Latest / KDAV Radio / Better Offline - JensenGate and Anthropic's IPO w/ Ed Elson

Transcript

Better Offline - JensenGate and Anthropic's IPO w/ Ed Elson · KDAV Radio

• 0:02
  Speaker 1: Hello and welcome to Bear Offline, I'm your host Ed Citron.
• 0:18
  Speaker 1: Bear Offline Good Lord, is it a slow news week.
• 0:23
  Speaker 1: And so to bring in someone just for a little
• 0:26
  Speaker 1: bit of entertainment, I have the mighty Ed Elson from
• 0:29
  Speaker 1: the Prof G Markets podcast. Ed, how are you doing?
• 0:32
  Speaker 2: I'm doing well. Here to bring the entertainment. How are
• 0:34
  Speaker 2: you doing?
• 0:35
  Speaker 1: Well, thankfully, 300 stories actually did run. So we're going
• 0:39
  Speaker 1: to start with Anthropix IPO. It seems like they're pricing
• 0:43
  Speaker 1: it somewhere between a trillion and 1.4 trillion. But it
• 0:46
  Speaker 1: wasn't even that simple. They filed something, they retracted something.
• 0:50
  Speaker 1: It's been kind of a mess.
• 0:52
  Speaker 2: Yeah, it's been one of the ugliest rollouts of an
• 0:56
  Speaker 2: IPO that we've seen, but not something that you and
• 0:58
  Speaker 2: I didn't expect. I mean, this is a company that
• 1:02
  Speaker 2: wanted to go out at $ 2 trillion, despite the fact
• 1:05
  Speaker 2: that as we learned, it was going to spend about
• 1:08
  Speaker 2: $ 800 billion over the next several years, almost $ 500 billion
• 1:12
  Speaker 2: over the next three years. And this is a company
• 1:14
  Speaker 2: that investors were saying, oh, it's about to be profitable,
• 1:18
  Speaker 2: which I understand based on the income statement, which told
• 1:21
  Speaker 2: a very different story of operating losses coming down 70%
• 1:25
  Speaker 2: to 75% over the past year, which I also was
• 1:28
  Speaker 2: initially optimistic about. To me, that was like an answer
• 1:31
  Speaker 2: to the question of, Does the LLM business actually work?
• 1:35
  Speaker 2: And my response was, oh, maybe it will. Maybe it
• 1:38
  Speaker 2: will imminently or next year. But then, of course, we
• 1:41
  Speaker 2: get this giant caveat, which is like actually what we
• 1:44
  Speaker 2: learned from 2026 in terms of the fundamentals of this
• 1:47
  Speaker 2: business tells us basically zero about what the business is
• 1:51
  Speaker 2: actually going to look like next year. And so that
• 1:54
  Speaker 2: was when I was like, okay, this $ 2 trillion number
• 1:57
  Speaker 2: is seems too high. My prediction was that Anthropic was
• 2:01
  Speaker 2: either going to delay the IPO, or it was going
• 2:04
  Speaker 2: to reduce the price. It seems like the latter has occurred.
• 2:08
  Speaker 1: Yeah, and it was interesting. The boosters really got hot
• 2:12
  Speaker 1: and heavy over the Q2 numbers, because without the Q2
• 2:15
  Speaker 1: numbers first, everyone was like, where's Ed? He had a
• 2:18
  Speaker 1: family member in the hospital, so that's something more important. Yeah,
• 2:21
  Speaker 1: they were real.
• 2:23
  Speaker 2: I remember just, you were not online and everyone was
• 2:27
  Speaker 2: freaking out. They thought it was this giant admission of defeat. Yeah,
• 2:31
  Speaker 2: they did not have all of the facts in front
• 2:32
  Speaker 2: of them.
• 2:32
  Speaker 1: I saw all the dickheads deleting the posts afterwards, which
• 2:35
  Speaker 1: I really enjoyed. But what was interesting was they blanked
• 2:38
  Speaker 1: out the Q3 numbers. Because when the Q3 numbers came out,
• 2:42
  Speaker 1: revenue growth had slowed, costs had exploded, and they'd added
• 2:46
  Speaker 1: that they had $ 895 billion worth of commitments in an
• 2:52
  Speaker 1: indeterminate period. It's just a really, it's just a, I
• 2:56
  Speaker 1: don't know if you agree on this one, but it
• 2:58
  Speaker 1: just feels like kind of a dodgy company. They obfuscate details.
• 3:02
  Speaker 1: They are very vague about the actual timelines, because even
• 3:06
  Speaker 1: in the updated S1, the IPO doc, the filing, Even
• 3:10
  Speaker 1: when they updated it, they said, oh yeah, the minimum
• 3:14
  Speaker 1: amount we'll have to spend in 2027 is $ 30 billion.
• 3:17
  Speaker 1: And it's like, we know you're not spending that little.
• 3:20
  Speaker 1: And then they're like, actually, we expect to spend $ 70
• 3:23
  Speaker 1: billion in the next 12 months. But it's unclear if
• 3:26
  Speaker 1: that includes Elon Musk, who's charging them $ 3. 1 billion a month.
• 3:30
  Speaker 1: It's just, I realize that they're allowed to put whatever
• 3:33
  Speaker 1: they want in these things, as long as they follow
• 3:36
  Speaker 1: GAAP or what have you, or say when it isn't GAAP.
• 3:38
  Speaker 1: But this just feels... so wriggly to me.
• 3:45
  Speaker 2: Yeah, I actually, on the spectrum of dodginess in AI world,
• 3:50
  Speaker 2: I actually put Anthropic on the side of less dodgy
• 3:53
  Speaker 2: because something I do, I mean, that's not giving it,
• 3:56
  Speaker 2: I mean, that's not that much praise.
