chances are extremely low.Okay. So, here's a question I have foryou. Did you hear anything in the lastcouple of weeks that caused you tobelieve that we're on the verge of somebubble bursting or we're greatlyoverbuilding or or anything else? Or isit just the flags are up and now it's await and see?Yeah, sure. It's funny, they had HowardMarks on CNBC this morning. I'm a hugeHoward Marks fan, and they asked himthis question, and he he said, "Look,multiples are too low to for this to beyou can't be on bubble watch if themultiples aren't high enough." Andyou've been making this point for a longtime. And I would say like you'd have tobe a fool not to notice that thesenumbers you're talking about are so

00:21:39remarkably unprecedented from anythingwe've ever seen before. They aremassive. You know, I I've talked about,you know, to see the Mac 7 go frombeing massive cash producers to wherethey'remany of them are taking the majority oftheir free cash flow and the CapEx. It'sit's it's totally new, and clearlyeveryone believes that this wave ismaybe bigger than the previous waveswe've seen that have led to so muchvalue creation. So, all that'shappening. I like to believe that it'sokay to recognize that the market'sgreat and still think that thesetransactions shouldn't happen thisway. And I'm I'm able to keep both thosethings in my head at the same time. I II think I I think it's a super fairpoint. By the way, you just mentionedit, so we'll include this chart. This isMag 5 CapEx as a percentage of theiroperating free cash flow bill. And ifyou look at it in 2025,so that's this year, they'll spend about66%of their operating cash flow on CapEx.Yep.>> Right? And if you look at the the theconsensus forecast for their CapExrelative to their operating cash flow,

00:22:54this is the peak. Around 66%, and it hasit going down to about 45 or 50%. Now,embedded in there is they're going tokeep growing their operating free cashflow at 15 to 20% a year, right? So,there's still room for them to to growwith that coming down. But I think thatis another thing to keep your eye on.How much are they spending? And by theway, just give you an order ofmagnitude, Bill, in 2023, their totalCapEx was 156 billion.And this year, it's 379 billion.Right? So, radical step up in in yourpoint's a good one. And just remember acouple years ago in 2022,when Meta stepped up their CapExspending on Reality Labs,the stock got obliterated because peoplesaid, "What the hell are you doing? Thisis all about free cash flow per shareper share." Including myself, and I wassaying, you know,let's get fit here. Let's drive morefree cash flow out of the business. Youknow, so much so that the CFO sent me ahat that says free cash flow, right? So,they got real about free cash flow, butthere's a difference between investingthat free cash flow in data centers andAI than there was in Reality Labs. As aninvestor, let me just tell you my ownperspective. The reason

00:24:10Meta's stock'sdoing great, notwithstanding going backto high levels of CapEx spend, isbecause now the investors understand itand believe in it. We're seeing thebenefits, right, in the earnings of thebusiness. They're growing the earningsof the business. They don't have to hirea lot of new employees, and so it's it'sfundamentally different than the CapExthat was going into Reality Labs, whereinvestors were saying,"Hold on a second, we're going to spend100 plus billion dollars over the nextfive years. We don't even know whatwe're building."Right? We don't know what it'll be worthat the end of the day. So, I think fornow, at least, there's enough belief inthe the byproduct of generative AIbecause people are using ChatGPT,they're seeing the utility in theenterprise that they're going you know,that they're willing to tolerate thesecompanies giving over half of their freecash flow to these build-outs.I do think another dynamic is the racecondition created by the competitivedynamic. And it appears from where Isit, and you don't need to comment cuzyou're an investor and and maybe havemore information than I do. Probablyhave more information than I do. But itappears to me that OpenAI, um throughall these partnerships andannouncements, is trying to createescape velocity, you know? And thatcould be against the model

