Liberation Day, Tariffs, US v China Open Source, OpenAI, $CRWV, TikTok

00:00:00.000 | A riddle for you before we move on.

00:00:02.360 | We'll do Salesforce, Netflix, Square, Amazon, Palo Alto Networks, Facebook, Snap, Proofpoint, NetSuite, and CoreWeave have in common.

00:00:11.500 | No idea.

00:00:13.320 | They all broke issue.

00:00:14.840 | Oh, wow.

00:00:28.560 | And we're back.

00:00:29.700 | Bill, great to see you.

00:00:30.660 | Good to be seen.

00:00:31.660 | I mean, you have to be pretty stoked coming off those wins last weekend in San Francisco.

00:00:38.160 | Yeah, I'm repping the Gator hat still.

00:00:40.980 | I'd say we kind of eked by for those that did watch the game.

00:00:45.580 | That was an incredible final few minutes.

00:00:49.040 | Explain it.

00:00:49.680 | Take us through the final few minutes.

00:00:51.440 | Well, I mean, to be honest, I was afraid.

00:00:54.520 | Like, I didn't think there was a chance at this point because they were behind by so much as you were heading into the end of the game.

00:01:03.000 | And they basically, you know, scored four three-pointers, you know, against zero from the other side.

00:01:10.420 | They intentionally fouled and had missed free throws.

00:01:13.820 | So, I think someone said the Yahoo game predictor had it at, like, 98% Texas Tech with, like, very little time left on the clock.

00:01:24.960 | And they somehow eked it out.

00:01:26.600 | Now, the bullish people will say, oh, you lived through something like that.

00:01:30.560 | Now, you have confidence to deal with anything.

00:01:33.100 | But, you know, the odds makers have put Duke in front of Florida at this point.

00:01:38.140 | And that started in the opposite place.

00:01:40.460 | So, anyway, one game at a time.

00:01:41.980 | Super exciting.

00:01:43.240 | I was out in San Francisco at the Chase Center.

00:01:45.560 | Got to hang out with the coach a bit and some of the players' parents.

00:01:49.840 | And it was a good trip.

00:01:51.880 | And now, since I'm in Austin, it's right down the street here to San Antonio.

00:01:56.760 | So, a lot of friends coming in.

00:01:58.600 | And we'll see what happens.

00:02:00.420 | It's going to be a heck of a weekend.

00:02:02.680 | And maybe if a certain couple teams end up in the championship game, I'll sneak in there on Monday night.

00:02:08.520 | I guess as long as we're repping and rolling, you know, I have two degrees, one from Florida and one from Texas.

00:02:15.080 | The Texas ladies are playing in the Final Four in Tampa.

00:02:19.160 | So, a lot of good stuff happening.

00:02:20.680 | Speaking of a lot of good stuff happening, I think there are some people in the world that think a lot of not-so-great stuff is happening

00:02:27.080 | based upon market reaction to the president's announcements of these tariffs.

00:02:31.700 | So, why don't we kick it off talking about Liberation Day?

00:02:34.260 | Yeah.

00:02:34.580 | So, we're recording this right after Trump's presentation.

00:02:38.200 | I think you and I both watched it and then we jumped on the pod.

00:02:41.280 | You've been talking about this.

00:02:43.440 | It's obviously been choreographed that this was coming.

00:02:47.680 | And, you know, you've been saying for a very long time that Trump and his team are very serious about this,

00:02:55.300 | rather than the argument that, oh, it's just a means to an end, you know, and a way to get a negotiation started.

00:03:02.880 | You have made the point that you believe that they believe this is actually where we need to take the economy.

00:03:09.140 | And as such, you've been conservative and worried about where this would go.

00:03:14.700 | You gave a presentation last week on how to think about this.

00:03:18.860 | What did you talk about there?

00:03:20.400 | I think it was at the J.P. Morgan Tech Conference.

00:03:22.460 | It was pretty clear to me and you, we were talking about it in early Feb.

00:03:26.640 | You know, this was doctrinal.

00:03:28.240 | This is, you know, there was a philosophical belief around trade that they wanted to, you know, create a more fair and level playing field.

00:03:35.680 | And the real debate, you know, has been going on is like how big?

00:03:39.400 | And there are a couple of different camps.

00:03:40.640 | And so J.P. Morgan had this great event in Montana last week, 100 tech CEOs.

00:03:44.680 | But, you know, they had Howard Lutnick, Elon Sachs, Doug Bergram, all talking about, you know, various aspects of this.

00:03:52.420 | And they asked me to do a little bit of a presentation on decoding the Trump economic agenda.

00:03:56.180 | And really, it boils down to this, Bill.

00:03:58.760 | You know, at the top, I think that all the CEOs in the room are pretty excited about the golden age that people have been talking about.

00:04:05.960 | You know, a pro-growth administration, pro-business, pro-investment, lower taxes, less regulation, pro-M&A.

00:04:12.820 | We've seen this M&A flywheel starting to kick up, this AI super cycle.

00:04:16.640 | But everybody's been pretty terrified about these tariffs.

00:04:19.520 | And the real question going into today, Liberation Day, was were tariffs going to land closer to the $600 billion trillion tariff level that Peter Navarro had been talking about and Howard Lutnick had been talking about?

00:04:31.960 | Or maybe at a little bit lower end of the spectrum, which we heard a little bit more from Scott Bessent and Kevin Hassett.

00:04:39.220 | And I think everybody was holding their breath.

00:04:41.240 | Well, we got the answer.

00:04:42.400 | We got the answer today.

00:04:44.100 | And so I was framing that, you know, at the JP Morgan conference.

00:04:47.600 | I showed this slide you'll get a kick out of.

00:04:49.460 | You know, I ended the presentation, you know, with two planes coming in for landing.

00:04:55.400 | They both actually land, believe it or not, but it's like the glide path that we land here with tariffs and budget cuts matter a lot.

00:05:02.800 | And so, you know, we got the news tonight.

00:05:05.620 | You're right.

00:05:06.320 | We just we just listened to the president talk, you know, and he came in on the larger end of this.

00:05:11.120 | I mean, there's no other way to slice it.

00:05:12.980 | There was a there was a headline that hit right after the market closed from the Wall Street Journal, I believe, that it was 10 percent across the board and the markets jumped up like two and a half percent.

00:05:24.320 | Right.

00:05:25.360 | And then as the presentation started unfolding, they started coming in.

00:05:28.980 | People saw this chart that they presented on reciprocal tariffs, right, where they say tariffs on China going to 54 percent.

00:05:37.580 | Can you believe that?

00:05:38.360 | Thirty four percent on top of the 20 percent that already exists.

00:05:41.480 | And he starts going through this list and the futures, the S&P futures, the Q, the Nasdaq futures start sinking.

00:05:48.040 | They had a 600 basis point fall between where they initially jumped and where they ended up.

00:05:54.560 | So the market is not liking this at all.

00:05:56.800 | And depending what index you were looking at, we were already down, you know, eight to 15 percent on the year.

00:06:03.580 | Now, whatever comes in tomorrow will be on top of that.

00:06:06.860 | So that's the chart.

00:06:08.960 | That's kind of the initial reaction out of the market.

00:06:11.600 | We can we can break them down a little bit if you want.

00:06:14.340 | But that was the initial reaction.

00:06:16.240 | Right.

00:06:16.820 | And one of the pieces that I, you know, witnessed and I think everybody else did as well, maybe you have more data on it,

00:06:23.760 | is when they said reciprocal tariff, they kind of redefined what a tariff is for the reciprocal calculation.

00:06:32.300 | And they are including others, whatever.

00:06:35.880 | I don't know what else goes in there.

00:06:37.420 | Maybe you know exactly what what fills this in.

00:06:39.940 | But but but basically the Trump administration is calculating an effective tariff, if you will, for each and every other country, which may not be the explicit tariff.

00:06:49.720 | Correct.

00:06:50.780 | I mean, they call them non tariff trade barriers.

00:06:53.400 | This is everything from what they call currency manipulation to things like judicial actions that restrict free and fair trade of our products into their countries.

00:07:02.540 | And by the way, we know there are non tariff trade barriers.

00:07:05.220 | So, like, it's not totally surprising.

00:07:07.680 | But if you and I were to do the math on these, you can kind of make those numbers whatever you want to make them.

00:07:12.680 | And so, you know, if I had to go through this, the tariffs largely break down into let's call it three or four big buckets.

00:07:20.520 | Right.

00:07:21.320 | There's the auto tariffs.

00:07:22.560 | That's largely on Mexico, Canada and Germany.

00:07:24.960 | And the auto tariffs were imposed at 25 percent.

00:07:28.300 | This, you know, in fact, he had 20 members of the UAW union in the front row at the event.

00:07:35.620 | He invited, I think, the president of the UAW up on stage to make some comments.

00:07:40.280 | And he said these people used to be Democrats.

00:07:42.540 | The Republicans now were the only ones who fight for them.

00:07:45.500 | And by the way, this is really, you know, he said we won the state of Michigan because of this.

00:07:50.300 | This is what I campaigned on.

00:07:52.080 | You know, these are the promises we made and we're delivering on the promises.

00:07:55.540 | It is striking to me that just politically, this is what Democrats were running on 20 years ago.

00:08:01.500 | And it just shows how much the political parties have changed.

00:08:04.300 | So this is a big, you know, tariff differential as it relates to autos.

00:08:09.420 | Then he had the reciprocal tariffs, which are the ones that I outlined here, Bill.