• 3:59
  Speaker 1: I mean, we're dealing with companies.
• 4:00
  Speaker 2: That, as we just learned, are literally smuggling chips into
• 4:03
  Speaker 2: China and then having affairs with Chinese spies. So, you know,
• 4:06
  Speaker 2: dodginess is rife throughout the industry. So I'm talking about
• 4:09
  Speaker 2: as far as AI goes. And the reason I think
• 4:11
  Speaker 2: that is because You know, one of the things that
• 4:14
  Speaker 2: I expected and that you might have expected is that
• 4:17
  Speaker 2: they were going to do some ridiculous accounting gymnastics on
• 4:21
  Speaker 2: their non-gap adjustments. I was expecting that they were going
• 4:24
  Speaker 2: to say, we are profitable when you remove our training costs.
• 4:30
  Speaker 2: And I think that's something that a lot of people expected.
• 4:32
  Speaker 2: And I was waiting for that to happen. Yeah. With
• 4:36
  Speaker 2: the gross margins in some of that original reporting, but
• 4:39
  Speaker 2: not in the actual income statement itself. Not when we
• 4:41
  Speaker 2: look at the operating losses, which did account for the
• 4:44
  Speaker 2: training costs. So on that front, I was like, okay...
• 4:48
  Speaker 2: you know, credit to you for not bullshitting us too extravagantly.
• 4:54
  Speaker 2: But the trouble is that they did not bring enough
• 4:56
  Speaker 2: attention to that $ 895 billion. But again, maybe you'd blame
• 5:01
  Speaker 2: that on investors for just not doing their homework and
• 5:04
  Speaker 2: actually reading the document. The other thing that I do
• 5:07
  Speaker 2: give Anthropic credit for is, you know, when you read
• 5:10
  Speaker 2: the full thing and you read what they're explaining about
• 5:14
  Speaker 2: what is about to happen to the company, I just
• 5:16
  Speaker 2: want to read you this one quote in the filing
• 5:18
  Speaker 2: where they said, quote, our increasing efforts to build out
• 5:21
  Speaker 2: our compute and infrastructure require substantial capital expenditures that are
• 5:27
  Speaker 2: changing the cashflow of our business. That is a businessy
• 5:31
  Speaker 2: way to say the most important thing in the entire document,
• 5:34
  Speaker 2: which is that massive losses are about to come. So
• 5:38
  Speaker 2: they're saying our cashflow, the cashflow profile of our business,
• 5:42
  Speaker 2: which might look okay-ish or that it's about to be
• 5:46
  Speaker 2: pretty good, that's going to change substantially because we have
• 5:50
  Speaker 2: all of these issues where we're basically reliant on our
• 5:53
  Speaker 2: competitors compute, which means that we're probably going to be
• 5:57
  Speaker 2: locked into some extremely onerous agreements, which of course we're
• 6:01
  Speaker 2: beginning to see show up in the business profile of
• 6:05
  Speaker 2: this company. But they're saying we're going to have to
• 6:07
  Speaker 2: do something about that. And it probably means ramping up
• 6:10
  Speaker 2: capital expenditures, which is Latin for probably building our own
• 6:13
  Speaker 2: data centers, which is going to completely transform the way
• 6:17
  Speaker 2: this business actually works. So I actually credit them for
• 6:20
  Speaker 2: at least saying that, though perhaps they could have been
• 6:23
  Speaker 2: a bit more forthcoming about the significance of that element.
• 6:28
  Speaker 1: Well, I mean, that's the thing. The way I describe
• 6:32
  Speaker 1: Q2 is that is a company that no longer exists.
• 6:36
  Speaker 1: Just the Q2 one was discounted SpaceX compute. It was,
• 6:40
  Speaker 1: I reckon, I've heard this from a few people, but
• 6:44
  Speaker 1: you can't really report on what your mate told you.
• 6:46
  Speaker 1: But I've heard people tell me that they ramped down
• 6:50
  Speaker 1: training costs in Q2. But the problem is Q3 was
• 6:53
  Speaker 1: such an aggressive ramp up of costs. And it's so
• 6:57
  Speaker 1: obvious that they're cramming this thing out the door at
• 6:59
  Speaker 1: the worst time. It's so obvious that Q4 is getting
• 7:03
  Speaker 1: worse because Ramp is saying, and I had a word
• 7:05
  Speaker 1: with Ara about this, Ara Karazian, he's probably going to
• 7:08
  Speaker 1: kill me for I just said his second name, from Ramp.
• 7:11
  Speaker 1: And he was saying that right now, Q4 is trending
• 7:15
  Speaker 1: towards 15% to 25% quarter-over-quarter growth, down from 47%, I think.
• 7:21
  Speaker 1: And that may seem like a lot, and indeed it
• 7:23
  Speaker 1: is a lot in a vacuum. But this is a
• 7:26
  Speaker 1: company whose expenses are going to ramp up. Well, they
• 7:29
  Speaker 1: ramped up 77% compute-wise between Q2 and Q3. That mathematics
• 7:35
  Speaker 1: stops working very quickly. And it's funny, you've got Goldman
• 7:38
  Speaker 1: Sachs saying, oh, they'll have $ 600 billion of revenue. in
• 7:42
  Speaker 1: uh in 2030 it just feels i i actually fundamentally
• 7:46
  Speaker 1: disagree with you about their transparency this is a 300
• 7:50
  Speaker 1: page document that sells a company based on terms that
• 7:55
  Speaker 1: really don't apply to the new company like in a
• 7:59
  Speaker 1: year's time anthropic is not going to be profitable or
• 8:03
  Speaker 1: indeed dealing with what $ 7 billion worth of compute costs,
• 8:07
  Speaker 1: 6 billion. In fact, I find the compute costs extremely confusing.
• 8:11
  Speaker 1: Something does not add up there.