00:25:25providers, itcould be against a hosting provider,depending on how you think the marketplays out. Um it could could be on theconsumer side, it could be on the APIside. But it it it creates aninteresting stress test for anyone elsein the ecosystem to say, "Are you goingto lay chase?" Because of all thenumbers you laid out there, they'rethey're gargantuan. And it it'll beinteresting. That's my opinion. It justfeels like they're they're daring peopleto to follow them, and and I suspect abunch don't. I mean, it may work. Well,it's you've seen this before, I know.We've talked about many times on here.This was in Lyft. Um ultimately, I thinka couple things to remember. These theseannouncements are frameworks. It itallows people to begin working, butthey're not etched in stone. These arenot contractual obligations. You know,everybody's got to deliver their parts.If the demand comes in lower, then thesepeople are not going you know, leave>> Oracle wouldn't have to be contractualfor it to be RPO. Yeah, you know, I youmay know more>> Well, for sureI shouldn't say that they're all rightframeworks, but I know, for example,like in the case of AMD, they're goingto have to deliver a workable chip, oryou're not going to build six gigs'worth. Yeah.>> Right? Uh in the case of this Broadcomannouncement, obviously, they have tobuild a workable

00:26:40chip. So, I think yourspeculation, and again, it makes senseto me. If you said, "What are theadvantages of getting out there andlocking up all of these deals?" Right? II can't imagine it it it doesn't help alot with recruiting. All the bestresearchers in the world want to work atthe place that has the most compute, andso you want to lock up the compute. I II have to imagine it helps with thesupply chain, because, you know, nowyou're you're locking up that supply.So, I think your speculation is a prettybig one, a pretty good one. But at theend of the day, if you add up all thesedeals, I tried to do this, and we may bewe may be off by a bit, and I'dencourage people who have a betterestimate to let me know. But if you addthem all up, it looks like to me OpenAIwould be on the hook for like 150billion of CapEx in 2030.Okay? And so, like, the question is,Bill, what how much revenue do they needin 2030 to justify 100, you know, 150billion in CapEx? Well, I think youwould need at least 150 billion ofrevenue.Right? And, you know, like, at aminimum. We just talked about Meta andthese companies spending 66% on CapEx.But if they had 150 billion of revenue,then the question is, is it plausiblethey could have 150 billion dollars ofrevenue in 2030? And I would

00:27:55argue as aninvestor that it's it's more than morethan plausible that they could have 150billion of revenue. But I think it makesit very, very difficult for anybody elseother than the hyperscalers. Obviously,Google's going to be there.Obviously, Meta's going to be there.Obviously, Amazon, you know, can bethere. But it makes it very, verydifficult for anybody else in theecosystem, right? Who believes this is agame of of scale compute competes. So, Ithink your point's a good one.All right.>> maybe shift a little bit, you know, toan area of passion for both you and I,which is this AI regulation. We'vetalked on the pod many times about theconcerning patchwork of these emergingstate regulations that, under the guiseof doing good, and maybe they're evenwell-intentioned, cause a hell of a lotmore confusion at best, and at worst,they set back our leading frontier labs,you know, and and really hamper us inthe race to to stay in the in the leadin global AI. Well, it's gone from moretheoretical to now more more moreproblematic. I tweeted over the weekendin particular about this Colorado AIAct, which is now passed into law,signed into law. It defines somethingcalled algorithmic discriminationby outlining

00:29:10these 12 protected classes.Of course, age, color, religion, butalso limited proficiency in Englishlanguage, reproductive health. Andbasically said if the algo providesinfo,right? If the chatbot providesinformation that's used to discriminate,then there's liability back at thefrontier model level,right? And just this morning, GavinNewsom signed SB 243, which mandatessafety protocols for AI chatbotcompanions, and gives any consumer aprivate right of action to sue thesecompanies for any emotional harm thatcomes out of a chatbot. I mean, Iyou not. You can't make this stuff up.And you said recently that China is socompetitive with the United Statesbecause it's run by engineers andAmerica's run by lawyers, and that's thegreatest risk we have. Talk to us aboutthe need for federal preemption and youknow, again, just let's dive back intoyour concern about these two laws thatwere just passed.As an aside, I just consumed JonathanHaidt's book Anxious Generation where hetalks about what he believes is somesocial harms caused by some of the appson the internet ecosystem.