00:08:13.660 | Remember, Trump is the negotiator in chief.

00:08:16.660 | This is the starting point.

00:08:18.220 | All these tariffs go into place and we'll put these charts up on April 9th, country by country.

00:08:23.080 | And so we're going to hear all these ad hoc negotiations going on, some of which I'm sure, like Vietnam, he'll declare victory on even before we get to April 9th because they've already capitulated on a bunch of tariffs.

00:08:35.360 | He's also declared that there are $6 trillion of new investments, right, that people have committed to in the United States.

00:08:42.620 | He mentioned NVIDIA, Apple, TSMC, SoftBank, OpenAI in his remarks.

00:08:48.700 | In fact, I particularly noted when he talked about SoftBank and OpenAI, he said great companies.

00:08:54.880 | So, you know, for the people who are, you know, watching the battle between OpenAI and NX, that was notable.

00:09:01.600 | And then he said, we're going to have a minimum tariff of 10% on all countries.

00:09:05.920 | So even if you're not on this list, we're going to have a minimum tariff of 10% on all countries.

00:09:10.700 | And then, of course, China is kind of in this bucket on its own, right?

00:09:14.160 | That's going to be a huge negotiation on its own.

00:09:17.240 | There are a lot of things that go into that negotiation, everything from, you know, the Panama Canal to the TikTok stuff.

00:09:23.780 | So set that aside, if you will, for a second.

00:09:26.160 | There's no way that lands, I think, at 54%.

00:09:29.780 | That's the headline tariffs, okay?

00:09:32.160 | When we do the math and we add all of these up and, you know, say, what does this come up with, okay?

00:09:39.320 | The headline is we were at $77 billion last year, and we end up at about $750 billion, right?

00:09:47.060 | Remember, Peter Navarro, the hawk, the person who had been saying we're going to land big tariffs, he was estimating $600 billion.

00:09:54.400 | So this definitely landed on the larger side.

00:09:57.060 | But then they came out right after that, and there was a footnote about things that were exempted from the tariffs.

00:10:02.220 | Exemptions.

00:10:02.840 | Exemptions, which included pharmaceuticals and, notably for you and I, semiconductors.

00:10:09.020 | Wow.

00:10:09.620 | Right?

00:10:10.020 | So Taiwan's got a 32% tariff, but semiconductors are exempted.

00:10:15.900 | And so we're going through the value of those exemptions right now.

00:10:19.880 | My hunch is that this is going to land right around $600 billion.

00:10:23.040 | But, Bill, I have to ask you, you know, here we are.

00:10:27.940 | You and I, we're trying to make our way through this, make sense out of it.

00:10:32.760 | I don't think there are many CEOs we know who support this or like this.

00:10:36.780 | In fact, I think a lot of congressional Republicans don't like this.

00:10:39.940 | You've made an eloquent defense of, you know, the benefits of free and largely fair trade.

00:10:46.660 | When you start hearing things like this, like, okay, this category got exemption, you know, or this category got exempted.

00:10:53.580 | But, like, just give me your reaction, right?

00:10:57.220 | As somebody who I think totally understands the benefits of free trade, when you see the Republican Party doing this, you know, how does it make you feel?

00:11:07.180 | Well, yeah, I mean, at a high level, I'm a believer in open markets, free trade, and comparative advantage.

00:11:14.000 | And that's been studied for a very long time.

00:11:19.280 | There are very solid mathematical arguments why, if you pull up the trade walls between multiple countries, that you're going to hurt the efficiency of both of them in the long run.

00:11:32.120 | And I, you know, at least from a theoretical perspective, I'm a believer in that.

00:11:35.940 | I think there's another issue for the markets and for the CEOs, which has to do with both the slow pace at which they could realistically respond to this,

00:11:49.280 | And then the amount of ambiguity that's been out there.

00:11:52.460 | And so, let's say what the administration wants to encourage is for you to relocate a factory that you have, or let's call it production, that you have in Thailand, you know, and put it on the American shores.

00:12:08.320 | That's not a quick process.

00:12:10.120 | And if you start that process today, it might take three years, you know, before you'd have the type of volume that'd be capable of bringing that back.

00:12:20.360 | And probably at a higher cost.

00:12:22.380 | I mean, one of the things that I've said over and over again, I don't think our labor is globally competitive, nor do I think it wants to be.

00:12:30.320 | And so, even bringing it back, you're going to have a higher cost of production because we're going to have a higher cost of labor.

00:12:36.600 | That being said, you know, this ambiguity, you know, there's a lot of people even going right into this announcement that didn't know what percentage of it was real versus bluster.

00:12:48.200 | And there's, I think, at this point, just reading the papers, not making my own assessment, the administration has a reputation of maybe some of this is for negotiation, maybe some of it's not real.

00:13:01.600 | And so, you're left not knowing what's going to be the policy three months from now, six months from now, 12 months from now, which makes it very hard to allocate CapEx in any meaningful way whatsoever.

00:13:13.980 | I think it's so well stated.

00:13:16.260 | You and I have said, you know, markets and business abhors uncertainty.

00:13:21.980 | Yeah.

00:13:22.560 | Right?

00:13:22.940 | It can deal with almost anything, but it has to have predictability.

00:13:26.380 | So, it can build a forecast.

00:13:27.880 | So, it can look at an investment and say, is that MPV positive?

00:13:32.280 | And I was literally texting with some big CEOs during the president's announcement, and they are asking questions.

00:13:40.280 | Do you see us exempted?

00:13:41.680 | You know, are we in there?

00:13:43.120 | And this is just amazing to me, right, that you have this level of uncertainty.

00:13:47.580 | You know, I was with those 100 CEOs in Montana last week, and I would say almost to a one, they said things are slowing down in February and March because nobody knows what to do.

00:13:57.840 | And remember, you know, the Fed had just come out last week, had taken down their forecast for GDP growth, had taken up their forecast for inflation, had taken up their forecast for unemployment by the end of the year.

00:14:11.940 | So, the economy, you know, most major economists are increasing their probability of recession, are slowing the rate of growth.

00:14:24.580 | And I think your point is, can I count on this, and how long can I count on this, and can I really plan a year out based upon this, or is this going to change yet again over the course of the year?

00:14:45.040 | So, and by the way, there's cascading effects.

00:14:47.960 | So, if you're unsure about things like this, you're not going to hire a bunch of people, for example.

00:14:55.240 | You're probably going to pause hiring because you don't know, you know, you don't know how much earnings you're going to have.

00:15:02.040 | So, those kind of things can proliferate downward and affect unemployment and eventually affect consumer spending just by freezing, you know, nearly everything in the economy.

00:15:15.160 | Let me tell you two other cascading effects.

00:15:17.640 | I heard from one of the CEOs last week that four contracts with him had actually been canceled, right, because they had like three European contracts canceled and one Asian contract canceled because the countries were so upset with America.

00:15:31.920 | You know, that they're going to do deals instead with European countries, companies, or whatever the case may be.

00:15:38.420 | You know, and then I saw a couple of tweets, Bill.

00:15:40.440 | One was that, you know, China and Korea and Japan were actually going to collaborate in a response to the U.S. tariffs.

00:15:48.420 | And somebody said, you know, we haven't seen the Koreans and the Japanese and the Chinese combined forces on anything since the Mongols caused them, you know, to get together.

00:15:58.200 | And so, it is causing a lot of strange bedfellows.

00:16:00.860 | And I sent you the tweet where the Europeans, the president of Europe, the president, Macron, they're all going to China and they want to talk about closer trade negotiations with China.

00:16:12.220 | And you warned us about this.

00:16:13.820 | I think you said the meeting was in Vietnam.

00:16:15.700 | And obviously, I got that from the president of the economists who had predicted that that would happen.

00:16:20.840 | But, yes, you know, there's no reason that that wouldn't happen.

00:16:24.420 | And there's a lot of ramifications of this.

00:16:28.200 | I don't think that anything came out of it that's going to be good for the markets and good for stocks.

00:16:37.040 | But that's more your world than mine.

00:16:38.460 | I've always shunned macro analysis.

00:16:40.800 | Let me just maybe opine on that for a second.

00:16:43.900 | Like, where do I think we go from here?

00:16:45.500 | After hours with the NASDAQ down, peak to trough post-Trump at almost 18%, right?

00:16:52.420 | That's down a lot.

00:16:53.560 | A lot of names in the NASDAQ are down 40% or 50%.

00:16:57.500 | So we're starting to get some of that fear into the market.

00:17:00.740 | Somebody asked, and I said, as very rude, I do believe the president wants to do deals.

00:17:05.600 | He believes in fairer trade.

00:17:07.840 | I think we're going to land the plane here closer to $300 billion or $400 billion in tariffs, not $600 or $700 billion, and certainly not a trillion in tariffs, even though it feels today like it was bigger than that.

00:17:19.240 | And one of the things I think that's going to kind of force the president's hand, he talked at the press conference.

00:17:25.540 | He had a bunch of senators there and House members.

00:17:27.420 | The senators and House members are hearing from their constituent CEOs that they don't like these tariffs.

00:17:34.080 | And remember, most important to the president, he wants to get this reconciliation package passed, which he calls a big, beautiful bill.

00:17:41.520 | He wants to get this thing passed, which puts in place no taxes on tips and the permanency of the tax cuts, which he passed in his first administration.

00:17:50.720 | He can't afford to lose a single Republican vote.

00:17:54.240 | And so I think that that also is going to guide him a little bit more to the center.

00:17:58.760 | And that's what, you know, we'll see whether the market believes that.