• 8:13
  Speaker 2: Well, I think this is a problem with S1s is
• 8:16
  Speaker 2: that it's supposed to be like, it's almost presented as
• 8:19
  Speaker 2: if it's this financial audit of a company that's showing you,
• 8:22
  Speaker 2: here are all the things that you need to know.
• 8:23
  Speaker 2: Here are the risk disclosures. In reality, it is literally
• 8:26
  Speaker 2: a sales pitch. It's like, here's why you should buy
• 8:29
  Speaker 2: this thing. And that's why so many companies end up
• 8:32
  Speaker 2: doing some ridiculous things in there. I mean, the worst
• 8:35
  Speaker 2: perpetrator would be WeWork with the community-adjusted EBITDA, which is
• 8:38
  Speaker 2: like basically fraud. But I mean, with Anthropic, I mean,
• 8:46
  Speaker 2: I just want to give them at least the courtesy
• 8:49
  Speaker 2: of the credit that they didn't do that and that
• 8:51
  Speaker 2: they said what is going to happen. But the thing
• 8:55
  Speaker 2: that is a little bit, you know, to not give
• 8:58
  Speaker 2: them too much credit it was buried like 160 pages
• 9:03
  Speaker 2: into the document, the most crucial details of this thing.
• 9:06
  Speaker 2: And then as you point out, I mean, they had
• 9:09
  Speaker 2: to then go back and give the updated numbers on
• 9:12
  Speaker 2: Q3 and also give the updated semblance of a timeline
• 9:15
  Speaker 2: on how those $ 895 billion in costs are going to
• 9:19
  Speaker 2: be spread out over the next several years. So just
• 9:22
  Speaker 2: to your point about the 25% revenue growth, you know,
• 9:27
  Speaker 2: you've got a company here that spent around, $ 10 billion
• 9:31
  Speaker 2: on compute in 2026. And so let's assume that it's
• 9:38
  Speaker 2: going to be $ 20 billion. They're going to spend $ 895
• 9:44
  Speaker 2: billion in the near future, and almost $ 500 billion of
• 9:47
  Speaker 2: that is going to be in the next three years.
• 9:51
  Speaker 2: So you're looking at compute costs that are going to
• 9:53
  Speaker 2: rise next year to at least $ 100 to $ 150 billion.
• 9:57
  Speaker 2: So that's when we're living in this world of, oh, 25%
• 10:01
  Speaker 2: quarter over quarter revenue growth, that might sound okay. but
• 10:14
  Speaker 2: actually it just doesn't cut it anymore. That's significantly lower
• 10:19
  Speaker 2: than the amount that the spending is going to increase.
• 10:22
  Speaker 2: This is the problem. We can talk about the income statement,
• 10:25
  Speaker 2: we can talk about the numbers in Q2, we can
• 10:27
  Speaker 2: even talk about Q3, but not even that tells us
• 10:30
  Speaker 2: The full picture. The full picture is happening next year
• 10:34
  Speaker 2: and the year after that, where they're going to have
• 10:36
  Speaker 2: to raise hundreds of billions of dollars. And they were
• 10:39
  Speaker 2: able to do it pretty easily over the last couple
• 10:42
  Speaker 2: of years because it's been the era of AI euphoria
• 10:45
  Speaker 2: and exuberance. Capital was cheap. Capital was readily available. You
• 10:48
  Speaker 2: could go out to Silicon Valley. You could go out
• 10:50
  Speaker 2: to the Gulf states. Everyone's willing to just hand you
• 10:53
  Speaker 2: money to go build your AI. But now we're reaching
• 10:56
  Speaker 2: a different moment. where suddenly investors are actually getting a
• 10:59
  Speaker 2: little bit anxious. And we're seeing this reflected in the
• 11:02
  Speaker 2: markets today. We're seeing it reflected in all of the
• 11:05
  Speaker 2: AI names. We saw this happen after the midterms. We're
• 11:08
  Speaker 2: seeing what's happened to Nvidia, which by the way, has
• 11:11
  Speaker 2: lost one and a half trillion dollars in market value
• 11:14
  Speaker 2: in about five weeks. It has been an absolutely stunning
• 11:17
  Speaker 2: fall for Nvidia. And so the question is, do you think...
• 11:21
  Speaker 2: that Anthropic can continue to raise billions, hundreds of billions
• 11:25
  Speaker 2: of dollars ad infinitum? Or is there a point at
• 11:28
  Speaker 2: which investors say, actually, we're going to be a little
• 11:31
  Speaker 2: bit more discerning right now. We're a little less excited
• 11:34
  Speaker 2: about your AI future. We're a little bit more concerned.
• 11:37
  Speaker 2: about how much you're going to spend. I think that
• 11:39
  Speaker 2: they are now running up against that obstacle, which is
• 11:42
  Speaker 2: why they've lowered the amount that they're going to raise.
• 11:45
  Speaker 2: They wanted to do $ 100 billion. Now we're down to
• 11:47
  Speaker 2: $ 65 billion. And in addition, they have reduced the price
• 11:52
  Speaker 2: by nearly 50%. Yeah.
• 11:53
  Speaker 1: It's like 1 trillion to 1.4 trillion, depending on what
• 11:58
  Speaker 1: article you're reading. Yes. Which is very frustrating. Again, very frustrating.
• 12:03
  Speaker 1: The reporting on this has been fucking strange to me
• 12:05
  Speaker 1: because Anthropic put out a document on Monday night saying, oh,
• 12:11
  Speaker 1: this many shares. And then they retracted it. Everyone calculated
• 12:15
  Speaker 1: 1 to 1.25. Now it's 1 trillion to 1.4 trillion.
