E130: DeSantis's Twitter Spaces, debt ceiling, Nvidia rips, state of VC, startup failure & more

1. 0:00 – 3:03

   Bestie intros!

   1. JCJason Calacanis0:00

      Look at John Quincy Adams here.

   2. CPChamath Palihapitiya0:01

      Yeah. Hey, hey, Quincy. How you doing? (laughs)

   3. JCJason Calacanis0:04

      That hair. (laughs) Look at that hair.

   4. DSDavid Sacks0:07

      John Quincy Adams needed a week to achieve this look.

   5. CPChamath Palihapitiya0:09

      He did. He did. That's true.

   6. DSDavid Sacks0:10

      "I, I did it au naturel."

   7. JCJason Calacanis0:11

      I mean, it's a lot. It's a, it's a look. You're like some movie star from the 1970s who's still working. Do me a favor. Pull up a picture of Graydon Carter for a second, and then put that side by side with Sachs. Let's see.

   8. CPChamath Palihapitiya0:23

      (laughs)

   9. JCJason Calacanis0:23

      Just, yeah, zoom in on that. Sachs, this is where you're headed, by the way.

   10. DFDavid Friedberg0:27

       Ugh.

   11. JCJason Calacanis0:27

       You're headed to crazy town.

   12. DFDavid Friedberg0:28

       (sighs)

   13. JCJason Calacanis0:29

       This is where you're headed. You can just poof it out-

   14. DFDavid Friedberg0:31

       (laughs)

   15. JCJason Calacanis0:31

       ... on the sides, and you get a little woof, swoop on the top.

   16. DFDavid Friedberg0:34

       Crazy town. (laughs)

   17. JCJason Calacanis0:36

       This is where you're headed, eccentric Sachs. Pull up a picture of Steve Bannon's hair.

   18. DFDavid Friedberg0:41

       He's a wise statesman.

   19. JCJason Calacanis0:42

       I think this is what happens. When you get too close to power, you get more eccentric with your hair. So like, too much power equals crazy hair. Now look, this is my theory. Graydon Carter, too much power, Vanity Fair, he went hair crazy. Now you look at Bannon. You get me a Bannon photo here. You start to see, they go longer. They go-

   20. DFDavid Friedberg1:00

       (laughs)

   21. JCJason Calacanis1:00

       ... wafty. They just, they get volume. And they're just like, "Fuck it. I'm just, I'm not gonna cut it. I'm gonna let it go wild." And of course the, the ult- I mean, he's the penultimate.

   22. DFDavid Friedberg1:10

       (laughs)

   23. JCJason Calacanis1:10

       But you look at Trump's hair, this is where it gets truly crazy. This is where you're headed, Sachs. You keep getting this close to power.

   24. DFDavid Friedberg1:18

       (laughs)

   25. JCJason Calacanis1:18

       This is where your hair is headed. (laughs)

   26. CPChamath Palihapitiya1:20

       I'm so glad this podcast never broke up, 'cause where would we get this amazing insight from J-Cal, if not for keeping this all together? This makes it all worth it.

   27. JCJason Calacanis1:28

       This is-

   28. CPChamath Palihapitiya1:28

       It's very true.

   29. DFDavid Friedberg1:29

       It's, yeah.

   30. JCJason Calacanis1:29

       It is true. You get like, eccentric, and you get crazy hair. This is men-
2. 3:03 – 28:18

   Sacks goes behind the scenes on the DeSantis Twitter Spaces experience

   1. JCJason Calacanis3:03

      Hey, everybody. Welcome to episode 130, maybe, of the All In podcast. We're still here, cooking with oil. With me, of course, the dictator himself, Chamath Palihapitiya, prince of panic attacks, sultan of science, the queen of quinoa, David Friedberg, and the power broker himself, the emperor, David Sachs.

   2. DSDavid Sacks3:29

      (laughs)

   3. JCJason Calacanis3:29

      The emperor of his new republic.

   4. DSDavid Sacks3:31

      (laughs)

   5. JCJason Calacanis3:32

      Anybody have any interest, anything going on interesting this week? Any interesting-

   6. DSDavid Sacks3:36

      (laughs)

   7. JCJason Calacanis3:36

      ... moments for people on the national stage?