00:18:02.800 | Certainly didn't believe it after hours today.

00:18:04.940 | We're going to get a little bit more positive on the companies we like the best because we think some of this fear is now getting priced in.

00:18:11.040 | What will you be looking for in the next 30 to 60 days as this plays out?

00:18:17.380 | Yeah, I think we're still in the fog of war, certainly.

00:18:22.040 | But I will be looking at, you know, do these exemptions on things like semiconductors and pharmaceuticals hold?

00:18:29.020 | Are we seeing the country by country renegotiation on some of these things?

00:18:34.200 | And probably most importantly, Bill, it's really about China.

00:18:37.420 | China is the second largest economic power in the world.

00:18:43.320 | It scares me how big the tariffs are that we are suggesting are going to go in place on China.

00:18:50.160 | And I think, you know, it's imminent that he and she are going to have to talk and get a big trade deal done.

00:18:55.140 | And so those are the things I'm going to be watching for.

00:18:57.760 | I don't think I see any clearing event here for at least another 30 to 60 days.

00:19:03.680 | But remember, the best opportunities to buy something are when people are a little fearful.

00:19:08.400 | So you may have to just, you know, take a bit of a leap of faith on this one if you want to be if you want to purchase at the best prices.

00:19:16.240 | Makes sense.

00:19:18.240 | You know, another thing that we heard a lot about this week, Bill, speaking of China, is, you know, some developments in Chinese open source and some developments on the U.S. open source front, particularly with respect to these frontier models.

00:19:32.140 | You have a lot of, I think, understanding about kind of the history in China around open source as well as around the history of the United States around open source.

00:19:42.320 | So help us unpack, if you will.

00:19:44.660 | What do you think strategically is going on in China with respect to these open source models?

00:19:50.480 | I've seen some people tweet that maybe that, you know, DeepSeq was forced into open source by Xi.

00:19:56.920 | Do you think that's going on or is there something else going on here?

00:20:00.440 | And by the way, I mean, all this culminated in the past week with open AI moving, you know, are talking again about open weighted models, which is a, I think, a really important data point.

00:20:11.840 | So how did how do we get from where we were to where we are?

00:20:14.760 | So, yeah, I read that same tweet and I think it was remarkably misplaced.

00:20:19.420 | China has been supportive of open source for well over a decade now.

00:20:24.260 | If you look on most of the major open source products and look at the management page and who the sponsors are for these, like Linux, you know, you'll see many of the major Chinese companies have been there and been supporting it for a while.

00:20:39.100 | Why?

00:20:40.500 | They've been accused of stealing tech IP for years.

00:20:44.420 | And so when something like open source comes along, this looks like the best thing possible, right?

00:20:51.720 | There's no one that can accuse us of IP theft because there is no IP ownership in an open source world.

00:20:57.880 | And so having dealt with those accusations for probably 40 or 50 years, I think everyone in China, you know, the government and the entrepreneurs writ large view open source as a very positive movement for their country relative to the West.

00:21:14.840 | And so they've been they've been in on it for a long time.

00:21:17.600 | They're very adept at it.

00:21:19.700 | They're very they're very big believers in it.

00:21:23.160 | You know, when we talked about the interview with the DeepSeek founder, I would say he had as much kind of emotional like his his entire emotional mindset was tied to open source.

00:21:37.180 | Like he believes in it and wants to support it.

00:21:40.400 | So that's I think that's an important backdrop.

00:21:42.340 | So I don't think China got there in some calculated way or do I think it was some recent move.

00:21:50.640 | I think they embraced open source, you know, over a decade ago because it made a ton of sense for them in a world that had had pointed a finger at them from an IP standpoint.

00:22:02.320 | So let me just double click on that.

00:22:04.420 | So basically what you're saying is they may have been fearful that they would have been cut off from other type of software products in the United States.

00:22:13.160 | Like, you know, there might have been export controls or other things put on them.

00:22:16.900 | So they said, I may as well support Linux because I may not be able to get Windows.

00:22:20.760 | Right.

00:22:21.160 | But I think that that discussion happened a long time ago.

00:22:25.800 | It wasn't recent, but I think it's important because that lays the foundation, right?

00:22:31.440 | That if that's a foundational belief among Chinese entrepreneurs and Chinese companies, then it's understandable that this new generation of entrepreneurs might also see the advantages of open source.

00:22:42.340 | Another thing that I think people have to remember is that within also within the past decade and maybe 15 years, many U.S. companies have learned to use open source as a defensive tool rather than just an offensive tool.

00:22:58.740 | And this is a more.

00:23:00.180 | Yeah.

00:23:00.540 | So this is the biggest companies out there.

00:23:02.960 | If they get in a position where they feel like they're behind the eight balls, so they're not in a leadership position, they will embrace open source as an attempt to level the playing field.

00:23:13.880 | And a great example is Kubernetes.

00:23:16.600 | So Amazon took this huge lead with AWS in the hosted server business.

00:23:23.500 | Everyone was afraid of that.

00:23:25.420 | Google had a piece of technology called Kubernetes.

00:23:29.200 | It was orchestration that would allow you to move a workload if that became a standard from one large server vendor to another, right?

00:23:38.120 | It basically created ease of distribution.

00:23:41.440 | So you could run on multiple clouds.

00:23:43.420 | They went to the Linux Foundation.

00:23:45.840 | They recruited IBM, a whole bunch of other people, got everyone behind it.

00:23:49.960 | And it gained so much momentum that Amazon had to embrace Kubernetes.

00:23:54.980 | So it worked and we don't have a monopolist in that cloud business right now, you know, perhaps because of that deft move made by Google.

00:24:05.020 | But they did it with Android against Apple, you know, being notable one.

00:24:10.120 | And Meta did it with Llama here, right?

00:24:13.980 | They came to the table.

00:24:15.760 | They weren't first to the table in the AI space, but they were disruptive with open source.

00:24:20.540 | But one other thing I would point out about that type of move and attention, in addition to saying it's defensive, I think it's great for consumers.

00:24:28.980 | Like if you study economics in business school, you know, there's this notion of pure competition.

00:24:36.340 | Like where do you have the most fluid competition, which leads to the lowest prices for a consumer?

00:24:43.300 | And certainly open source does that versus proprietary code.

00:24:47.260 | Like it's just hyper competitive.

00:24:49.560 | And that's why it's disruptive.

00:24:51.000 | And that's why people use it in this way.

00:24:52.980 | So that's a huge backdrop to where we are today.

00:24:55.780 | I believe DeepSeek has been remarkably successful in the enterprise.

00:25:01.200 | And that's, you know, it's hosted by AWS.

00:25:03.420 | It's hosted by Google.

00:25:05.180 | It's being used around the world.

00:25:07.300 | I've heard, you know, from Hugging Face that it's been, you know, forked over 1,500 times on their site.

00:25:13.620 | And so it's prolific.

00:25:15.600 | I'm beginning to hear, you know, concerns that D.C. may take action to limit the use of DeepSeek.

00:25:23.740 | You're saying Washington may intervene to take action because, you know, there are people perhaps lobbying against or other concerns that Washington may have about open source Chinese models being used by American enterprise.

00:25:38.380 | I think it's safe to say both of those things are happening.

00:25:40.600 | There's people that are really concerned about, you know, China technology getting underneath our products.

00:25:47.580 | Whether or not this particular code could be bad or not, they just might have that default.

00:25:53.360 | And then I think there are people that are lobbying because it would benefit their company.

00:25:57.380 | Either way, if that gets traction, you have a window in the U.S. right now where someone might move to go left of DeepSeek in terms of openness on one of their models, either in an offensive or defensive move.

00:26:13.120 | And I think it's a short window.

00:26:14.760 | And so it'll be very interesting from my point of view to see whether Google does that or Meta.

00:26:21.680 | I think they have an announcement coming up in a few weeks of their next model.

00:26:25.040 | I don't think Anthropica would do it.

00:26:27.140 | They've been so anti-open source, it would be very out of character for them.

00:26:31.540 | But it will, you know, it will be interesting to see what happens on this front.

00:26:35.460 | And that leads us up to OpenAI making an announcement, which I'll let you describe.

00:26:40.560 | Well, you know, listen, we've talked here and Sam has been dropping the breadcrumbs on Twitter, right?

00:26:48.260 | That, you know, they wanted to launch an open source model.

00:26:51.400 | They've been GPU constrained.

00:26:52.740 | They've been bandwidth constrained.

00:26:53.800 | Right.

00:26:54.200 | But he got the announcement out this week, which I was thrilled with, where he said, you know, we're excited to release a powerful new open weight language model with reasoning in the coming months.

00:27:04.000 | And we want to talk to devs.

00:27:05.700 | So he's inviting all these developers to participate and give feedback.

00:27:11.040 | He said, we want to make it a very, very good model.

00:27:13.560 | We're planning to release an open weight language model, one of our first since GPT-2.

00:27:18.800 | And he says, we've been thinking about this for a long time.

00:27:22.220 | And the interesting thing is somebody, you know, in the replies I saw, somebody said, are you going to make people buy licenses if they get a lot of users like Meta is doing with Llama?

00:27:33.640 | And he kind of takes a jab at Meta and he says, no, we're not going to.

00:27:37.520 | Well, he says no.

00:27:37.740 | Yes, he said no.

00:27:38.840 | He says no, which indicates maybe we're going to out-open, you know, Llama.

00:27:43.820 | Right.

00:27:44.520 | So that's on the one hand, you know, kind of open AI.