• 12:21
  Speaker 1: It isn't really clear. And Anthropic is 26 billion dollars
• 12:27
  Speaker 1: of those shares are being sold to Google, Nvidia, AMD, Broadcom, Micron, Microsoft,
• 12:32
  Speaker 1: and TSMC.
• 12:33
  Speaker 2: That's the other thing. They needed to bring in some
• 12:36
  Speaker 2: help to help support the price.
• 12:37
  Speaker 1: But that immediately makes me think something, which is, there's
• 12:40
  Speaker 1: not going to be any money left if OpenAI even
• 12:42
  Speaker 1: tries to go public. Because OpenAI tried to raise... This
• 12:45
  Speaker 1: is also somewhat of a diversion from the Anthropic one,
• 12:48
  Speaker 1: but OpenAI last week raised $ 11 billion at a $ 1. 25
• 12:52
  Speaker 1: trillion private valuation. And that was down from raising $ 30
• 12:56
  Speaker 1: billion at $ 1. 4. I mean, it kind of feels like
• 13:02
  Speaker 1: the money has run out, or is running out, or
• 13:05
  Speaker 1: will run out very soon. Because... And also just OpenAI
• 13:10
  Speaker 1: can't go public now. How? How?
• 13:16
  Speaker 2: Well, I think this is the ultimate question is, is
• 13:19
  Speaker 2: the money beginning to run out? And by the way,
• 13:21
  Speaker 2: it's not just OpenAI that's going to put pressure on
• 13:24
  Speaker 2: the capital markets. It's all of the big tech companies,
• 13:27
  Speaker 2: which are also going out and raising tens of billions
• 13:29
  Speaker 2: of dollars in equity raises and also debt issuance. I mean,
• 13:33
  Speaker 2: Meta's got a $ 60 billion equity raise that it needs
• 13:37
  Speaker 2: to carry out. next year. They've already planned it. And
• 13:41
  Speaker 2: this is in addition to the tens of billions of
• 13:43
  Speaker 2: dollars that have been raised by Microsoft and Amazon and Oracle.
• 13:48
  Speaker 2: I mean, you look at the debt issuance, we're looking
• 13:51
  Speaker 2: at nearly a trillion dollars in debt issuance by 2028,
• 13:55
  Speaker 2: purely among the hyperscalers. So the question you have to
• 13:58
  Speaker 2: ask yourself is, is the money there? Do investors actually
• 14:03
  Speaker 2: have enough money to pay that amount to build this
• 14:06
  Speaker 2: thing that increasingly they're becoming very skeptical of? So for me,
• 14:11
  Speaker 2: I mean, the way that I like to think about
• 14:15
  Speaker 2: it and the way that it's often explained on Wall
• 14:19
  Speaker 2: Street is it's called the music.
• 14:23
  Speaker 1: Is the music playing or is it stopping?
• 14:25
  Speaker 2: Margin call. Where Jeremy Irons says, my only job is
• 14:28
  Speaker 2: to figure out what's happening to the music, what's going
• 14:31
  Speaker 2: to happen in the music tomorrow and the day after.
• 14:34
  Speaker 2: So this is the most important thing for the AI
• 14:36
  Speaker 2: trade is how does the music actually sound? I think
• 14:39
  Speaker 2: that if there's one thing that I can say with
• 14:41
  Speaker 2: certainty is that music is still playing.
• 14:44
  Speaker 1: Yes.
• 14:44
  Speaker 2: Because you had Meta, which had this $ 25 billion stock sale,
• 14:49
  Speaker 2: which was five times oversubscribed. You had Amazon with a
• 14:52
  Speaker 2: $ 25 billion bond deal. It drew more than $ 60 billion
• 14:55
  Speaker 2: in orders. You had Microsoft pull off one of the
• 14:58
  Speaker 2: most sought-after debt offerings on record. But that's because Microsoft's
• 15:01
  Speaker 2: high quality. Exactly. So that's the other thing, is that
• 15:06
  Speaker 2: investors feel a lot more safe about Microsoft because they
• 15:08
  Speaker 2: haven't obliterated their free cash flow. But then you have
• 15:11
  Speaker 2: all of these other signals, like Corweave's bond yields exploding
• 15:17
  Speaker 2: into space, and you have OpenAI reducing its funding target
• 15:22
  Speaker 2: by $ 10 billion. And now this, to me, this anthropic
• 15:25
  Speaker 2: IPO is the clearest signal of what's happening to the music.
• 15:29
  Speaker 2: And the fact that they've reduced the price as much
• 15:30
  Speaker 2: as they have, they've reduced the amount that they want
• 15:32
  Speaker 2: to raise, to me, that is the clearest signal, okay,
• 15:35
  Speaker 2: the music's beginning to get a little bit fainter. It's
• 15:37
  Speaker 2: beginning to get a little bit quiet.
• 15:38
  Speaker 1: It's slowing down.
• 15:39
  Speaker 2: It's slowing down. And so if you're expecting that investors
• 15:44
  Speaker 2: are going to save you. I mean, a lot of
• 15:47
  Speaker 2: these companies are dangerously dependent on the capital markets at
• 15:51
  Speaker 2: this point to fund their ambitions. And that's a new
• 15:55
  Speaker 2: dynamic because for years, big tech was funding their ambitions
• 15:58
  Speaker 2: on their own. Yeah, they were spending $ 80 billion on
• 16:01
  Speaker 2: the metaverse, but they were at least using their own
• 16:03
  Speaker 2: profits to fund that operation. We're no longer in that
• 16:06
  Speaker 2: territory now. The more that companies say, okay, we need
• 16:10
  Speaker 2: your money to fund what we want to do, well,
• 16:13
  Speaker 2: that puts yourself in a very vulnerable position because that
• 16:15
  Speaker 2: means that the people have to be willing to give
• 16:17
  Speaker 2: you the money. Increasingly, investors seem pretty unwilling.