   8. DSDavid Sacks3:39

      (laughs)

   9. JCJason Calacanis3:39

      No? Okay. Well, let's get right up into the docket. First story-

   10. DSDavid Sacks3:42

       (laughs)

   11. JCJason Calacanis3:42

       ... on the docket, Ron DeSantis, Governor Ron DeSantis announced his bid on the internet, uh, on something called Twitter Spaces. And it looks like almost 10 million viewers have seen it so far across all the different spaces. And Donald Trump wasn't too pleased. He said, "Rob, my big red button is bigger, better, stronger, and it's working." Truth, because when Elon fired up the Twitter Spaces, it went to 650,000 viewers in under five minutes, and then blew up everybody's phones. My phone was melting. I could have cooked an egg on the back of it, the Twitter app crashed so many times. But then Sachs, with his meager following of a half million people or something, then restarted the stream. And so 15 minutes of technical snafus, we're relieved, and then there was a, uh, announcement. And I'll let you take it from there, Sachs.

   12. DSDavid Sacks4:38

       You want me to take you behind the scenes?

   13. JCJason Calacanis4:40

       Take us behind the scenes.

   14. DSDavid Sacks4:41

       (laughs)

   15. JCJason Calacanis4:42

       Take us behind the scenes.

   16. CPChamath Palihapitiya4:43

       How did it come together, Sachs?

   17. JCJason Calacanis4:45

       Oh yeah, even better.

   18. DSDavid Sacks4:46

       Yeah, the way it came together is, I think, the DeSantis team were interested in potentially, you know, doing something different for their announcement. He also did an appearance on Fox News afterwards, and I think he did a town hall. But I think they saw an opportunity to, to break new ground here in terms of presidential announcements by doing it on Twitter Spaces. And so the DeSantis campaign connected with Twitter, and Elon and I agreed to kind of co-host the space, uh, with him. And he did his announcement. Now, you're right, we had about 15 minutes of technical difficulties, because the interest was so intense. At the time the, the room crashed, it had over 700,000 people in it. And there were... It crashed 'cause so many people were trying to get in it.

   19. JCJason Calacanis5:28

       Yeah.

   20. DSDavid Sacks5:28

       I think there was well over a million people trying to get in it. So, you normally don't have this kind of interest. I think this is by far the biggest Twitter Space. The, the engineers there told me that the previous order of magnitude was more like 100,000, not a million. And then you combine that with the fact that Elon's account has over 100 million followers, and that basically led to a new level of scale. And you guys understand that when you get to a new level of scale as a platform, there's always going to be some-

   21. JCJason Calacanis5:55

       Yeah.

   22. DSDavid Sacks5:55

       ... some challenges. So, in any event, the engineers were there trying to figure out, how do we, you know, solve this? And we realized the simplest thing to do would just be to restart the room on my account instead of Elon's. And-... then Elon joins the cohost and we brought DeSantis in. And it all worked perfectly at that point. The audio was crisp. We had over 300,000 people in the room. There was also a- another room that had been set up by Mario Nawal, who's like a big Twitter Spaces host, and he had hundreds of thousands of people on there. And then he had live commentary from people he invited. And so this ability to fork Twitter Spaces into many different rooms, and each room gets to decide who they want to be their hosts and their speakers, allows you to do live commentary in- in a way that you could never have done before. So it was really innovative, I think.

   23. JCJason Calacanis6:44

       Super innovative. And for-

   24. DSDavid Sacks6:45

       Yeah.

   25. JCJason Calacanis6:45

       ... people who don't know, Twitter Spaces was really a rush job at Twitter. They did that in reaction to Clubhouse. They- it's just, it's still basically a beta product that predates Elon being there. And it doesn't have yet the infrastructure or scale of the code base, I don't think, like YouTube and Twitch do, which, you know, have been working on this problem for, I don't know, 15 years, maybe the live products, over a decade.

   26. CPChamath Palihapitiya7:08

       I have two observations.

   27. JCJason Calacanis7:09

       Yeah, go ahead.

   28. CPChamath Palihapitiya7:11

       The first is I thought DeSantis did a really good job just rolling with the punches.

   29. JCJason Calacanis7:16

       Okay.

   30. CPChamath Palihapitiya7:17

       Because I think whether he wins, you're not gonna look back on this moment as the defining moment of the campaign, nor whether he loses will you say that this was where it was it all the beginning of the end. Instead, what this was, was a really seminal moment, I think, in further divorcing ourselves away from the mainstream media. And you know that it was that important because Biden tried to troll the whole thing. Nick, you can show this link. This link works. And I actually think this is a really terrible idea by the Biden team because-
3. 28:18 – 40:44

   Debt ceiling, government spending, and lack of accountability

   1. JCJason Calacanis28:18

      of science. US debt ceiling is at 31.4 trillion currently. Uh, Treasury Department recently warned the federal government could be unable to pay its bills as soon as June 1st. Fitch put the US credit rating, which is the highest rank of triple A, on negative watch. So negative watch means it's trending towards bad and that it's imminent that they might lower it due to this debt limit deadline. It seems like we're not making much progress. Every couple of days we seem to swing one way or the other. The stock market has kind of shrugged it off. And the last time the US credit was downgraded was in 2011, but we avoided that. Chamath, what's going on here? What do you think the eventual outcome? And is there a chance that these knuckleheads who represent us are going to default and cause chaos? Or do you think this is all kabuki theater and we're gonna wind up in the same place, which is they raise it and make some modest concessions?