00:27:47.580 | On the other hand, just hold your thoughts in case people don't know, you know, openness is a continuum.

00:27:54.940 | It's not black or white.

00:27:56.280 | And that's true of all the open source technologies.

00:27:58.840 | In the open source model world, some of the players, most notably Meta with Llama, have a usage restriction against the free use of the model at $700 million.

00:28:12.700 | And that's what you were referring to.

00:28:14.060 | And so at least in a tweet, Sam suggested they won't have that in theirs.

00:28:18.980 | So back to you.

00:28:20.780 | Yeah, no, which I think is notable.

00:28:23.320 | Because remember, at the end of this month, we have LlamaCon, which is the developer event for Llama.

00:28:28.540 | It's a big deal for Meta.

00:28:31.300 | The launch of Llama 4 has been rumored for a long time.

00:28:34.980 | And in fact, I think there are just a lot of people who are surprised they haven't released it.

00:28:38.940 | But it seems to everybody, they're going to have to release it ahead of LlamaCon.

00:28:42.320 | What I'm hearing, Bill, is that it'll be a $400 billion parameter model.

00:28:47.360 | It'll be a mixture of expert model using, you know, 50, 60, 70 experts.

00:28:53.500 | It's going to have huge context window, like 10 million context window.

00:28:57.600 | You know, and it's going to launch this month.

00:29:00.020 | I think it's terrific that OpenAI has now fully committed.

00:29:04.240 | You know, everybody on the team, Brockman, Kevin, et cetera, we're all retweeting this.

00:29:08.820 | Fully committed and maybe even suggesting that this is going to be a more capable model and even more open.

00:29:14.300 | I think it's good that we have competition in the U.S. for an open source model.

00:29:18.460 | And when it comes to the administration and what Washington, D.C. is going to do, my best sense is they do not want to see a Huawei DeepSeq Belt and Road, either with chips or with open source models.

00:29:33.440 | They do not want to see the world run on DeepSeq on, you know, Huawei 910 chips.

00:29:39.800 | And so, you know, this gets back to the AI diffusion bill and how we're going to restrict these things.

00:29:45.500 | I think they would love to see the world continue to run on, you know, U.S. compute, U.S. silicon.

00:29:51.920 | And they would love to see Llama and perhaps OpenAI's open source model around the world.

00:29:57.240 | They know it has a lot of distribution.

00:29:58.780 | So I think this was a really positive step forward on that.

00:30:03.380 | And you bring up an important point, which is, you know, you don't, you don't have to be, let's say something does happen to limit DeepSeq usage.

00:30:12.280 | Whoever's going to jump in to try and lead the open source movement in the West, if they want to be a global player, they don't just have to get left of everyone else in the West.

00:30:24.620 | They still have to compete with DeepSeq on a global basis.

00:30:29.260 | And I don't think, you know, it'll be interesting to watch.

00:30:31.960 | I'm very, very curious how this plays out.

00:30:34.380 | I've already asked Clem at HuggingFace if maybe he will create a continuum so we can rank all these different models from an openness perspective.

00:30:43.180 | Because there's so many different facets by which you could be open or not.

00:30:46.940 | Oh, but actually, I had one more thing.

00:30:48.920 | Since you're involved at OpenAI, you can correct me if I'm wrong.

00:30:54.160 | But you have been, you know, saying for, I would say, a couple quarters now that the real opportunity for OpenAI is on the product side versus the model side, which hints at being more of a consumer product than, say, the enterprise APIs business that they've also been in.

00:31:17.980 | If they think that also, and I'm not involved, so this is a conjecture of my part, being more open with your model is a really deft move.

00:31:29.280 | Because it will put more pressure on other players to try and keep up.

00:31:35.260 | And it will allow your model to have more pervasive usage globally.

00:31:40.360 | And you talked about running out of compute.

00:31:43.580 | You know, the minute you put that model out where other people can download it, they're doing that on their servers, not on yours.

00:31:50.500 | And so I just think it's a very clever move for the same reason Google would have supported Kubernetes.

00:31:56.480 | You're kind of wiping out the business opportunity for other models to play on the API side if you make yours open, which helps protect the competitive flank.

00:32:10.340 | And once again, great for consumers.

00:32:13.320 | Yeah, I think it's, I think, you know, it makes sense on a bunch of fronts.

00:32:16.280 | On the first front, I think they want developers to develop on their platform and build applications for, you know, for open AI.

00:32:23.380 | And so this brings them into the ecosystem.

00:32:25.320 | Number two, I think that they fundamentally have a view that they want to build, you know, they want to build products and applications that move humanity forward with AI.

00:32:35.540 | And this is another way to do it at scale.

00:32:38.100 | You know, Sam has said publicly now, I've heard him several times say that he thinks that models are commoditizing.

00:32:45.320 | They're anti into the game.

00:32:46.700 | You know, they will continue to push the frontier.

00:32:49.200 | So he thinks they'll have the best models.

00:32:51.040 | But that general intelligence, you know, that average level intelligence is we already see is going to be widely distributed.

00:32:57.440 | And that the battleground bill is really going to be fought around products and services.

00:33:02.340 | And I wouldn't say that they view themselves as an exclusively a consumer company.

00:33:06.760 | But clearly, Chad GPT is a major thrust, a major focus of the business.

00:33:11.720 | It's the market leading consumer application, probably has 80 to 90 percent market share.

00:33:17.760 | I think network effects are kicking in and other things.

00:33:20.140 | But I also think their enterprise business is, if not the biggest, one of the biggest and also growing at the fastest rate.

00:33:26.900 | Because remember, the consumerization of the enterprise.

00:33:30.760 | One of the fascinating things that's happening in the enterprise is these are all users of Chad GPT.

00:33:35.940 | So when the CEO shows up in the boardroom and somebody says, yeah, we're looking at bringing AI into the company.

00:33:41.600 | And we're looking at, you know, Chad GPT enterprise.

00:33:44.640 | It's an easy yes.

00:33:45.980 | It reminds me of when every CEO said, hey, get me an iPhone, you know, in the enterprise.

00:33:51.380 | And they were on BlackBerry and they all switched to iPhones because they loved using them at home.

00:33:56.100 | I think the natural thing for them to do.

00:33:58.420 | So it's not to say enterprise is going to be a battle.

00:34:01.340 | It's not going to be winner take all.

00:34:03.020 | But I think these guys do have their eyes squarely set on building a big enterprise business.

00:34:07.320 | There's probably two different types of enterprise.

00:34:09.300 | There's probably a product that people buy user licenses for, for doing, you know, white collar research type work.

00:34:19.140 | Where they may, where I think what you said will be very relevant, like having the UI they're used to.

00:34:25.820 | And then there's the separate side, which is models that underlie the type, you know, types of business processes that you're building.

00:34:33.720 | Well, one of the things you pinged me on this week was the investment round around OpenAI.

00:34:39.700 | And, you know, I'm happy to share what I can share.

00:34:42.440 | But, you know.

00:34:43.400 | Tell us what happened.

00:34:44.460 | What was announced?

00:34:45.980 | Yeah, well, I mean, they announced the long rumored investment that was led by SoftBank, which many people described in the headline as a $40 billion investment round.

00:34:57.360 | I think if you read the breakdown of it, it's, you know, it comes in a couple tranches.

00:35:02.580 | The first tranche being closer to $10 billion.

00:35:05.120 | The second tranche being closer to $30 billion.

00:35:08.340 | And it's an extraordinary amount of money.

00:35:11.240 | It's bigger than any, bigger than I think the largest IPO sans maybe one or two that has ever been done.

00:35:17.780 | I've often described these as private IPOs.

00:35:20.740 | Altimeter participated along with several others who were reported.

00:35:24.940 | And the valuation was like $260 pre, which would make it, if all the money were to come in, a $300 post valuation.

00:35:33.120 | And so it certainly got a lot of attention this week.

00:35:36.900 | And, you know, you asked me the question, I think, Bill, just around kind of valuation, right?

00:35:43.700 | How did we think about valuation?

00:35:46.160 | The first thing I would say is market leaders never look cheap.

00:35:49.100 | When I invested in Google in 2005, when I invested in, you know, Meta, when the IPO broke and we looked at those late stage private rounds, I certainly remember the Microsoft round in a Meta at $15 billion that was roundly criticized as being incredibly expensive.

00:36:05.000 | None of these things, you're certainly not going to buy a market leader on the cheap.

00:36:08.820 | But if you really look at this, I think that they've said publicly they expect their revenues this year to be around $13 billion, right?

00:36:17.860 | To do $13 billion in revenue probably means you have to exit the year closer to $15 to $18 billion in run rate revenue.

00:36:25.760 | So as I look at this on a forward run rate for this year, you're paying something like, you know, 20 times revenue for the business.

00:36:33.860 | Now, we also had a couple other announcements, you know, this week.

00:36:38.200 | There's, you know, there's the Anthropic funding round and there's talk that they're doing a billion to $2 billion in revenue, a $60 billion funding round.

00:36:46.820 | So that to me looks like something like 50 times revenue.

00:36:50.580 | So again, you got OpenAI at 20 times, Anthropic at 50 times.

00:36:54.460 | And then we had the merger of X and X.AI, which are rumored to have around, let's call it $3 billion in revenue.

00:37:02.280 | And the combined market cap there is like $125 billion.

00:37:06.000 | So that looks like closer to 80 times revenue.

00:37:09.220 | So the market leader here, which usually trades at a premium, not at a discount, to me, you know, again, we can argue about the sustainability and could somebody disrupt them?