00:30:26And I dothink a lot of the passion for writingsome of these states' laws comes fromthat place. Like there are localcongressmen that that feel like theyshould have been out in front of socialmedia more, and so they want to get ajump on this. And and and I I think someof that comes from there. You run thismassive risk of of trying to regulate abrand new technologyat a state-by-state level. And you know,you could ask yourself, you know, andand and by the way, II've said this a lot about policy. Theintent of the policy is different fromwhat happens once the policy isimplemented. And so people can come inwith great intentions, and this goesback to my speech at All-in onregulatory capture, and you can end upwith the exact opposite outcome of whatyou intended because you just don't knowenough about the way you write theregulation. Right now, a lot of peoplebelieve we're in this global competitionto to see, you know, whosetech stack for AI is used on a globalbasis. Andif we implement 50 different state rulesthat these companies have to jumpthrough, and companies that are that arecompetitors that are competing

00:31:42in thebroader world don't have any of them,there is zero chance that's not going tocreate mud and slow down the US players.There's just zero chance. AndI'm certain the people that are writingthese laws don't understand that theremight be some global, you know,consequence of what they're doing. Butbut it's bad. I mean, I can rememberwhen Obama was excited about removingsome of the state-by-state requirementson like hair stylists and whatnotbecause it makes it such that they can'tmove between states. And it's kind ofridiculous that they would havedifferent laws and different licenses.This is likeif that's a problem, this is really aproblem. And so I, you know, from aglobal competitiveness standpoint, Iwould certainly hope that that they'reable to federalize this and preempt it.I don't know if it's there's too much,you know, water under the bridge or not.I don't know enough about what it takesin Congress to make that happen. But Ithink this is bad forI think it's bad for the US. it's badfor innovation broadly. It's going tomake it harder for startups to do thingsjust because they're going to have toworry about all this stuff. Yeah, it'sit's it's way worse for little tech,right? Because smaller

00:32:57companies don'thave phalanxes of lawyers. They can, youknow, go out and comply. You know, Ithink the other thing is listen, wealready have the Civil Rights Act, wehave the Fair Housing Act, we have theAmerican with Disabilities Act. Ofcourse, we don't want discrimination.But this just seems like broadoverreach. It's no like I don't evenknow how you comply or enforce. And soit just ends up bogging down the entiresystem in uncertainty and litigation.And again, it's important to say thisisn't even about whether or not AIshould be regulated.It's just a question of who shouldregulate it. And what we're saying isthat, you know, there is ampleopportunityfor this administration and Congress toget together and write legislation tothe extent it needs to be written,right? To provide a national These areinherently interstate technologies.There's no way to keep it in a singlestate. And so write a piece of nationallegislation that allows us to continuemoving forward very quickly, but all theat the same time addresses any of theseconcerns. I frankly think we need a weneed a moratorium on all state laws.Postpone all state laws until thefederal government has time to act. Andif states are going to pass these laws,Bill, then I wonder whether or notOpenAI or some other company should

00:34:12consider blocking the citizens of thosestates until it's resolved at thenational level. Somebody needs to getthe attention of these states that theycan't do this on a state-by-state level.It's bad for the companies, it's bad forthe country. But hopefully we'll getaction out of Congress soon. I thinkthere's good momentum. We almost had itpassed as part of the I think the bigbeautiful bill.And so I think there's a lot of movementafoot in order to do it. I wanted tohighlight it because I think it's it'sone of the high priority issues facingthe new Congress. All this AI all thetime stuff, Bill, and you you pinged meand you said, "Hey, I want to talk aboutstablecoin, right?" We have thisparallel development in the world. So Ithink if we have three major trends inthe world,AI is clearly the largest supercyclegoing on. The reindustrialization ofAmerica is massive. All these criticalsupply chains. And I would say the thirdone is kind of the digitization andtokenization of finance. It's going onas a result of the administration