   2. CPChamath Palihapitiya29:16

      August 5th, 2011, the S&P downgraded the United States from triple A to double A+. You know what happened? Absolutely fucking nothing.

   3. JCJason Calacanis29:29

      Okay.

   4. CPChamath Palihapitiya29:30

      So I do think that this is another opportunity for the Little Red Riding Hoods to get their panties in a bunch.

   5. DFDavid Friedberg29:38

      (laughs)

   6. CPChamath Palihapitiya29:40

      But these downgrades don't mean much of anything.

   7. DSDavid Sacks29:42

      Who are you referring to there? Are you talking about Freebird?

   8. DFDavid Friedberg29:44

      Yeah, what are you talking about?

   9. CPChamath Palihapitiya29:45

      Talking about, talking about all you guys.

   10. JCJason Calacanis29:47

       Wait a second, is that Biden-

   11. CPChamath Palihapitiya29:49

       Talking about all you guys-

   12. JCJason Calacanis29:50

       ... talking about the media?

   13. CPChamath Palihapitiya29:51

       I think that these third party credit rating agencies are not particularly that accurate and sophisticated. They don't know anything that you don't know. They're not getting access to-

   14. DSDavid Sacks30:03

       Moody's gave SVB an A rating the week before it went into receivership.

   15. DFDavid Friedberg30:07

       (laughs)

   16. JCJason Calacanis30:07

       Wow, well done.

   17. CPChamath Palihapitiya30:09

       This is my exact point.

   18. JCJason Calacanis30:10

       (laughs)

   19. CPChamath Palihapitiya30:10

       This is my exact point. None of these companies know what they're doing. These companies are in the business of putting a letter on a document and then selling access to that document. So if you're gonna trust these guys to know what they're saying-... either right or wrong, independent of what side you are outsourcing your decision-making to the wrong body. So whether it's S&P, Fitch, or Moody's, I would e- tell you to ignore it and come to your own conclusion. I think that this budget ceiling thing is happening now every couple of years and it seems like the real thing we should be talking about is whether Biden's gonna use the 14th Amendment and just ram a budget through. And I think that that's-

   20. JCJason Calacanis30:51

       Explain- Yeah.

   21. CPChamath Palihapitiya30:51

       So under, under the 14th Amendment, the President of the United States has in their discretion the ability to make sure that the United States can pay their debts. And that wasn't necessarily thought of as a way to work around the debt ceiling impasse. But because Congress refuses to pass any structural laws that allow the budget to ebb and flow with tax receipts, we get in- caught in this situation, again, roughly every s- handful of years. So what Biden could do is he could say, "The 14th Amendment gives me the right. I'm gonna pass a budget via executive order." From a game theory perspective, what that does is it forces Republicans to sue Biden, take him to the Supreme Court, and say, "This is unconstitutional." The problem with that is that that probably really does tank the economy in the way that it creates enough uncertainty where capital markets freeze up and liquidity just absolutely goes away. And again, liquidity's been shrinking for the last 18 months anyways, so it'd get even worse. So I think the brinksmanship right now between McCarthy and Biden is basically that veiled threat. If Biden effectively isn't gonna get something done, I think he's gonna test the 14th Amendment. Now, if the 14th Amendment turns out to be a reasonable way in order to pass a budget, the good news is, not just for Biden but for any f- future president, including Republican presidents, will not have to be held hostage by Congress. They will, in the 11th hour, be able to pass a budget that works. The implications of that though is that now you will not get consensus and whatever's happening in that moment, you'll see more of. So if you have a spending president, they'll just continue to spend and y- if you have an austerity president, they'll continue to cut, and that'll have implications on either side.

   22. JCJason Calacanis32:37

       Sax, do you think this is over issues of 14th Amendment or do you think it's a valid use of it?