00:37:19.820 | And is 20 times a good valuation in this environment?

00:37:22.560 | And yeah, but aren't they spending a lot of money on compute?

00:37:25.280 | And is it really high value revenue?

00:37:28.120 | But apples for apples relative to their peers, it certainly appears to me like, you know, 20 versus 50 versus 80.

00:37:36.000 | It's hard to say that this would be more expensive on a multiple basis than Anthropic or X.AI.

00:37:42.980 | Yeah.

00:37:43.600 | And I also read that there are still contingencies on whether the full conversion from a nonprofit to a for-profit happens.

00:37:52.760 | So I think there's some stuff.

00:37:54.120 | If that's true, there's still some stuff to play out.

00:37:57.200 | But, you know, one thing I would say when I witnessed this from afar, and once again, you're involved, I'm not.

00:38:03.060 | So correct me if I'm wrong.

00:38:04.380 | But, you know, having lived through the Uber Lyft situation, which oddly had Masa coming to the table also, you know, our world has just evolved into one where so many people believe in power at all.

00:38:19.040 | So many people believe in network effects and that these markets are winner take all that you end up with these massive capital raising rounds.

00:38:28.900 | And, you know, it's not lost on me that both with Stargate and with this one, the headline number is bigger than, you know, the piecemeal when you start to unpack it, which, you know, for better or for worse, from my position, smells of being promotional.

00:38:43.040 | And, you know, and so why would you do that?

00:38:45.960 | Why are you trying to have a bigger headline number?

00:38:48.560 | And that number does get repeated in the press.

00:38:51.640 | So it does work in that way.

00:38:53.640 | And I come back to, you know, I suspect the company's sending a message to Anthropic and anyone else in the game that we're here for the long haul.

00:39:03.560 | And, you know, they probably didn't anticipate all the moves that Elon's made with X and Twitter.

00:39:08.940 | And obviously that is another deep pocketed player.

00:39:13.320 | But, but boy, you know, if you're, if you're on the Anthropic, you know, investment, you know, side, I, you know, I'd be scared.

00:39:22.360 | Like, like, you know, and I've lived through this before.

00:39:25.620 | It is a, it is a, uh, a sport of Kings, if you will.

00:39:29.560 | And then lastly, one thing that naturally falls out of that is unit economics get postponed.

00:39:36.200 | You have to believe in the power law and the network effect.

00:39:38.980 | They've, in addition to that headline number, you know, I think they've said publicly, they expect to lose, you know, five to 7 billion this year.

00:39:47.480 | And, you know, with an employee count of the, I think six to 7,000, that's not going to run you more than a billion or two.

00:39:55.580 | So there's some number between the revenue number you talked about and subtracting 2 billion for expenses.

00:40:02.760 | And the rest is your operating cost of keeping this AI machine going.

00:40:08.900 | And, you know, I ran some loose numbers and I come up around 15 or 20 bucks a year for a non-paid user, you know, just to, to run the servers on their behalf.

00:40:19.840 | And, you know, eventually you might get to advertising.

00:40:22.840 | Eventually you may convert more of them to paid.

00:40:25.360 | These are, these are things we've seen play out over time.

00:40:28.980 | They played out with ad models.

00:40:30.360 | But once again, if you're going to try and lay chase to them, um, you got to be prepared to underwrite that cost yourself.

00:40:36.780 | Right.

00:40:37.740 | You know, listen, I think the analogy is a fair one, Bill.

00:40:41.400 | And obviously Masa was involved in the Uber lift battle.

00:40:44.780 | So it's an easy one, particularly with his involvement here to say, you've referred to it before as weapons of economic destruction, all of this capital.

00:40:53.020 | Um, but I would remind you, there was a moment in time in 2020 where the headlines were that Uber would never be profitable.

00:41:00.320 | It was a failed business model.

00:41:01.780 | It will never make money.

00:41:02.900 | And here's a business that's going to do 6 billion in free cashflow.

00:41:05.980 | No doubt.

00:41:06.520 | Right.

00:41:06.860 | No doubt.

00:41:07.540 | So, and so the winner does take all the winner does take most.

00:41:11.820 | I will tell you as a shareholder, I speak with the leadership of the company all the time about unit economics.

00:41:17.980 | Obviously, if I'm investing in the business, I feel confident in their leadership around unit economics.

00:41:23.640 | I'm, you know, one of the things that I think is really important here is just like what's happening in the business.

00:41:30.420 | Right.

00:41:30.940 | Sam tweeted this week, uh, they added a million chat GPT users in an hour, in an hour that they crossed 20 million subscribers, paying subscribers for chat GPT.

00:41:42.600 | They crossed 500 million weekly average users of chat GPT.

00:41:47.800 | In fact, they were, they're, they're going so gangbusters they're throttling all their demand.

00:41:52.600 | In fact, I don't know if you saw David Sachs's tweet where he said America's leading AI companies are all reporting that demand is off the charts.

00:41:59.340 | So much so that they're being forced to impose rate limits.

00:42:02.000 | And he said, fortunately, we have massive new infrastructure projects coming online, which gets me to the point of why are they raising so much money?

00:42:09.680 | Right.

00:42:10.740 | And you and I are talking about taking the pod down to Abilene, Texas to see Stargate, to Denton, Texas, um, to see, you know, the core weave facility that they're standing up for open AI.

00:42:21.180 | And the fact of the matter is, I think that they need to bring on a massive amount of compute just to support the demand they currently have.

00:42:28.760 | I can tell you when you look at the product pipeline for open AI, right?

00:42:33.280 | Whether it's, uh, you know, there, there are two or three models they already have completed on the shelf.

00:42:38.680 | There's a lot of agent, uh, stuff that they want to do that's on the shelf.

00:42:43.200 | I think there's a lot of stuff they want to do around pricing, but they can't do these things today and open source with their current level of GPU demand.

00:42:52.600 | And, you know, Sam went on online, Seth, anybody has a cluster of a hundred thousand GPUs, you know, send me a DM.

00:42:58.900 | And, you know, you may say it's promotional and hyperbole, but, but the round was already raised.

00:43:03.280 | He's not doing that.

00:43:04.080 | It could be both.

00:43:04.760 | I actually, I, I actually think, uh, you know, in this case, it's true.

00:43:09.620 | I know they were pulling, you know, a lot of things offline just to support the demand.

00:43:14.700 | Now, the irony is where, what was this demand coming from?

00:43:17.580 | Right.

00:43:18.440 | And the demand, you know, because we didn't mention Gemini 2.5 that happened to release, uh, you know, in this last week.

00:43:26.960 | And part of the reason we didn't mention it is because literally on the day that they, that they launch it, OpenAI launches this upgrade to Imogen where people are making all these anime photos of themselves that literally blew up.

00:43:39.880 | Like demand for a billion anime photos a day from the United States all the way to India, and they can't support it.

00:43:46.080 | And some people may say, oh, well, this is an example of how dumb AI is.

00:43:49.280 | People are using it to make anime photos.

00:43:51.440 | But I would point them to Chris Dixon's, you know, blog that he wrote, you know, some time ago where, where he says, listen, the next big thing will first appear as a toy.

00:44:00.580 | Right.

00:44:01.560 | And there are a lot of things that we do for entertainment.

00:44:04.580 | A lot of things that, you know, we know that OpenAI and ChatGPT are being used for a lot of deep research.

00:44:10.840 | Um, but the fact of the matter is at least as to this one, and I'm not going to get into the other, you know, valuations for the other models, but I'd say at least as to this one.

00:44:19.100 | I was an early investor in Google.

00:44:21.120 | I was early investor in Meta.

00:44:22.500 | I saw what those early consumer products look like, what those demand curves look like, what that cohort retention look like.

00:44:30.140 | And I would just say to you that, you know, what I see out of ChatGPT reminds me a lot of kind of those winner-take-all consumer applications.

00:44:40.700 | They're not infallible.

00:44:41.960 | It's not that they can't be assaulted.

00:44:43.680 | You know, Grok has been a great model, you know, launched by Elon, but I think they really do have a big moat.

00:44:50.100 | And I think the network effects are kicking in.

00:44:52.120 | And I think that not only are they order of magnitude bigger, but they're also growing a lot faster.

00:44:57.380 | And so, you know, I think that the consumer business here will ultimately be valuable.

00:45:01.900 | But to your point, the unit economics can be crap along the way, and it's up to the company to launch the things like advertising, pace, you know, different pricing tiers, et cetera, that, you know, bring all of those things.

00:45:14.660 | Let me ask you a question since you went down that avenue.

00:45:20.700 | I think they've announced, is it 20 million paid users and 500 million total users?

00:45:29.040 | So you have a 4% conversion rate.

00:45:31.860 | How do you think about paid versus advertising that conversion rate?

00:45:37.520 | What is, how do you think about the business model with those facts on the table?

00:45:41.420 | Yeah.

00:45:42.200 | I mean, honestly, I think that right now we probably, you know, we're throttling ChatGPT.

00:45:47.220 | So when you bring on more compute, all those numbers would be higher, right?

00:45:51.160 | If you just have more compute.

00:45:52.500 | Secondly, I think most of ChatGPT users are using a model that's like a year old.

00:45:58.560 | Right.

00:45:59.940 | Because we haven't been able to upgrade the, they haven't been able to upgrade the models because I don't think they can support the things they want to do in the upgraded models from a GPU perspective.

00:46:09.260 | So my suspicion is when they're able to do that, that they're going to have a lot more flexibility around things like pricing tiers.