00:37:54basically turning>> And you have to fill out pages and pagesof crap. Sometimes get a verbal. And itand it and it and it's, you know, we areso behind. I am going to go out on alimb and say I hope someone, whoeveragitated this to happen in the Trumpadministration, is someone who is kindof caught in a regulatory captureposition getting lobbied by somebody. II think if the Trump administrationstudied this, they shouldn't be criticalof PIX. They should be envious of it. Weshould have done this a long time agowith FedNow. But we may be on the vergeof stablecoin just being able to do thisanyway. And the rails have tons oftransaction on them as you've talkedabout. We have this interestingsituation where Coinbase and Circle havedone this deal where Coinbase will allowyou to earn 4% on your stablecoinbalance. Which, you know, to to get thatkind of return at another bank,even a neo bank, you have to have yourdirect deposit go there. Here, whetherit's 10 bucks or or a million bucks, youknow, you put it in stablecoin withCoinbase and you start earning 4% daily.And on top of that, and this gets backto the PIX thing,you can transact immediately

00:39:10out of thataccount. So you don't have to like moveit from your savings to your checking toget it to do ACH. Like you can sendstablecoin immediately in microseconds,and it'll cost you a few pennies. It isThe rails are there. They're ready. Andit's working. And I think the UI is alittle difficult, but there's no reasonwhy that won't get better and faster.And so I look up, you know, and I I'mjust I wonder what the team at Meta likethey might just be kicking themselves.Like with all the money they spent onthat coin, everything they wanted to doon WhatsApp, like they should be runningback at it. Like I don't know>> Well, I think so maybe they should, youknow, remember the guy who did the theLibra network, David Marcus, has startedthe lightning network now. It's astartup. Maybe maybe Meta should go buyyou know, should go buy lightning andbring David back in house becauseyou know, all the things they talkedabout all the things they talked aboutare now what's happening, Bill. And letme tell you one I think you're ontosomething big here, but one of thechallenges we still have I think firstanybody who looks at our current systemright, knows that it's dreadfully behindthe rest of the world. Right? And weknow it's the result of regulatorycapture by not only

00:40:25our issuers, but thebanks and everybody else who who likesthat who benefit from the status quo.But if you look at Visa and MasterCardtoday, I think that they're doingsomething like 50,000 transactions persecond. And I checked with our goodbuddy Vinnie Lingham. And he said, youknow, on on both Solana and ETH today,they're still under 4,000 transactionsper second. So they're trying to come upwith these solutions to actually makethe rails have the functional throughputand efficient settlement required toreally become a consumer product. But Ithink you nailed the other one that youknow, Patrick Collison had a tweet onthis that I replied to, which is youknow, when Genius Act was passed, therewas massive lobbying by the banks toprevent the crypto companies from payinginterest on stable coins.So the settlement was that they couldpay rewards, not interest. Okay? And umbut what the way in which it'smanifested itself because Coinbase isnot the issuer, Circle is the issuer andthey did this deal. So Coinbase ispromoting it as though it was interest.So from a consumer perspective, a rewardand interest if it's 4% isindistinguishable. No doubt. Right? So Iput my money>> and obviously Brian's been out on BrianArmstrong of Coinbase

00:41:40has been out on Xlikearguing his side of the argument. Sohe's clearly he's either gettingopposition or expecting opposition onthe regulatory front. And when I see,you know, whether you know, whether it'sVisa or Nasdaq or any of these peoplekind of run at the tokenization I alwaysworry like cuz if you look at thehistory of like the debit card versusthe credit card, it was supposed to bedisruptive. It was supposed to be analternative that would change things,but they just they just run at it andstrangle it and mix it up a little bitand then it's not as disruptive as itwas. That's their go-to move. But youknow, when I read this thing on PIXagain, like I'm going to read this outloudas part of its aggressive economic andpolitical campaign against Brazil isinvestigating PIX accusing the paymentsystem of unfairly utter undercutting USfinancial and technology companies likeVisa and Apple.I meanthat's the most absurd thing I've everheard. Undercutting Visa? Like do dothey realize they have Visa MasterCardhave like the top two operating incomesin the history of American business?Like like there's there's no one thatneeds less protection than these guys.If anything, there should be aninvestigation