• 16:21
  Speaker 1: Well, also, the other thing is, just back to open
• 16:25
  Speaker 1: AI for a second, let's think practically here. Anthropic goes out,
• 16:29
  Speaker 1: best case scenario, let's say 1.4 trillion, maybe pops to 1.5, 1.6.
• 16:33
  Speaker 1: How in blue blazes does OpenAI actually navigate that? At
• 16:41
  Speaker 1: a 1.25, also everyone who invested at 1.25 in private
• 16:44
  Speaker 1: is a moron, but putting all that aside, what the
• 16:47
  Speaker 1: fuck does an IPO even look like for that company?
• 16:50
  Speaker 1: They've got lower revenues, higher costs, more of a drag
• 16:53
  Speaker 1: on their revenues. I mean, on their costs, like they
• 16:56
  Speaker 1: have $ 750 billion worth of commitments through 2030. I mean,
• 17:01
  Speaker 1: I don't, like, I think it's worth discussing what OpenAI
• 17:05
  Speaker 1: does full stop. How the hell does it navigate out
• 17:08
  Speaker 1: of this? Because put aside whether the money is available
• 17:11
  Speaker 1: or not, Anthropic's IPO better go absolutely perfectly, because that's
• 17:15
  Speaker 1: the only condition under which OpenAI could go public. Yeah.
• 17:19
  Speaker 2: I mean, it seems that as you bring this up,
• 17:22
  Speaker 2: Anthropic has been getting all of the attention because it's
• 17:26
  Speaker 2: the company that's going public. And so we've been looking
• 17:28
  Speaker 2: at their financials, we've been scrutinizing them, we've been assessing
• 17:30
  Speaker 2: the risk. But what we know about OpenAI is... Whatever's
• 17:36
  Speaker 2: gone wrong at Anthropic, it's 10 times worse at OpenAI,
• 17:41
  Speaker 2: where they lost... I mean, we're talking about the operating
• 17:45
  Speaker 2: losses at Anthropic of more than a billion dollars through
• 17:49
  Speaker 2: the first three quarters of this year, which is like, okay,
• 17:52
  Speaker 2: that's not great, but maybe they'll figure it out. I mean,
• 17:56
  Speaker 2: we can have that debate. OpenAI lost more than $ 20
• 18:00
  Speaker 2: billion on an operating basis last year. So... What would
• 18:05
  Speaker 2: happen if OpenAI were trying to go public right now
• 18:09
  Speaker 2: and all of Wall Street in the world was scrutinizing
• 18:13
  Speaker 2: their financial statements? What would happen then? I mean, I
• 18:17
  Speaker 2: think it would be probably a crisis. And so my
• 18:21
  Speaker 2: expectation is that there is a little bit of crisis
• 18:24
  Speaker 2: mode going on within open AI because they're realizing, oh no,
• 18:28
  Speaker 2: we are actually dependent on these people who are currently
• 18:33
  Speaker 2: selling like crazy on the AI trade. And so we
• 18:36
  Speaker 2: need to do whatever we can to figure out how
• 18:38
  Speaker 2: to become independent. We need to get the money from somewhere.
• 18:41
  Speaker 2: Oh no, we can't get the money from our own
• 18:44
  Speaker 2: business because our own business is losing tens of billions
• 18:47
  Speaker 2: of dollars a year. So I would bet that, you know,
• 18:50
  Speaker 2: I'm sure people at Anthropic are quite nervous. I would
• 18:53
  Speaker 2: bet that people at OpenAI are a lot more nervous
• 18:56
  Speaker 2: right now because from my understanding, their spending has been
• 19:00
  Speaker 2: way more irrational than Anthropic's. And if there is a
• 19:06
  Speaker 2: company that's going to get burned, in my view, it's
• 19:08
  Speaker 2: probably going to be OpenAI.
• 19:20
  Speaker 1: Well, also, they have the burden of these hundreds of
• 19:22
  Speaker 1: millions of free subscribers because they're making what? They were like, oh, yeah,
• 19:29
  Speaker 1: we've got a billion dollars of annualized run rate on ads.
• 19:33
  Speaker 1: Who cares? I'm sorry. Be an adult. Stop using run rates.
• 19:37
  Speaker 1: But also, that's not enough. The inference costs those free
• 19:43
  Speaker 1: bastards are really dragging them down. But then there's just
• 19:46
  Speaker 1: the value proposition of what is it an investor would
• 19:49
  Speaker 1: be investing in, in OpenAI versus Anthropic? It's a very
• 19:52
  Speaker 1: similar company with worse economics, with a more- Exactly. I
• 19:55
  Speaker 1: was about to say more annoying CEO, but I think
• 19:58
  Speaker 1: that Dario and Sam, that's a competitive field.
• 20:02
  Speaker 2: But on top of all- The only thing, just to
• 20:05
  Speaker 2: build on that, the only thing- that would be compelling
• 20:08
  Speaker 2: for an investor to invest in OpenAI versus Anthropic is
• 20:12
  Speaker 2: a dramatically lower price.
• 20:14
  Speaker 1: Yes.
• 20:15
  Speaker 2: That's basically it. Like, yeah, like a 500 billion. You're
• 20:18
  Speaker 2: the worst version. I mean, you're the company that is
• 20:22
  Speaker 2: at significant risk compared to Anthropic. Okay, I want to
• 20:27
  Speaker 2: see that reflected in the price. I want to see
• 20:28
  Speaker 2: a dramatic discount, which to me likely means that we're
• 20:32
  Speaker 2: probably going to see the era of down rounds. We'll
• 20:35
  Speaker 2: look at these unbelievable valuations that these companies commanded in
• 20:39
  Speaker 2: the private markets in Silicon Valley, where investors don't really
• 20:42
  Speaker 2: care about profits. They don't really care about margins. They
• 20:44
  Speaker 2: talk They care about technology and founder vision and how
• 20:49
  Speaker 2: cracked the engineering team is. They don't care about the
• 20:52
  Speaker 2: numbers as much. And then suddenly they're going to enter
• 20:55
  Speaker 2: the coliseum of price discovery, which is the public markets.