00:46:16.480 | Sam has said he doesn't particularly like advertising, but at some point they will obviously have something that they think is beneficial to consumers that will be around that.

00:46:24.480 | If you look at operator, if I say to operator, book me the Four Seasons Hotel in New York next Tuesday and it does that for me, which I think they're getting a lot closer to you and I have this back and forth on that.

00:46:36.400 | But, you know, let's just stipulate that even you believe at some point they're going to get there.

00:46:41.340 | And if we're driving that kind of value for users, either the user will pay or the end merchant will pay.

00:46:46.980 | I think there are all sorts of business models that will evolve around that.

00:46:50.780 | So my hunch is you're going to see a mixture of advertising.

00:46:53.040 | You're going to see a lot of different pricing tiers.

00:46:55.480 | You're going to see models.

00:46:56.740 | I don't think we're going to have this, you know, this long menu of model choices that forces the consumer to understand the difference between O4 Mini and O3 and O1 and, you know, all these different models.

00:47:08.600 | I think you're just going to have a smart model, ChatGPT5 or ChatGPT6, that may be coming sooner rather than later, that's just going to make those choices for you.

00:47:18.360 | And so I think there are a lot of ways –

00:47:20.420 | By the way, you know, we've talked about this in the past, but I've often felt that one of the reasons that Google is so susceptible to disruption is how they've maximized the revenue per visitor.

00:47:32.320 | And I personally don't think there's any way when that world you're talking about, that agent world evolves, that that partner in a hotel is going to pay a fee anywhere close to the fee that's paid to Google by someone that's marketing a service that I always say using LTV math versus transactional math.

00:47:56.820 | I just don't think there's any way you can get there.

00:47:58.920 | And so that's a huge disruptive advantage for – I think it's a huge disruptive advantage for OpenAI.

00:48:07.040 | Well, let's click on that for a second.

00:48:09.020 | You know, we have our friend Glenn Fogel who runs Booking.com and incredible CEO.

00:48:14.600 | They've built an incredible business.

00:48:16.080 | You know, and they're one of the largest advertisers.

00:48:18.520 | They've historically been one of the largest advertisers on Google.

00:48:21.380 | I think it's been reported that Google generates – it's one of their largest advertising categories in the world is travel, hotels specifically.

00:48:29.140 | Booking is one of their largest global advertisers.

00:48:31.800 | So you sell a $100 hotel room and you take $20, right?

00:48:36.200 | Youbooking.com, let's call it, take $18 to $20.

00:48:39.160 | And then you pay a portion of that to Google, maybe half of it, maybe more than half of it.

00:48:44.020 | Well, actually, to be fair, in many circumstances, you know, they'll be using what I call LTV math and they'll pay –

00:48:52.240 | They'll pay more than 100%.

00:48:53.640 | Oh, they'll pay $50 instead of $20 because – and then they'll say, well, if the customer comes back twice in a year, we get to break even in the first year and we're going to hold them forever.

00:49:04.560 | And this is why that won't work in the agent world, which is no one's going to think that way because if you're a white-label service that's underneath the agent model, the most you can share is 10 of the 20, right?

00:49:18.780 | Or whatever, you know?

00:49:20.560 | It's a –

00:49:21.640 | So there's no doubt there's competition coming to that.

00:49:26.140 | You know, Google's traded down from 200 to 150 and change and, you know, they see that disruption coming their way.

00:49:33.520 | The irony here, right, is we still have, you know, this antitrust investigation with Google.

00:49:39.060 | I always get a, you know, a laugh out of the fact that finally in 2025, the first time we've actually had competition for Google, like very clear that competition is coming to all those categories.

00:49:49.860 | And now we get around to talking about breaking up their search monopoly.

00:49:52.980 | I mean, it's ridiculous.

00:49:54.240 | I don't think that's – I think that's the last of our problems.

00:49:56.460 | But I do think we're going to see business model evolution around these different categories.

00:50:01.180 | So let's just say, Bill, that it settles out at 10 bucks, right?

00:50:04.660 | That the hotels are clearly willing to give 20 bucks.

00:50:07.460 | Let's say it settles out at 10 bucks.

00:50:09.480 | Well, hell, that's all upside for OpenAI.

00:50:11.260 | Right.

00:50:11.700 | But it's replacing 50 for Google.

00:50:13.980 | That's my only point.

00:50:14.900 | Yeah, no, very good point.

00:50:16.780 | By the way, another thing played out in this space, a little out of order of our competition, but there's this acronym.

00:50:23.620 | People are starting to use MCP that is a way for you to represent your service, like if you were booking.com to a model so that it's not simply scraping your website, which is certainly not the ideal way to do this thing.

00:50:39.020 | And that standard, I believe, got started in Anthropic, but OpenAI has agreed to support it.

00:50:45.380 | And so another factor that plays out with whoever's most aggressive with the open standards is they might be able to take a lead in defining these things, which could be advantageous to them.

00:50:58.500 | And I got to tell you, you know, I meant to say this earlier when we were talking about the open source stuff, but I got to believe the anxiety is high at Google.

00:51:08.720 | I got to believe the anxiety is high at Meta.

00:51:11.900 | We've seen some high executive shifts and departures, also at Apple, in terms of who's in charge of these things.

00:51:19.480 | And so I see those moves, and I think that must represent anxiety.

00:51:23.360 | You know, there's a lot at stake.

00:51:26.460 | And so it'll be really interesting to see how open people are, how willing they are to be open with their models, how aggressive they are, because I think it's a really critical window.

00:51:38.700 | I'm talking about like next three to six months could dictate who's standing on top of the hill five, 10 years from now.

00:51:46.700 | Well, the tectonic plates, as I've said, as we've said for now two years, this is the first time they've shifted in this magnitude in 20 years, right?

00:51:56.300 | This is a 20-year event.

00:51:58.660 | For 20 years, the search paradigm ruled everything in consumer internet, and Google stood at the top of that mountain, and it took something, you know, it took an AI-level shift.

00:52:09.800 | It took ChatGPT moment, and them getting to the scale of maybe a billion monthlies and 500 million weeklies to even lead to this conversation.

00:52:19.620 | But things happen very slowly, Bill, as we know, and then they happen very fast, and I think that's your point.

00:52:25.440 | Before we run out of time, we had an IPO, which we haven't had very many of.

00:52:30.740 | Let's talk IPOs, both core weave, but after that, let's just talk about IPOs in general.

00:52:37.020 | Well, it's great.

00:52:39.100 | You know, like, as you know, we're shareholders in core weave.

00:52:44.380 | Since a couple rounds ago, we were one of the largest buyers in the IPO, and we're happy we did.

00:52:50.200 | You know, I have to say, on Friday, I was pretty damn nervous, Bill.

00:52:54.500 | It broke the offering price, went down to about $37 a share.

00:52:58.200 | I think today they had some announcements of this deal with Google where they're going to provide NVIDIA, Grace Blackwells through core weave.

00:53:07.340 | So Google is going to be buying a bunch of NVIDIA chips through core weave.

00:53:10.460 | And one of the big criticisms of this company was they were too dependent upon Microsoft, but now they've diversified.

00:53:16.540 | They have Microsoft, Meta, now Google, OpenAI, NVIDIA, Cohere, Mistral.

00:53:21.740 | And so I think they've really, you know, emerged as the leading AI kind of cloud.

00:53:28.140 | And the stock in the last couple days, despite the fact they took it public last Friday.

00:53:32.900 | I mean, talk about taking a company public into a Category 5 hurricane.

00:53:37.100 | I mean, we had like Liberation Day staring us in the face, and they had to fly into that.

00:53:43.160 | And, you know, as you pointed out, it wasn't the least controversial of IPOs.

00:53:48.040 | But I have to give credit to Mike Intrator and his team.

00:53:51.960 | And listen, I'm here in Silicon Valley.

00:53:53.940 | They started this company five years ago.

00:53:55.860 | It's worth over $30 billion today.

00:53:57.780 | It's played a really important role in standing up OpenAI and a lot of the leading AI labs.

00:54:03.320 | And I just think that's a good thing for all of us.

00:54:05.140 | But it's not, you know, it's also fair to ask the questions that you've asked, you know, around, you know, kind of the durability, if you will, about CoreWeave and the revenue.

00:54:15.200 | Yeah, and look, you're absolutely right.

00:54:18.000 | It's so funny.

00:54:18.680 | We sit around and complain about the IPO market not being open.

00:54:23.480 | And, you know, for the entirety of 2024, the markets were up 30%.

00:54:28.600 | You know, the sunshine was out and no one was going.

00:54:32.060 | So, here, someone finally gets the guts to go.

00:54:35.320 | And the markets, of course, have turned the other way.

00:54:38.400 | And that's why, man, like anyone that tells you when you're ready to go public that you have to wait on the markets to be in a particular place, I would tell them to shut the fuck up and, like, take your company out.

00:54:52.320 | Like, you can't control that thing.

00:54:54.020 | That's an external factor.

00:54:55.120 | And I also believe a lot of people, because of my stance on direct listing, said, was this a good IPO, a bad IPO?

00:55:01.040 | I mean, as you said, there's stuff moving all over the place.

00:55:04.400 | Like, you have a leader in their field, unquestionably, huge revenue growth.

00:55:10.660 | Their business model isn't fully unpacked because the CapEx is invested so far ahead of the product.

00:55:18.560 | So, you can't look at the income statement and say, oh, the right, unit economics are perfect.