00:42:56>> dinner at the White House?>> cabal? I don't know. I I just like it'sa it's so bizarre to me. I'm so thrilledto to see this kind of disruption. Ithink that it's super interesting what'spossible. I suspect all the big guysshould be paying attention to this.Apple, Google, Amazon anybody that mighthave payment on their rails I'll go outon a limb, Bill. You're going to see theyou're going to see the hyperscalers.You're going to see Amazon and Meta andthese guys back involved in the stablebusiness. At the end of the day, we knowmoney is a network effects business. Andthe challenge of Circle and some ofthese stables from a consumerperspective isuniversalVisa and MasterCard are universal.So you got to get to all the merchants.Well, who has all the merchants? Amazonand Meta.Right? And so I think they're in a greatposition to partner with or do some ofthese things themselves. Clearly theyhave the instinct to do it. That's whythey did Libra in the first place. It'salso amazing for innovation. And and oneof the reasons why I think those biggercompanies should run at this is if youlook at the history of the PIX likealternatives I mentioned in the UK andChina and India the startups that dofinancial innovation scale up way moreaggressively successfully on

00:44:11those railsthat are cheaper and faster. And it ifif anything the havingmore rigid, high friction, hightransaction cost rails makes it harderfor a startup to think about, you know,using one of those technologies. And sothe success of WeChat Pay and Alipay,which I as I described for my China chipare universal. They're only way peoplepay in China happened because of thatgovernment instant pay product, not notin spite of it. And and the same thing Italked to the CEO of Nubank. He said PIXwas huge for his business. So it'sprobably bad for a laggard bank, but fora bank that embraces it, it just becomesa better feature. And I think the samething about Coinbase and what they'redoing here. So I'm applauding theinnovation. I'm I'm I'm jumping on boardthe crypto train and I hope I hope theincumbents aren't able to strangle thisthing in Washington. Here, here. Here,here.As we move towards the end and and talkabout my book and what I'm going to donext, I do want to share with you thatthat both my book and and the the nextproject are outside of what I've I'vespent my career doing. And and as Imentioned, you know, that that's kind ofmoving outside my comfort zone, but it'salso,

00:45:26you know, trying to have an impactand give back in areasthat that I don't know as well, but butwith a hope towards having an impact.I've said this to you before, but I I'vebeen justinspired, frankly, to go do this basedon your success with Invest America.When you first told me about it youknow, I had doubts that you could get itdone, real doubts. AndI've watched other people in your shoestry and do these types of things overdecades and be unsuccessful. So you madeit look easy. I know it's just gettingstarted, um but I wanted you to know howmuch that inspired what I'm going to godo. And and could you give us an updateon where things are? Well, that means alot, Bill. Maybe to talk about it firstis just a reminder. You know, we I thinkyou and I agree, we kind of have thisbattle for the soul of America when itcomes to capitalism right now. Right?And that that's fundamentally becausetoo many people feel left out and leftbehind. 60% of people will never ownassets that compound. Mondami's winningthe mayoral race in New York City. Andyou know, they're doing it by beinganti-capitalist.But if you look at these two charts,I've shown them many times before itjust shows you that free marketcapitalism is the most productive forcein the history of the world.

00:46:42Right? Thisfirst chart just shows that GDP on aglobal basis went parabolic at the exactsame time that capitalism was reallyintroduced and started taking off. Andyou know, remind reminder like GDP isimportant because it's that surplus forhumanitywith a fixed amount of labor and capitalthat then leads to better schools andbetter hospitals and and and drugs thatsave lives and all the things that makeour lives better. And you can look atthis chart that shows the results. Fewermothers die in childbirth. The averageage you know, of life is extended. Thequality of life is higher. Literacyrates are higher. Bill Gates extols uponthis in his in his annual letters. Soit's not just an investment account.This is really a much much bigger battleover where we want the country to go.And I was very concerned as you know, afew years ago that we were headed downthis path. Right? And the path is thatyou can't have so many people left outand left behind. So I think the answerto socialism, which has not worked forEurope, right? It Europe's in a disasterrelative to where they were 30 years agoon a on on a global competitive basis.And China, as you well noted, havepulled themselves out of poverty byleveraging