   23. DSDavid Sacks32:42

       No, I- it's not gonna fly. I mean, it's true that the 14th Amendment has some language about, uh, the full faith and credit of the US shall not be compromised. Uh, however, it's never been tested or, or tried, so no one exactly knows what it means. Progressives are now saying that the language means that Biden could just keep spending without the debt limit being raised, but Biden himself threw cold water on this. He said that he didn't think he had that authority and even Lawrence O'Donnell the other night, who's a big progressive on the media, he was saying that if the Dems were gonna take this tact, they needed to do it months ago before they started negotiating on a debt limit increase. So it's too late now. In other words, if you're gonna take that position, why would you have negotiated? Then you're- y- you, the president are effectively conceding you don't have that authority when you start negotiating. So I think it's too late to revoke the 14th Amendment. They're gonna have to raise the debt ceiling. That being said, I think they're gonna be able to. Reuters had a report this morning that they're only 70- 72 billion apart now in their positions, which is a relatively small amount. So my guess is they're gonna- they're gonna work this out. Now what should the fate of the debt limit be moving forward? I mean, the- the thing that's so stupid about our budget process is we spend the money and then argue about whether we're gonna pay the credit card bill.

   24. JCJason Calacanis33:57

       (laughs)

   25. DSDavid Sacks33:57

       The way that the debt limit should work is you raise the debt before you spend it. Congress should have to vote first on whether they want a deficit or debt spend.

   26. JCJason Calacanis34:07

       Mm-hmm.

   27. DSDavid Sacks34:07

       Then you decide how much you're gonna spend. So this thing- this vote needs to be moved up before they spend it. And I think if you did that, it'd be a lot harder for the politicians to spend money because, you know, if you had an up or down vote very early on saying, "Should we even be s- in deficit?" I think a lot of people would say-

   28. JCJason Calacanis34:23

       Huh.

   29. DSDavid Sacks34:24

       ... "No, we're not in- we're not in a war, so why- why would you deficit spend?"

   30. JCJason Calacanis34:27

       Feel free to bring any thoughts on the debt ceiling. If not, we'll go on to the, uh, government accountability vis-a-vis the defense spending.
4. 40:44 – 57:46

   Ways to force more government accountability, who really wins from higher taxes

   1. JCJason Calacanis40:44

      And Chamath, I guess, is, seems like there's a couple of unpopular stances to take when you're running for office. One of them is Social Security retirement age, as we saw in other countries like France, where people are arguing over 62, 64 years old, whatever it is.

   2. DSDavid Sacks40:59

      Yeah.

   3. JCJason Calacanis40:59

      And so these entitlements, job requirements if you wanna get unemployment, and then of course defense spending, you seem un-American. If you don't wanna take care of old people, you seem un-American and, you know, weak if you don't wanna support the government. Is there a path towards celebrating an administrator as CEO and executive who tells honest truth to the American people, which is, "Hey, we, we've been on a binge. We've been going on vacation. Nobody's looking at the, the bills, and, and we need to have a staycation. We need to cut costs. We need some austerity here." Is there a path for somebody, Ron DeSantis, Chris Christie, RFK, whoever it is, to win over the American public, to win over moderates with an austerity and a balance-the-budget message? Or is it just too unpopular to even bring that up?

   4. DSDavid Sacks41:49

      I just think that you guys don't psychologically understand how to get what you want. I think the best way to get what you want, which I would want too, which is a healthy economy where there's accountability for spending, is to not look at expenses, and all of this talk is always about expenses, but to look at revenue and limit revenue more drastically. And I think the best way to limit revenue dramatically at the federal and state level is to just minimize taxation as much as possible. And I think that is something that Democrats and Republicans have a hard time fighting. Nobody wants to raise their hand and say, "I want new taxes." Nobody says that.

   5. CPChamath Palihapitiya42:34

      ... right? And I think that if you attach that to some sort of spending guidelines, like what David says with a sensible foreign policy, you just end up spending a lot less. You wanna fight a foreign war? Okay, great. Well, you know what? It's part of the existing budget. Let's go figure out why we wanna do it. We wanna have more accountability in defense spending? Okay, well look, the defense budget is half of what it used to be, because we just have half the revenue. I think that if you start to go and talk about austerity and cutting Social Security and healthcare benefits, it's literally a non-starter. People close their eyes, they plug their ears, and you get nowhere. Now, separately, I think the step even before you look at taxation, so minimizing government revenue, is to figure out how to refinance. So if you're a homeowner and you got a mortgage in the '80s, you were paying 12, 14, 15, 16%. If when rates kept going down, you or the people around you didn't have the common sense to refinance that, that was negligence. Similarly, we're in a position today to refinance our debt wall and push out these maturities past 100 plus years. We are the only country that has a viable, stable economy that looks like it can still continue to thrive at scale. Take advantage of that. You lose nothing by giving us the optionality, generating some 100-year debt, refinancing a bunch of this short-term stuff. And then second, inflation helps us, and it helps us because it allows us to inflate the value of these dollars that allows us to pay off our short-term maturities. So, these are two practical, simple things that are uncontroversial that should happen today. And then separately, I think you need to look at minimizing revenue, and then you can cut expenses. But if you flip them around, I'm just telling you, it's not like I want any of this to happen, but I'm telling you, nothing will change and you guys will still be crying wolf. In five years, it'll still be the same. It'll just be a different debt-to-GDP number that gives you anxiety.