00:55:22.680 | And there's questions about what is the appropriate depreciation schedule and all these things.

00:55:29.120 | But I am glad that they got out, and I'm glad that it's done fine.

00:55:33.540 | They've had basically two customer announcements since they went out, which shows how fast this AI world moves.

00:55:39.900 | And I think that's why the stock went from 40 to 36 and way back up above that now.

00:55:44.920 | Let's hope that it brings more IPOs to the table.

00:55:48.540 | Well, this company needed capital because it's a capital-intensive business, and those are the ones that tend to, you know, come to the markets eventually, no matter what.

00:55:58.880 | We have this offsetting reality, you know, that the stripes of the world and Databricks and others are choosing to delay being public and have massive access to private capital.

00:56:10.320 | So, maybe that's a discussion for another day.

00:56:12.520 | But there are a few others, Klarna's in the pipeline.

00:56:15.880 | You know, we've talked about Cerebrus being in the pipeline.

00:56:18.920 | And so, I'm always hopeful.

00:56:21.060 | I'm always kind of just wanting for there to be more companies that are willing to move into the public markets.

00:56:28.280 | But the offset of what we talked about at the beginning is going to be there.

00:56:32.420 | And so, we'll see how those things interact with a choppy market.

00:56:36.700 | Well, you know, somebody else I should mention is Morgan Stanley.

00:56:40.440 | I mean, they took a lot of heat last Friday, you know, on this deal.

00:56:44.120 | Why are you bringing it now?

00:56:45.560 | You know, the stock went below 40.

00:56:47.900 | CNBC was roundly critical of, you know, of the company, you know, and of Morgan Stanley.

00:56:54.740 | And the stock's at 60 bucks or whatever it ended today at.

00:56:57.800 | And so, again, you know, we feel good as shareholders.

00:57:01.720 | I feel good for the people involved in the company.

00:57:04.680 | Obviously, there are still questions, you know, that remain about the business.

00:57:09.120 | You mentioned depreciation.

00:57:10.860 | The one thing I'd tell you about the depreciation argument as it relates to this company is, you know, a lot of people push on, you know, there's a statement by Jensen at GTC that, you know, hoppers may not have value because Grace Blackwell is so much better.

00:57:27.660 | He was a little more aggressive than that.

00:57:29.700 | He called himself the chief revenue destroyer and basically made statements that I think if you interpreted or interpreted accurately would imply that maybe you should have a two-year depreciation, not a six.

00:57:48.320 | Like, he made it sound that way.

00:57:50.360 | It's very abrupt.

00:57:51.100 | And he took a lot of heat.

00:57:52.180 | He probably wishes he hadn't said it, but he said it.

00:57:54.760 | My view on this is like, listen, because we've got to square this circle.

00:57:58.300 | We have, you know, the Saks tweet.

00:58:00.080 | You know, we know the inference demand is off the charts.

00:58:03.240 | Everybody has, you know, is demonstrating their need for more GPUs to run inference.

00:58:08.620 | Everything in the world is becoming, you know, inference.

00:58:11.020 | We've talked about that at length.

00:58:13.280 | And so my view is this, you know, like when you talk about two years for GPUs, they're going to, cutting-edge GPUs are going to be used for cutting-edge training for the, you know, frontier models in that first two-year period.

00:58:24.900 | But all these things are going to continue to get used for inference.

00:58:27.680 | And so the right way to think about CoreWeave, you know, and, you know, I think the consensus margins for this business are like 25% EBIT, you know, over the course of the next couple of years.

00:58:37.280 | How do they get there?

00:58:38.460 | You know, so think about their unit economics bill.

00:58:42.180 | You know, their CapEx, their OPEX.

00:58:45.020 | So they got to get a data center.

00:58:47.740 | They got to pay for all their operating expenses.

00:58:49.460 | Then they got to buy the servers.

00:58:50.960 | So the way this works, I think, is they sell a four-year deal to Microsoft or a four- or five-year deal to OpenAI or a four-year deal to Google or whatever.

00:58:59.340 | They expect to pay back all the CapEx, OPEX, and GPUs in three years.

00:59:04.080 | And so the fourth year, which is a four-year guaranteed contract, the fourth year is your profit margin, right?

00:59:10.660 | And then anything you earn past the four years, that's all gravy on top.

00:59:16.000 | And the consensus earnings are not giving them any credit for anything after the end of those contract periods.

00:59:22.040 | Now, what I'll tell you is, you know, and we've done a lot of research on this, there's still a lot of A100s in use in the world today.

00:59:30.340 | In fact, Jensen has talked at length about that.

00:59:32.460 | That's a 2020 product.

00:59:34.320 | So we're in the fifth year, and A100s are still out there being used by almost everyone that bought A100s.

00:59:40.840 | You know, and then if you look at it, I think Jensen at GTC said last year that OpenAI had just retired the V100s.

00:59:50.820 | That was a 2017 GPU.

00:59:53.600 | So that's like a seven-year lifecycle that they were using those for.

00:59:57.440 | And so I think that we have a lot of comfort that at a minimum, people are going to be using these things for four years, a couple years for training, a couple years for inference.

01:00:07.120 | I've yet to hear of anybody throwing away any GPU because it doesn't have value.

01:00:12.080 | Remember the way CUDA works.

01:00:13.560 | The software that runs these GPUs, it constantly gets upgraded.

01:00:17.760 | It's like my Tesla, right?

01:00:19.380 | I had an old Tesla Model S, like seven years old, but it felt like a new car because my software got updated all the time.

01:00:27.300 | And frankly, it still got me to the places I needed to get to.

01:00:32.320 | It wasn't as good as the new model I bought in December with full FSD and everything else, but it didn't feel like a really old car because the software was constantly updating.

01:00:41.400 | I kind of think of that the same way for these GPUs.

01:00:43.920 | The GPUs are getting better every year, even though the hardware remains the same.

01:00:48.240 | So I'm not nearly as worried about that depreciation schedule.

01:00:51.660 | It seems to be a big hit on the company and lots of people are talking about it, but they're out the door and kudos to them on this big new deal today.

01:01:00.980 | But look, the pushback on that is obviously that it's not a zero or a one.

01:01:06.140 | Like you make it sound like it's binary.

01:01:07.760 | You either throw it away or it's super valuable.

01:01:10.520 | And what inevitably happens is the earning power of that product drops over time.

01:01:16.340 | And so there is, I think, a reasonable question.

01:01:19.580 | You know, should it be more of like an accelerated depreciation schedule?

01:01:23.940 | The idea with depreciation is to kind of, you know, they say the useful life.

01:01:29.240 | So you'd kind of want it to mirror the earnability of the asset over time.

01:01:34.600 | And so six years straight probably isn't the best fit for that, but we'll see.

01:01:39.760 | We'll see what happens over time.

01:01:41.120 | You know, their peers, the incumbents in this world were at four years ago and pushed it to six, which wasn't.

01:01:49.800 | And the answer may lie somewhere in between.

01:01:54.160 | And, you know, like I said, I don't think they need it to be more than four in order to achieve the margins that they have.

01:02:02.500 | But they're also, to your point, it's a highly levered business.

01:02:05.500 | They got to delever the business.

01:02:06.840 | So there's a lot of things in play here with CoreWeave.

01:02:09.580 | That's why, again, if you look at the multiples it's trading at, well, I don't know what they are today.

01:02:15.600 | But the multiples it came public at were not overly taxing from our perspective.

01:02:19.880 | But there's a lot of headwind for all these AI companies.

01:02:22.420 | I mean, you have NVIDIA trading at 19 times fully taxed earnings.

01:02:26.300 | And so, you know, there's a lot of skepticism in the world.

01:02:29.440 | Notwithstanding all the stuff we hear about demand, a lot of skepticism in the world about AI demand.

01:02:34.580 | Riddle for you before we move on.

01:02:36.120 | Yes.

01:02:36.380 | We'll do Salesforce, Netflix, Square, Amazon, Palo Alto Networks, Facebook, Snap, Proofpoint, NetSuite, and CoreWeave have in common.

01:02:45.540 | No idea.

01:02:47.360 | They all broke issue.

01:02:48.900 | Oh, wow.

01:02:50.340 | And so when the talking heads on CNBC and others are critical of a company because they trade below their IPO price, it's just such a wrong way to look at things.

01:03:03.280 | And, you know, I think one of the reasons those high-quality companies get priced to perfection is the founders are stronger-minded and have more leverage and negotiate more on this agreed-upon price, which would also go away with a direct listing.

01:03:20.620 | But, boy, what a silly way to think about quality, whether or not you give away more and pop.

01:03:29.600 | That's what a lot of people think.

01:03:31.280 | You know, I don't like that.

01:03:32.260 | You know, well, I kind of thought that maybe you thought this was a perfect IPO because it ended day one at precisely $40, which was the offering price.

01:03:41.720 | Which was properly engineered.

01:03:42.300 | Let's be realistic.

01:03:43.400 | You also, there's some peculiar terms in this company that you may have played a part in.

01:03:51.060 | The Series C has a put right at like $38.75.

01:03:55.000 | No doubt in my mind, people wanted to make sure it priced above that, which may have played a factor here.

01:04:01.300 | Who knows?

01:04:02.600 | But let's move on.

01:04:03.660 | Let's talk about one more thing before we go.

01:04:05.420 | So much news in one week.

01:04:07.560 | The TikTok thing.

01:04:10.660 | There's new information as we speak.

01:04:13.140 | Tell me what you know.