00:47:57capitalism, right? So theanswer to this drift into socialism ismore capitalism.And the Invest America accounts, nowknown as the Trump accounts right, aremore capitalism. They make every child acapitalist from birth, a private ownergives them a thousand bucks in a 401klike account that they own and controltheir family has on their phone. And soI think that's a game changer, you know,but you're right. We just got it passed.So what where are we now? The TreasurySecretary, Scott Bessent, has toimplement this. And by the way, this isone of the largest consumer launches inthe history of government. So at thestart today, there are 65 million kidsin the country qualified for an InvestAmerica account. Every kid under the ageof 18.And every kid under the age of two willautomatically get a thousand bucks intheir account. Right? So you'll probablyhear a launch starting in maybe earlyDecember. We'll the launch the website.People can sign up for this. Rememberthe accounts have to be funded andestablished by our 250th birthday, July4, 2026. That's only nine months fromnow. And I can tell you I've been blownaway by the secretary, SecretaryBessent, the assistant secretary, LukePettit, and the team at Treasury workingwith the White House. They're attackingthis

00:49:12the way I would attack I wouldexpect a a Silicon Valley startup toattack the problem. They've gotten agreat a great group of technologists.Joe Gebbia helping to design the frontend of this, of course, from Airbnb. Soyou know, we're we're on the verge nowof some major announcements, the thestart where people can start signing uptheir kids. You know, and then the goalis once we launch this, all these kidswill have these accounts. they'll beable to roll them over into theirfavorite, you know, bank, whether it'sSchwab or Fidelity or JP Morgan or whathave you. And starting on July 4th ofnext year, Bill, as close to automaticaccount creation as possible. So, youhave a child, the child's born, they geta social security number, and they getan account seeded with a thousand bucks.From a kid's perspective, it's going tolook like I own a little bit ofMicrosoft, and I own a little bit ofUnited Healthcare and Nvidia andwhatever. We're going to be able toteach us, in fact, you know, your buddyTim, who's teaching financial literacy,you know, we now have 30 states requirefinancial literacy requirements. We'regoing to have this embedded in theschools. Every kid's going to have thison their phone. So, at any rate, it'sgoing incredibly well. I give I givethem a very high score, but we got toget it done.So,let's talk about your book. Well, I'vebeen a huge fan ofof the speech you made

00:50:28on this, but Ilove the title, Running Down a Dream.What's the thrust of the book? Irememberyou know, kind of that lecture, but whatreally compelled you to write it?So, I was years ago, and this this isprobably going back 10 years. I wasreading a lot of biographies, and Inoticed certain patterns among peoplewith extraordinary careers. And as VCs,we see a lot of patterns in businessesand pattern recognition, and I just sawpatterns with people. And in the back ofmy mind, I always wanted to do thispresentation. I I I kind of kept noteson it like I would a unwritten Above theCrowd blog post. And I had anopportunity, I got invited, had anopportunity to give the presentation tothe MBA class at the University ofTexas, and so I, you know, I worked onit and put it together, made it nice,and gave that presentation. They endedup putting that on YouTube, and manypeople have come to me and said thatit's changed their lives, encouragedthem to do different things. Certainpeople in the in the in the mediaindustry noticed David Senra, who's gotthe new podcast where he interviewedDaniel Ek and Mike Cole. He's a big fanof the presentation, talks about it alot in on his podcast. James Clear, whowrote Atomic Habits, maybe one of thebest

00:51:43self-help personal developmentwriters out there. He retweeted it andput a transcript on his own website. Andthen a few people who are influential inmy life started prodding me, "Hey, youshould turn that into a book." And so,eventually I got convinced. We talked topublishers, they were interested, and soI started working on this. Now, it tooka long time. And the the thing I wouldsay about it is that I hope that thattime equates to quality. Like, I was outthere really wanting to make it great.And so, the book has an interesting kindof novel architecture. We combine what Icall profiles, so stories of successwith principles, tools of success. So,they alternate. So, you get a a storymaybe about someone you didn't know, andhow they started at the very bottom andbecame successful, and then the types ofthings they did. I mean, I do I do thinkthat the principles, these tools thatare in the book, are things people canuse. And I really wanted to be great.It's done. I still need to record theaudio version. All the podcast fans ofours tell me I have to do it. It has tobe my voice. So, I'm going to do it. Ihope people love it. I I hope it changestheir lives for the better. And what Ireally want to encourage is