   6. JCJason Calacanis44:37

      Speaking of anxiety, Friedberg, your response.

   7. DSDavid Sacks44:39

      I don't agree. I think we need to balance the budget and I think that if we don't, we continue-

   8. CPChamath Palihapitiya44:44

      Wait, what don't you agree with? Financing, pushing out 100-year?

   9. DSDavid Sacks44:47

      No, no, I don't think reducing revenue solves the problem.

   10. CPChamath Palihapitiya44:50

       For the country's debt?

   11. DSDavid Sacks44:50

       I think that... And by the way, one of the problems with a democracy, Chamath, is that you speak about it as if everyone benefits from a tax cut. Generally, there's some disproportionate benefit to a tax cut and that's why it's less likely to happen, because the majority will benefit by keeping taxes high for a minority, whether that's some corporate minority or whether it's some wealthy individual minority. And that why I think the opposite is more likely to happen, which is we're more likely to see taxes go up in order to bridge the gap to continue to fund programs that everyone wants... Everyone wants an and, they don't want an or. And as a result, they'll kinda continue to seek revenue because there are other places to get revenue that don't affect me, meaning it doesn't affect the majority. And I don't mean me personally, I'm just speaking about a voter. And a voter would say, "If there's a way to tax other people for me to get the things I want, they will vote yes for that." And that's ultimately what the system ends up finding, and I think that's what's more likely to end up happening. That's all.

   12. CPChamath Palihapitiya45:51

       Sachs.

   13. DSDavid Sacks45:53

       Yeah, look, I mean, I, I agree with part of both of what you're saying which is, I agree with Friedberg that we need to balance the budget. There's no excuse for running peacetime deficits this large. We're really gonna regret this one day. On the other hand, I agree with Chamath that if you just try to solve the problem by raising taxes, the politicians will just keep spending. I mean, you have to starve the beast, I think, in order to control it. At least that's been a view, I think, for a long time.

   14. Do you, you agree with Chamath on that point?

   15. Here's the thing, you cannot solve this problem by raising up to 70% tax rates like we had in the 1970s and I can prove it. Pull up this chart.

   16. France is a good example of this, too.

   17. So what you see in this chart here is, this is federal receipts as a percentage of GDP. What I see when I eyeball this is, if you were to put a regression line on that, it'd be at around 17%, with a plus or minus of 2%. And the times when you get up to 19% or a little bit close to 20% are when we have great economic conditions or money printing. So, the Clinton era from, was it '92 to 2000, was an economic boom and we got up to almost 20%. But we've never ever been able to get more than 20% of GDP in federal tax receipts, even during the 1970s when the top marginal rate was 70%.

   18. What happens, Sachs? Just explain what happens.

   19. CPChamath Palihapitiya47:10

       And why... Yeah, why is that? Do people stop doing economic activity?

   20. JCJason Calacanis47:13

       No, it curtails investment and economic activity. This has been, like, documented-

   21. DSDavid Sacks47:17

       Yeah, that's what I... That was my guess. Yeah.

   22. JCJason Calacanis47:18

       ... for hundreds of years. Taxation doesn't solve this problem.

   23. DSDavid Sacks47:21

       I don't disagree. I don't disagree.

   24. JCJason Calacanis47:23

       They huma-

   25. DSDavid Sacks47:23

       You can only get so much blood from a stone.

   26. I don't-

   27. And the re- and the reality is, is that when you try to tax rich people in a confiscatory way, they spend a lot more on lawyers and accountants to figure out how to structure their income in a way that-

   28. JCJason Calacanis47:36

       Or they leave.

   29. DSDavid Sacks47:36

       Yeah, or they leave. Yeah.