• 20:58
  Speaker 2: And investors are going to say, yeah, this doesn't work
• 21:01
  Speaker 2: for us. And it'll be a rude awakening for OpenAI,
• 21:04
  Speaker 2: for sure.
• 21:04
  Speaker 1: And let's be clear that OpenAI raised that 1.25 valuation,
• 21:09
  Speaker 1: 11 billion, from the Sauds. They went to the Middle
• 21:14
  Speaker 1: East and they said, we want $ 30 billion, $ 1. 4 trillion valuation.
• 21:17
  Speaker 1: And they said, hold your horses, that's too much. The
• 21:18
  Speaker 1: Middle East. The middle, the literal, like the place you
• 21:31
  Speaker 1: go when you have your dumbest ideas and you want
• 21:33
  Speaker 1: the most money. Because they already shook down poor Masayoshi's
• 21:38
  Speaker 1: son and his geese. And they had to go and
• 21:40
  Speaker 1: bother the... And now we're in a situation where they
• 21:44
  Speaker 1: weren't able... And now they're saying they're going to go
• 21:45
  Speaker 1: around and they're going to raise from family offices. Yes.
• 21:50
  Speaker 1: But they very specifically didn't say venture capital. So I
• 21:53
  Speaker 1: think venture capital is out as well.
• 21:57
  Speaker 2: I mean, this is what we mean by the music.
• 21:59
  Speaker 2: I mean, going, it's a great point. You go to
• 22:01
  Speaker 2: the middle East, the people who will pay any amount
• 22:04
  Speaker 2: of money for literally anything and have done for, for
• 22:08
  Speaker 2: years at this point. And they say, oh, this is
• 22:10
  Speaker 2: a little bit too much. We're going to downsize our
• 22:13
  Speaker 2: quantum here. That to me is like a giant note
• 22:16
  Speaker 2: out of tune. Uh, that is, that's a real problem.
• 22:20
  Speaker 2: It's a real signal. of investor sentiment. And by the way,
• 22:24
  Speaker 2: usually I don't really give a crap about investor sentiment
• 22:28
  Speaker 2: because I'm someone who cares more about the fundamentals of
• 22:31
  Speaker 2: the business. Is the business working? What does the business
• 22:34
  Speaker 2: look like? Who cares about what the noise is? But
• 22:36
  Speaker 2: the trouble when it comes to AI and specifically these
• 22:38
  Speaker 2: companies is that they are completely levered to investor sentiment.
• 22:43
  Speaker 2: I mean, investor sentiment will dictate whether they live or
• 22:47
  Speaker 2: if they die. And so you actually have to keep
• 22:50
  Speaker 2: an eye on what's happening in best of sentiment. You
• 22:53
  Speaker 2: have to understand what the music is, what the music
• 22:55
  Speaker 2: is sounding like, because if it starts to go out
• 22:57
  Speaker 2: of tune, then yeah, these companies genuinely go out of
• 23:01
  Speaker 2: business that, or they have to massively restructure their business models.
• 23:05
  Speaker 2: And that will be very, very ugly.
• 23:07
  Speaker 1: And I mean, there was that story that came out
• 23:09
  Speaker 1: and yeah, just a tip of the hat here to myself,
• 23:12
  Speaker 1: because I called this. Apparently, Blackstone was telling investors, by
• 23:16
  Speaker 1: which I mean the marks in the whole Broadcom deal,
• 23:20
  Speaker 1: that Anthropic's expected to get a below investment grade rating.
• 23:25
  Speaker 1: And this is... Remember, this company's now raising, what, $ 65,
• 23:28
  Speaker 1: $ 69 billion at IPO. They had a big cash position,
• 23:33
  Speaker 1: like $ 100 and something billion.
• 23:35
  Speaker 2: Which they're going to spend in the next year.
• 23:36
  Speaker 1: Which they're going to send a chunk of that out
• 23:38
  Speaker 1: the door. But here's the thing. I... Junk rating will
• 23:42
  Speaker 1: be very difficult to navigate for this company because junk
• 23:48
  Speaker 1: rating means most investment funds, pension funds, and so on
• 23:51
  Speaker 1: and so forth can't invest in your debt. But on
• 23:53
  Speaker 1: top of that, the Fed's talking about multiple rate hikes
• 23:57
  Speaker 1: in the next few months. Yes. What the fuck are they?
• 24:01
  Speaker 1: They're going to be borrowing at 10%. It may as
• 24:03
  Speaker 1: well go on fucking SoFi.com, not a sponsor.
• 24:05
  Speaker 2: They'll be borrowing at 15%. That's what Corweave is borrowing at, 15%.
• 24:09
  Speaker 1: Well, I mean, Corweave's just not borrowing at the moment. No,
• 24:11
  Speaker 1: he's not borrowing.
• 24:12
  Speaker 2: Yeah.
• 24:12
  Speaker 1: It's just multiple stories. But that's the thing, though. This
• 24:15
  Speaker 1: entire industry, on investor sentiment, sure, but- The problem is,
• 24:20
  Speaker 1: is that they're no longer dealing with equity investors. They're
• 24:23
  Speaker 1: now dealing with the bond market.
• 24:25
  Speaker 2: Exactly.
• 24:26
  Speaker 1: The bond market, which is a lot meaner, because you
• 24:29
  Speaker 1: can pump... Well, the thing, the way I described it
• 24:32
  Speaker 1: to a mate of mine in fixed income, hello, Mike,
• 24:35
  Speaker 1: is that the stock market measures how excited you should be,
• 24:38
  Speaker 1: and the bond market measures how scared you should be.