00:52:58people totake a chance and and do what theyreally love.I think it's such an important topic.You know, one of the things parents areasking me so much these days is, youknow, what what what should my kids do?Particularly give There's a lot ofanxiety in the world.Right today about future careers. Andso, maybe just talk a little bit aboutwhy this is so important now, because Ithink the timing is really profoundhere.So, in in the introduction chapter, weunpack a lot of this. And I don't thinkanyone would be surprised when I readsome of this, but but Gallup Poll does acareer engagement study. They've beendoing it for a long time. I think in the2023 when only 23% of people said theywere thriving or engaged at work, and59% were unsatisfied. And that's just abig universal survey. Everyone seemsaware that we've kind of moved to thisgauntlet that we've put our kids in asthey approach college and go throughcollege. In The Coddling of the AmericanMind, Haidt and Lukianoff called it aresume arms race. Andwe've really taught them to be grinders,but Angela Duckworth highlights that ifyou have persistence but not passion,you eventually recognize you're in agrind. And and and and when you come outof that, you're in a really tough spot.And so,

00:54:13we've got people We've got thesekids on this runway. We We're tellingthem like they have to pick a a majoreven in their applications. So, they're17. What do you want to do with yourlife? What do you want to do? And Andthey they really don't know. And onething we stumbled upon doing research onthe book, I was working with aresearcher. We did this survey and askedpeople, if you could start your careerover again, would you do thingsdifferently? And in that survey, 70% ofpeople said yes. And we did it againwith Wharton just to make sure we had,you know, a true academic survey goingon, and they did a lot more people, andthat number was still six in 10. Six in10 said if they could start a careerover, they'd do it again. Andthere's a there's a great book. I read alot of books in writing my book, butthere's a great book called The Power ofRegret by Daniel Pink. He's He's awell-known author, but he he has thisthing he calls boldness regrets. And hesaid, "One of the most robust findingsin the academic research and on my ownis that over time we are much morelikely to regret the chances we didn'ttake than the chances we did."He says again, the surface domain,whether the risk involved are educationor work or love lives, doesn't mattermuch. What haunts us is the inactionitself. And so, I think that

00:55:29I thinkthat ties really nicely with this ideathat if people could start over, they'ddo something they'd do somethingdifferent. There's a great video thatthat Pink references that we can put inthe show notes where Bezos is askedabout the decision to leave D.E. Shawandand start Amazon. And he said he used aregret minimization framework. He saidonly a nerd could do that, but he saidumthat he imagined himself being 80, andwould he care that he left D.E. Shawmaybe for winning a bonus, you know, orwould he or would he care more that hedidn't take this chance, this kind ofinstinctive chance that he felt like hehad to take. And he immediately afterthinking about it in in that way wantedto go do it. And so,>> Hey Hey Bill I have Bezos's regretminimizationframework taped to my computer monitor.There you go. I didn't even know that.>> literally taped to my monitor. Yeah.Powerful.Yeah. So, so that's that's that's whatthis is about. That's what this book isabout. That's who it's for. And I wantmore people to take a flyer and go dowhat they love. You know, we have aphrase I use in the book, life is a useit or lose it proposition. Totally.

00:56:44One shot deal, man. Well, how do youplan to promote it? It comes out inFebruary, late February. And so, I'mjust getting started in that process. Ifpeople have ideas they want to sharewith me, please reach out and let meknow.>> I'll host I'll host a book launch event,Bill. Okay. We We've got We've got a lotof fun stuff planned, but but but uh butum I need to record the audio book. I Iknow that's going to take a a lot oftime. I'm excited. There's a handful ofpeople that have read it, maybe maybe 50in the publisher and whatnot, and nearlyevery one of them tells me that theyimmediately thought of three or fourpeople they want to give it to. And so,I hope there's kind of a viral componentto it because people have that reaction.But, you know, if you're feeling stuckin your career, I you should read it. Ifyou're a teenager or young adult whofeels overwhelmed by people telling youWhat do you want to major in? Where doyou want to go? What are you going to dowith your Like, I think this bookwon't put more pressure on them. I thinkit'll actually relax them and give thema framework that feels like a lot morepersonal to themselves and like they'rea lot more in control. If you're aparent that wants to help a child onthat journey, I think parents sometimesoverly push kids into the lawyer doctorbanker framework.