   30. JCJason Calacanis47:37

       Or they leave. Or they tap out. You know, like if you're paying 50, 60% taxes, like, what's the incentive to go to work? And if you're sitting on a big nut, you can just be like, "No, you know what, I'll just, I'll just enjoy my life a little bit more because each incremental dollar, 50, 60%'s going to taxes and then I have to pay my team." And you just sort of get diminishing returns.
5. 57:46 – 1:06:20

   Nvidia up 30% due to massive Q2 guidance, rebuilding and upgrading cloud infrastructure, understanding phases of value creation in technology

   1. CPChamath Palihapitiya57:46

      modern history?

   2. JCJason Calacanis57:46

      Well, and now it's going to be, CHIPS is the modern resource. Nvidia shares jumped as much as 30% after reporting huge revenue guidance due to AI demand. Q1 revenue, 7.2 bil, up 19% quarter over quarter, not year over year, quarter over quarter. Revenue beat analyst assessments by more than 600 million. For people who don't know, Nvidia is working on GPUs as opposed to CPUs, the graphic processing units that are being leveraged in AI, and there is a massive cycle going on. There is a line out the door (fingers snapping) to buy these. When you see Twitter going into AI, Facebook, Google, and obviously Microsoft, all the cloud infrastructure is moving from CPU to GPU. Yes, uh, Friedberg?

   3. DSDavid Sacks58:32

      Well, I mean, I was talking with the CEO and CFO of a major data center REIT, and they shared with me that they're seeing more demand in the last couple of, um, months than they've seen in the prior 10 years. Almost all of that data center, uh, build-out demand, so they'll build data centers for software companies, for internet companies. It's coming from, you know, GPU racks. And these GPU racks are much more energy intensive, much more costly, but it's a pretty, uh, kind of significant shift underway that businesses that historically didn't even operate their own data centers are now building out their own data centers to have their own training systems, to have their own infrastructure, to be able to run oth- AI applications and tools. So it's, uh, it's a pretty significant shift underway. All of the growth that Nvidia is projecting and highlighted in this earnings report yesterday is coming from their, their data center line of products, and it's a pretty significant... I mean, I don't know. People have said they've never seen a beat like this or never seen an up- a guidance update like this of this scale, where I think the, the Street was looking for maybe a $7 billion guide and, on revenue for this, and they bas- they came out and said, "We're going to guide to 11 billion next quarter," which is just an insane bump for a, a 90-day outlook.

   4. JCJason Calacanis59:51

      I was told from one of their major customers that they had to beg, uh, like get on their knees and beg for thousands of GPUs and, like the lobbying effort. He said there's a line around the corner to buy these. And you, you need only use ChatGPT or any, uh, Stable Diffusion, et cetera, and you see how long it takes to do, uh, generative AI to use these products. It's like we're back to dial-up modem sacks. (laughs) We're, literally, we're waiting for a computer to give us an answer. When's the last time that happened, that we had to sit there and wait for it to do its job? And I think this is the great renewal for America and other amazing American companies. Only three decades old. Its best years, its best decades are in front of it, obviously. This is gonna be a massive boon for-

   5. DSDavid Sacks1:00:37

      Yeah. And-

   6. JCJason Calacanis1:00:37

      Not only Nvidia, but for America, or as I say, 'Merica.

   7. DSDavid Sacks1:00:41

      Yeah, Nvidia has basically joined the trillion dollar club now in terms of market cap companies. It's really amazing. I mean, I'm kind of kicking myself because this was the easiest buy ever.... ChatGPT launched on November 30th. We all saw that that immediately ushered in a whole new era in Silicon Valley. There's a ton of VC funding that's poured into AI startups. They all need to train models and those models need to use GPUs. So, we all saw this coming and, yeah, I'm kind of annoyed with myself I didn't buy NVIDIA but-

   8. But Zach, let me, let me ask you, let me ask you a question. So, you say that you kick yourself. Obviously, the general statement that, you know, AI is coming and NVIDIA's gonna be a beneficiary or NVIDIA's products are gonna see a boon makes sense. But when you look at the valuation of the business, they are currently trading at 70 times the next 12 months EBITDA. So, how do you think about valuation? Even at a trillion dollar market cap, the, at a trillion dollar market cap, they're trading at 70 times next 12 months EBITDA. You know, this seems like it could take-

   9. JCJason Calacanis1:01:38

      They're doubling revenue every six months though, dude. They're doubling revenue-

   10. DSDavid Sacks1:01:41

       Well, they, I, the, there is the big question here-

   11. How high does it go? 'Cause at a trillion dollars-

   12. JCJason Calacanis1:01:46

       Who knows?

   13. DSDavid Sacks1:01:46

       ... what's the right EBITDA level for a business to be worth a trillion dollars? Is it 100 billion?