01:04:14.320 | Well, I mean, listen, there's a lot of rumors swirling, which not surprisingly, this deal is, you know, set to expire and need to be extended by April 5th.

01:04:25.020 | Under the terms of the first congressional extension that was made by Trump.

01:04:31.700 | They've, you know, they've made very clear that, you know, there are a lot of buyers for the TikTok asset bill and that the president has, wants to put together a deal.

01:04:42.040 | And, of course, we have all these tariffs going on on China.

01:04:45.440 | And so I'm sure this will end up as part of a big trade negotiation as it pertains to China.

01:04:50.020 | But as you know, just for everybody, you know, we're shareholders.

01:04:52.820 | I've been a shareholder in this company since 2015, one of the earliest venture capital rounds in ByteDance, the parent company, which, you know, which owns TikTok.

01:05:01.500 | You know, and for the last two years, I've agreed largely with Elon and Sachs and others that we should engage with China.

01:05:07.820 | We shouldn't just shut down TikTok.

01:05:09.660 | We should make TikTok abide by the rules and regulations, right, that we have in this country.

01:05:15.700 | And that's what this whole legislative unwind was about, you know, the forced sale of spin out TikTok U.S.

01:05:23.500 | So here's what I'm hearing.

01:05:24.560 | I'm hearing that there will be a new company stood up, you know, and I'm not privy to any information.

01:05:29.980 | I'm not party to these negotiations, but I'm, you know, let's call it TikTok U.S.

01:05:35.580 | And that TikTok U.S. will be partly owned by ByteDance.

01:05:40.300 | But I think they have to keep that ownership threshold under 20%.

01:05:44.100 | So let's call it 19.5% owned by ByteDance, that it will be owned partially by just the existing shareholders.

01:05:51.140 | Remember, the shareholders in ByteDance, 60% of those are U.S. investors like Altimeter.

01:05:57.080 | Right.

01:05:57.540 | So that we'll get our shares in ByteDance or in TikTok U.S.

01:06:01.440 | And then 50% of it or thereabouts will be new investors.

01:06:05.140 | So think, folks, like some of the rumors I've seen, Amazon, Andreessen, Oracle, you know, et cetera.

01:06:13.680 | And these are investors who are not currently in the cap table of ByteDance.

01:06:19.280 | So Altimeter or CO2, we're currently in the cap table of ByteDance.

01:06:22.860 | Right.

01:06:22.980 | So we're not going to be part of the new investor syndicate, or at least that's my understanding.

01:06:27.640 | So imagine they stand that up.

01:06:31.040 | And then the question-

01:06:31.720 | And where would that money go, Brad?

01:06:34.200 | So the money would go into this NUCO, right?

01:06:37.540 | So the NUCO would be capitalized with this new money.

01:06:40.660 | It would have a new board.

01:06:41.700 | So it would be new, fresh capital for NUCO.

01:06:44.860 | It wouldn't go to ByteDance.

01:06:46.320 | No, that's my understanding, that it would go into NUCO, that NUCO would get a license to the algorithm,

01:06:53.180 | and it would be up to NUCO to audit that, to audit the data.

01:06:57.260 | Because remember, that's the whole point here, Bill.

01:06:59.440 | Like, we want to have some control over, you know, the algorithm and the data.

01:07:04.100 | So it makes sense that Oracle would be involved in that.

01:07:07.040 | Because remember, TikTok runs on the Oracle cloud down in Texas.

01:07:11.100 | I think a logical question is, okay, like, what's the big, what's the so what here?

01:07:15.820 | And I'm hearing that the valuation for TikTok U.S. could be pretty low, which I would expect, right?

01:07:23.440 | Because remember, the, you know, Trump has said, maybe we'll put this in the U.S. Sovereign Wealth Fund.

01:07:29.800 | So he's negotiating the deal.

01:07:31.580 | I expect that he wants to get a pretty damn good deal.

01:07:33.860 | You didn't mention what percentage was for that.

01:07:36.340 | But is that part of the cap table, too?

01:07:38.460 | No, no, no idea.

01:07:40.320 | No idea.

01:07:41.060 | Yeah.

01:07:41.860 | One particular question.

01:07:43.260 | If you go back six months, maybe three or six months, there was a lot of discussion that would suggest that the parent company, ByteDance, had no interest in this deal.

01:07:54.320 | They'd rather shut it down than do this.

01:07:56.700 | Have they changed their mind for some reason?

01:07:59.140 | Is there a new perspective from their side?

01:08:02.520 | Well, I think, remember, if we go back six months, there was a camp that said, shut it down.

01:08:08.920 | And there's a camp saying, or we'll just take it, right?

01:08:13.100 | Right.

01:08:13.620 | And like, I think that the company's perspective, Yaming, the founder of the company, he basically said, there's no way to separate the algorithm between TikTok U.S. and TikTok rest of world.

01:08:24.380 | Because creators in the U.S. create content that go to the rest of the world and vice versa.

01:08:30.000 | And so, like, if you took away all the U.S., it does so much damage, you would be better to shut down TikTok U.S. and just invite the U.S. creators onto the French platform or the United Kingdom version of this or the Australian version of this via a VPN or something.

01:08:48.480 | So, I think the big change here, Bill, is this idea that U.S. TikTok and global TikTok will continue to use the same algorithm.

01:08:57.300 | And it's just a license to the U.S. TikTok would be my guess, was part of that bridge or breakthrough.

01:09:02.840 | I think a key thing here is, like, how does Altimeter or Sequoia or other U.S. investors, remember, 60% of the investors in ByteDance are U.S. investors.

01:09:16.520 | And the investors in places like Altimeter, they're pension funds, they're teachers, they're firefighters.

01:09:21.740 | And if you think about the fair value for ByteDance, I think most people, although it only trades at, let's call it, $300 billion, most people think the fair value of this is closer to a trillion dollars or certainly to $800 billion.

01:09:34.500 | So, if you take 60% of a trillion dollars, that's $600 billion in locked up venture capital value for all of the endowments and pension funds, et cetera, for U.S. investors.

01:09:49.220 | That's more than almost every other unrealized venture gain put together, Bill, right?

01:09:55.500 | And so, if you're able to take this company public, that turns into DPI, like hundreds of billions of dollars of DPI that goes out to the investors in these venture funds.

01:10:07.640 | Is this company being TikTok or ByteDance?

01:10:10.180 | This company being ByteDance.

01:10:12.080 | But we had to get the TikTok deal done as a condition required to get ByteDance public or ByteDance out the door.

01:10:20.920 | And so, remember, ByteDance, about 90% of ByteDance's business is not TikTok U.S.

01:10:28.240 | 90% of the value of the company is things like Doyan, which is the Chinese version of TikTok, and Daobao, which is the Chinese version of ChatGPT, and TikTok around the world.

01:10:39.880 | And so, there's a huge and profitable business inside of China and in the rest of the world.

01:10:44.920 | And we're just debating this piece in the United States.

01:10:47.780 | And so, as a shareholder, I will tell you that whatever the dilution is caused by this, it's nominal relative to the value of the total.

01:10:57.040 | And what I really want to see get done is just certainty, right?

01:11:00.920 | Certainty for the company.

01:11:02.020 | I think it's good for the U.S. that TikTok will remain.

01:11:05.000 | My kids love it.

01:11:05.960 | And I don't, you know, I'm glad we're going to make them abide by the rules of regulation.

01:11:10.880 | I think it's a win, you know, for Team Trump.

01:11:15.000 | I think it's a win for ByteDance.

01:11:16.800 | But remember now, we just hit them today, Bill, with 54% tariffs, okay?

01:11:21.900 | So, there may be a conversation that has to occur before she and Trump.

01:11:26.140 | I thought this deal would get approved by China.

01:11:28.300 | Now, I'm not so sure.

01:11:29.740 | Right.

01:11:30.100 | And the Chinese government could probably block the deal.

01:11:33.300 | Exactly.

01:11:34.220 | So, just because we announce a deal, if we do hear a deal announced over the course of the next few days or over the next week, doesn't mean that it's a done deal.

01:11:42.660 | But I'll leave on an optimistic note.

01:11:44.640 | Okay.

01:11:44.980 | Let's do that.

01:11:45.680 | I think that the president wants to do a deal with Xi.

01:11:49.040 | I don't believe we're going to have 54% tariffs against China.

01:11:54.980 | It's too important to the rest of the world that we can cooperate with China on things like ending the war in the Ukraine, things in the Middle East.

01:12:05.300 | Yes, there is a great competitive struggle between the two countries.

01:12:12.100 | But I think that ultimately, the president will cut a deal.

01:12:16.720 | He said that he likes Xi, invited him to the inauguration.

01:12:19.540 | And as we know, he's a dealmaker.

01:12:21.660 | And now we've got everything from the Panama Canal to negotiate over, to TikTok, to all the other trade deals between the two countries.

01:12:30.560 | So, I suspect that when we get back to what really came out of Liberation Day and what really matters, I think the most important thing that matters is U.S.-China bilateral trade relations.

01:12:42.540 | And I think that's going to really dictate the direction of global growth and the direction of U.S. and China economic growth over the course of the next few years.

01:12:52.780 | Important to watch.

01:12:54.240 | All right, man.

01:12:54.840 | Take care.

01:12:55.580 | Great seeing you.

01:12:56.300 | Have fun at the games.

01:12:57.540 | Take care.

01:12:58.140 | As a reminder to everybody, just our opinions, not investment advice.

Liberation Day, Tariffs, US v China Open Source, OpenAI, $CRWV, TikTok | BG2

Chapters