00:57:59I'm not sure that'shealthy, especially in this AI worldwhere those jobs may be under risk aswell. And then maybe if you're, youknow, an administrator or someone in thetype of role that guide people in careerdecisions, hopefully you'll like it aswell. But, we'll put the pre-order linkin the in in the show notes. I will putit on my X feed. Please go out andpre-order the book. I think it'll helpit be reach more people and and besuccessful. Well, I think it'sit's going to be hugely impactful. Andit's the type of stuff of consequence.I'm I'm I'm just, you know, I'm thrilledthat you're doing it. I'm thrilled thatyou're taking the time to do it. I knowthat's not the only thing that you'reyou're thinking about. You have theseother, you know, big topics that you'rethinking about as well, Bill. We'vetalked about them here, and I'm sure wewill continue to. Regulatory capture, UShealthcare, nuclear, et cetera. What areyou thinking aboutwith respect to those things? So, I'mYou going to write a book on every oneof them? No, I That's well, I could, butit's not it's not my goal. My My goal isto just go spend more times on thesereally big problems and see if I can behelpful in any way. I've spent a careerkind of breaking down and analyzingdifferent situations. I mean, two two ofmy favorite podcasts we've done, and

00:59:14andI hear I hear about this from thecommunity as well,the one we did at Diablo Canyon, andthenthe one about around my China trip. Andthose types of of work are veryrewarding for me, but they also werelearning expeditions. You know, I wentout and put in more hours for thoseepisodes than others.>> Right. And the the nuclear one inparticular, you know, one of the thingspeople ask me, why would I want to go dothis? You know, we were a small part ofa movement to kind of change the mindseton nuclear energy. And there were therewere people that put a lot more effortinto us. I'm not trying to take creditfor it. But, you know, the fact thatSteve Pinker was out there, Elon, youknow, our stuff like eventuallyhappened. Like overnight, it seeminglyovernight, we went from a very negativemindset towards nuclear energy torecognizing that it's very clean energyand something that that can really helpsave the planet. And so, that type ofmeme flip, if I could go achieve more ofthose in these other areas, I wouldconsider it a win. So, that's what'smotivating me. I'm really lookingforward to it. I'm kind of fired up andnervous at the same time. But but that'swhat that's what I'm thinking about.There'll be more to come on that interms of what the actual platform lookslike.

01:00:29I'm still working on it. But fornow, I'm going to I'm going to sprintinto February to make sure the book doeswell. I couldn't be more stoked for you.This has been a total blast. You and Ihave been chopping it up for, you know,for a couple of decades, but doing thislast 2 years together, pounding outthese has been a lot of fun for me. I'msure we'll continue to chop it up everyday. And, you know, I'm sure that you'llfind some topic that you can't livewithout exploring on>> Can't resist talking about. Exactly. So,so we'll we'll we'll get you back on.But I'm going to give you the last word,Bill. It's awesome awesome to hear aboutall of this, and I'm super excited forthe book. Well, I would I would just,you know, end the way I started, Brett.Thanks to you. It's been fun workingtogether on this and and and doing itevery week. It does force you to stayfresh. You have to read everything youpossibly can, which I'm sure is superhelpful to you as an investor. And then,thanks to thanks to all our listeners.LikeI'mI'm sure some of them are feel going tofeel like I'm letting them down, and I Ifeel the weight of that. Um buthopefully they'll recognize that I'mgoing to go try and put put my workeffort to good to good causes.

01:01:45As a reminder to everybody, just ouropinions, not investment advice.