   14. JCJason Calacanis1:01:50

       It depends on growth.

   15. DSDavid Sacks1:01:51

       Right, is it 100... But I mean, what's the number? And then the question is, if it's 100 billion, how many years does it take them to grow into that and, you know, are you really paying the right p- price or are you paying a premium to get, you know, uh, ownership in this business today?

   16. JCJason Calacanis1:02:01

       It's a, it's a momentum stock now and will be determined-

   17. DSDavid Sacks1:02:04

       Right.

   18. JCJason Calacanis1:02:04

       ... if they have competition. So is, is there competition for this company? Chamath, do you think there's a competitor that will emerge that, you know-

   19. CPChamath Palihapitiya1:02:10

       Yeah, I mean, it's important to understand what a GPU is maybe.

   20. JCJason Calacanis1:02:13

       Sure.

   21. CPChamath Palihapitiya1:02:14

       So, Intel had the run of the place for the first 40 years of compute because it turned out that most of the things that we used it for, Excel, Microsoft Word, a browser, operated well on a CPU which, essentially think about it as like a factory that takes in the first order and then puts it out, first in first out. And the great thing about GPUs is that it can take multiple streams of work at the same time and work on them at the same time, right? So it's very parallel and has this level of parallelism that makes it very well suited for AI applications. I think the thing to keep in mind is that it is a byproduct of a GPU that tries to also do other things. And so as a result of that, you are now seeing a lot of companies building their own silicon and most importantly, all the big tech companies now have some pretty well evolved efforts underway. So a lot of these companies have figured out how to do custom ASICs that can do this massively parallel processing and what you're now seeing is chips that are designed against specific models that are optimized for them. The other thing that you're also seeing is that some people are saying, "Well, you know what? For these massive models, actually, you should just run it all in memory." And so you're having folks that are doing it in, in massive arrays of FPGAs, that was Microsoft's first attempt at all of this. So what is the point of me telling you this? I think that, again, we talked about this last week, the biggest cheerleaders of this first point of value creation has really been Wall Street and family offices that wanted to front run where the value creation was gonna initially go. And they've been right, which is around chips. But there was a tweet, and Nick, I posted it, maybe you can throw it up here, that shows that if you compare this to the mobile internet, there's always this phased approach in terms of value creation where let's just say, initially in a new market, mobile a decade ago and AI today, the first dollar of profits tends to go to the chip companies. That makes a lot of sense, right? Because they're the ones that are in the bowels of making the elemental capabilities possible. And then you transition that value and what Friedberg says is people realize, "Hey, hold on, the profit dollars are not going to accrue there," because again, this beautiful principle of capitalism is that you can only over earn for a certain amount of time because then competitors emerge and say, "Hold on, I wanna steal those profit dollars from you and take them for myself." So margins compress, right? So in the end, NVIDIA's gains today will be then spread across NVIDIA, Facebook will have their own chip, Amazon will have their own chip, Google already does. Apple will have their own chip. All the memory companies will be in this space, right? So then the profits get smeared there, multiples compress. Then where does the value grow? To the device companies in the mobile internet. Here I think we still have to debate what is a device company in the world of AI? But the most important thing I think to remember is that where the real value gets accrued is five, six, seven years later when the software and services companies show up and create a huge moat. And those are the Googles and the Facebooks and the Apples of the world. And so it's a really dynamic moment. I think it's wonderful for NVIDIA. It's an amazing story for Jensen, who's been really at this game for a very long time. By the way, there's a, Jensen has a law of his own that is a sort of companion to Moore's Law called Wang's Law, named after himself, which you can read about, which just talks about the variability of the compute capabilities of GPUs versus CPUs, so if you wanna know that, he should get some credit for that. But I think it's great. I think we're in the inning one-

   22. JCJason Calacanis1:05:56

       And by the way, Apple is also making their own GPUs. That's why when you hear of the M2, that has GPUs in it, so you're absolutely correct that-

   23. CPChamath Palihapitiya1:06:03

       We're in inning one.

   24. JCJason Calacanis1:06:04

       We're in inning one.

   25. CPChamath Palihapitiya1:06:04

       So we're gonna, we're gonna have a few, we're gonna have a few quarters for sure of this hype.

   26. JCJason Calacanis1:06:08

       Mm-hmm.

   27. CPChamath Palihapitiya1:06:09

       And then the smart money will probably figure out where the next lily pad is and then they'll go to the next lily pad.