• 24:41
  Speaker 1: And I think that it's obviously because, yeah, you actually
• 24:48
  Speaker 1: need to pay back the bond market. You can do
• 24:50
  Speaker 1: all this dodgy shit, talking up your stock and your
• 24:52
  Speaker 1: book all you want. But in the bond market, you
• 24:55
  Speaker 1: actually have to pay someone. Yeah. And I don't know,
• 24:59
  Speaker 1: I don't think that... Well, Anthropic, when they were raising
• 25:01
  Speaker 1: for this Broadcom chip lease deal a few months ago,
• 25:04
  Speaker 1: were refusing to share their financials with some investors. You
• 25:09
  Speaker 1: can't play this charlatan shit in the bond market. And
• 25:13
  Speaker 1: so it's not impossible for them to raise debt, but
• 25:15
  Speaker 1: if they're looking for $ 50 billion a year... I just
• 25:19
  Speaker 1: don't think that's going to happen. This is the other
• 25:23
  Speaker 1: thing that pisses me off about this S-1. And I
• 25:25
  Speaker 1: feel like we need new financial regulations. They should have
• 25:29
  Speaker 1: to say how much money is going out the door.
• 25:31
  Speaker 1: They should have to estimate exactly how much they're going
• 25:34
  Speaker 1: to have to prepay. Because their tricksy thing they did
• 25:36
  Speaker 1: was they said, okay, here are our minimum spend commitments.
• 25:40
  Speaker 1: What are the prepayments? Because it's obviously going to be
• 25:44
  Speaker 1: higher than they're saying. Yeah.
• 25:46
  Speaker 2: It's the most crucial detail in the entire document. And
• 25:49
  Speaker 2: it's not there. Exactly. And it's not there. It's kind
• 25:52
  Speaker 2: of there, but not really. But just to your point,
• 25:56
  Speaker 2: the stock market measures how excited... Say that again?
• 26:02
  Speaker 1: Stock market measures how excited you should be, bond market
• 26:05
  Speaker 1: measures how scared you should be.
• 26:06
  Speaker 2: How scared you should be. Okay. So then my next
• 26:07
  Speaker 2: question is, what does venture capital measure? Oh, fuck, how
• 26:13
  Speaker 2: stupid you are. Yeah, that's beyond, that's another level. I mean,
• 26:20
  Speaker 2: the venture capital industry, the level of euphoria and hysteria
• 26:26
  Speaker 2: around AI and the myopic way that they see the
• 26:30
  Speaker 2: world where they really do live in a genuine bubble.
• 26:34
  Speaker 2: And they have literally been drinking this AI Kool-Aid through
• 26:37
  Speaker 2: a beer bong for years at this point. I mean,
• 26:40
  Speaker 2: what on earth are they measuring? And so this is
• 26:44
  Speaker 2: my problem. There are different levels of discernment throughout the
• 26:48
  Speaker 2: global markets. And we're only seeing anthropic face the music
• 26:52
  Speaker 2: when it enters the public markets as opposed to the
• 26:55
  Speaker 2: private markets. But your point is an important one, which
• 26:58
  Speaker 2: is like, we haven't even gotten to fixed income land yet.
• 27:02
  Speaker 2: We haven't gotten to the debt investors, which are even
• 27:05
  Speaker 2: meaner and crueler and more discerning than the rest of
• 27:08
  Speaker 2: them put together. Yeah. And so this is only the
• 27:11
  Speaker 2: beginning of the story. Like you're describing stock market investors
• 27:16
  Speaker 2: who actually by comparison to, uh, to private investors and
• 27:21
  Speaker 2: VCs are seemingly quite, quite difficult. Um, but yeah, there's,
• 27:28
  Speaker 2: there's more to come where that came from.
• 27:29
  Speaker 1: And the other thing with that is as well is
• 27:32
  Speaker 1: those like that debt they will be raising. And this
• 27:35
  Speaker 1: is assuming that things go perfectly because they will need to.
• 27:39
  Speaker 1: will be raised in a very different climate than even
• 27:42
  Speaker 1: about six months ago. Because about six months ago, Corweave
• 27:45
  Speaker 1: was issuing debt. They were taking 9.25%, but they were
• 27:49
  Speaker 1: issuing debt. There was still hunger for both high-yield debt
• 27:54
  Speaker 1: and also AI-related debt. I don't know. How does anyone
• 27:59
  Speaker 1: loaning money to anyone feel about a negative a bit...
• 28:04
  Speaker 1: software company. I can't find a comparison for Anthropic anywhere
• 28:09
  Speaker 1: in the market, anywhere. Software companies don't raise this much money.
• 28:14
  Speaker 1: Software companies don't need this much money. The whole point
• 28:16
  Speaker 1: of a software company is that they aren't this needy.
• 28:22
  Speaker 1: But returning to the venture capital thing at this point
• 28:25
  Speaker 1: as well, the other thing is the word is gullible
• 28:28
  Speaker 1: with venture capital. Because you could be gullible and still
• 28:31
  Speaker 1: make money. But- venture capital has been hopping on about
• 28:35
  Speaker 1: for years. They've been on a historic losing streak since
• 28:38
  Speaker 1: like 2017. Like the average venture capital funders makes you
• 28:41
  Speaker 1: about somewhere between 90 cents and $ 1. 30 for each dollar
• 28:45
  Speaker 1: you invest, which is not being the S & P 500.