   28. JCJason Calacanis1:06:15

       Beautifully said. Beautifully said. Any other thoughts, uh, Friedberg, as we wrap th- up

6. 1:06:20 – 1:11:43

   Adobe's new AI product, why they might have overpaid for Figma

   1. JCJason Calacanis1:06:20

      on that? I wanted to show this AI demo from Adobe Photoshop. You know, every week or so we see something that's insane.

   2. CPChamath Palihapitiya1:06:26

      This is incredible.

   3. JCJason Calacanis1:06:27

      And this one was, we shared in the group chat, but for those watching-

   4. CPChamath Palihapitiya1:06:30

      John Belsky put this out. Shout out to Belsky.

   5. JCJason Calacanis1:06:32

      Yeah. So what we see here, if you're listening is, uh, you know, taking a Photoshop, creating an area, and then instead of like cutting and pastes- pasting or refining it, you're putting in a text prompt and saying, "Oh, put um, uh, expand the, and make this wide screen or let me grab this deer out of a forest and then put it on a wet alley at night. And uh, then you know, let me highlight this wall over here-"

   6. CPChamath Palihapitiya1:06:59

      It's incredible.

   7. JCJason Calacanis1:06:59

      "... and put a sign."... and put a red arrow sign. And it just generates it and it's doing this, they made specific note when they launched this, that all of this is done with stock photography, they have the complete license to, so they can monetize this without getting sued like Microsoft is currently being sued and Microsoft actually, in addition to the GitHub lawsuit, it tu- turns out Twitter sent a letter over to, uh, Microsoft as well. So incredible demo. The producers here at All In as our production team grows, Producer Brian made a little video here. Here's Chamath in, before the Lora Piana days.

   8. CPChamath Palihapitiya1:07:41

      That's, that's Tom Ford, yeah. That's old.

   9. JCJason Calacanis1:07:43

      It's Tom Ford. This is when he-

   10. CPChamath Palihapitiya1:07:44

       Bad hair. Look at that terrible hair.

   11. JCJason Calacanis1:07:46

       The hair's terrible. You look tired. Uh, the watch looks great.

   12. DSDavid Sacks1:07:49

       The watch is good.

   13. JCJason Calacanis1:07:49

       What's the watch?

   14. CPChamath Palihapitiya1:07:50

       No, I had a long night that night.

   15. DSDavid Sacks1:07:52

       I think the hair's good.

   16. CPChamath Palihapitiya1:07:52

       My God, I played poker.

   17. DSDavid Sacks1:07:54

       (laughs)

   18. CPChamath Palihapitiya1:07:54

       Literally, I walked into that, to that CMBC head from poker.

   19. JCJason Calacanis1:07:57

       Oh my god, really? You went straight from the poker table to the-

   20. CPChamath Palihapitiya1:07:59

       That was brutal. Brutal.

   21. JCJason Calacanis1:07:59

       ... to the, uh, squawk box deck?

   22. CPChamath Palihapitiya1:08:01

       Yeah, brutal.

   23. DSDavid Sacks1:08:02

       Whoa, whoa, whoa. What's going on with the pin? Wait, no tie, but he's wearing a tie and pin?

   24. CPChamath Palihapitiya1:08:06

       Is that Tom Ford? Yeah, yeah. He said to, he said, "Chamath try this without the, without a tie." I said, "Okay."

   25. JCJason Calacanis1:08:10

       Oh, when you were talking to Tom Ford? Great. What a flex.

   26. CPChamath Palihapitiya1:08:13

       Hey, you know, Tom Ford has stopped me twice in my life. I can tell you both stories if you'd like.

   27. JCJason Calacanis1:08:16

       Okay, ho- hold on a second. He t- oh, so degenerative AI add, make a yellow sweater is the prompt here. Let's see if he goes to BERT, from Ernie to BERT. There he is, Sri Lankan BERT, as we call him in the, uh, poker room.

   28. CPChamath Palihapitiya1:08:29

       Not that good.

   29. DSDavid Sacks1:08:30

       Yeah, it doesn't quite get the borders right, but-

   30. CPChamath Palihapitiya1:08:31

       Looks terrible.
7. 1:11:43 – 1:33:36

   State of Silicon Valley, VC, and startups; dealing with failure

   1. CPChamath Palihapitiya1:11:43

      I had.

   2. JCJason Calacanis1:11:43

      Well, so we're here, let's talk a little bit about the state of Silicon Valley. Sax, we were talking offline, uh, after the show last week. You've, uh, slowed down investment at your firm, Craft, you're being thoughtful, you're working on the existing portfolios. How would you describe your activity so far in the first half of 2023 as a firm, if, if you're so willing to share that?

Episode duration: 1:37:42