• 28:48
  Speaker 1: And so I think the actual answer here is, is
• 28:53
  Speaker 1: that people give way too much credit to venture. The
• 28:56
  Speaker 1: venture capital is nowhere near, outside of like SpaceX, Anthropic, Rivian, Cerebrus,
• 29:03
  Speaker 1: and Cerebrus is now tanked, what, 50% from its valuation,
• 29:07
  Speaker 1: maybe more? I mean, for the most part, venture capital
• 29:11
  Speaker 1: has absolutely not, it's kind of absolutely failed right now.
• 29:16
  Speaker 1: Like it's just, and by the way, Cerebrus is down 70%
• 29:20
  Speaker 1: over the past six months. That's fucking crazy.
• 29:23
  Speaker 2: I knew that company was a joke when I saw
• 29:25
  Speaker 2: their filing and one of the first pages in the
• 29:28
  Speaker 2: document said, we have the biggest chips in the world.
• 29:32
  Speaker 2: That was the thing.
• 29:34
  Speaker 1: The biggest chips, we love them. Yeah, it really is like, no,
• 29:38
  Speaker 1: it's Trumpism, but.
• 29:41
  Speaker 1: The funny thing is, and the point I'm making, is
• 29:43
  Speaker 1: that everyone, for some reason, gives venture capital this insane credit,
• 29:50
  Speaker 1: when in fact, the average VC is pretty fucking terrible
• 29:54
  Speaker 1: at their job. So while there'll be VC payoffs in
• 29:57
  Speaker 1: this IPO, Those people who invested at the $ 965 billion valuation,
• 30:09
  Speaker 1: it's not really a great return on investment. And those
• 30:13
  Speaker 1: people who invested in the $ 1. 25 trillion valuation for OpenAI,
• 30:19
  Speaker 1: they're never seeing that money again. That money's gone. It's over.
• 30:23
  Speaker 1: You wasted that money. I could see OpenAI never listing.
• 30:28
  Speaker 1: No one wants to talk about this with me for
• 30:31
  Speaker 1: some reason, but there is a very real chance that
• 30:34
  Speaker 1: OpenAI simply runs out of money. That OpenAI does not
• 30:38
  Speaker 1: go public because a $ 500 billion offering is not going
• 30:42
  Speaker 1: to be it's still going to be bad. Like it's
• 30:49
  Speaker 1: still going to be like, I don't think people are
• 30:51
  Speaker 1: really thinking any of this through because open AI, let's
• 30:55
  Speaker 1: say they go out of 500 billion. This is assuming
• 30:57
  Speaker 1: they go out at all. How the fuck do they
• 31:00
  Speaker 1: pay all their obligations? How do they pay their people?
• 31:03
  Speaker 1: How do they keep up with all the promises they've made?
• 31:06
  Speaker 1: What the hell happens? And the answer is no one
• 31:09
  Speaker 1: wants to think about it. No one wants to think
• 31:12
  Speaker 1: about all the difficult things because the difficult things make
• 31:15
  Speaker 1: the whole thing less fun.
• 31:17
  Speaker 2: This is the trouble with, with needing money is it
• 31:20
  Speaker 2: just makes you vulnerable. And the analogy that I, that
• 31:24
  Speaker 2: I have written about recently is like, it's sort of
• 31:27
  Speaker 2: like being a kid, like your fate, you have to
• 31:31
  Speaker 2: basically just place your entire fate in someone else's hands. Now,
• 31:35
  Speaker 2: when you're a kid, the good news for most kids
• 31:37
  Speaker 2: is that parents are willing to help them no matter
• 31:39
  Speaker 2: what happens. And, you know, they'll try to make sure
• 31:42
  Speaker 2: that you're set up for a good future. And that's
• 31:45
  Speaker 2: what a lot of the AI companies have been convinced
• 31:48
  Speaker 2: by when it comes to Silicon Valley. You have all
• 31:51
  Speaker 2: of these VCs who, in my view, are essentially bottle
• 31:54
  Speaker 2: service girls, but for founders. And their job is to
• 31:58
  Speaker 2: just make the founders feel happy and follow into whoever
• 32:03
  Speaker 2: led the previous round. And don't worry about the price.
• 32:07
  Speaker 2: That doesn't matter. What matters is partnering with an awesome founder.
• 32:11
  Speaker 2: And that's sort of the story that they've gotten used
• 32:13
  Speaker 2: to over the past few years. And they've convinced themselves
• 32:16
  Speaker 2: that Sequoia and Andreessen, they're all going to help them
• 32:19
  Speaker 2: no matter what. But what they forget is that actually
• 32:22
  Speaker 2: investors are not like your parents. Investors will let you die.
• 32:28
  Speaker 2: They do not give a shit. And we saw that
• 32:30
  Speaker 2: play out in 2022. I don't know if you remember
• 32:34
  Speaker 2: where you saw a correction in in the stock market
• 32:37
  Speaker 2: where a lot of tech stocks got hammered. Meta was
• 32:41
  Speaker 2: one of them. They saw the value come down by
• 32:44
  Speaker 2: around 70%. I mean, it was a really stunning correction
• 32:48
  Speaker 2: in 2022, which then, of course, trickled down to the
• 32:51
  Speaker 2: private markets. We had all of these companies, these, you know.
• 32:56
  Speaker 2: Buy now, pay later companies like Klarna and these one-click
• 33:01
  Speaker 2: payment companies and these ultra-fast delivery companies that had gotten
• 33:05
  Speaker 2: these ridiculous valuations. And then suddenly the investors in the
• 33:09
  Speaker 2: private markets were absolutely brutal to them. And they did
• 33:13
  Speaker 2: not give them the valuations that they wanted. They were
• 33:15
  Speaker 2: barely giving out the money. And all of these VCs
• 33:18
  Speaker 2: were going on their VC podcast industrial complex shows where
• 33:24
  Speaker 2: they would just talk about how the most important thing
• 33:26
  Speaker 2: for founders is being efficient and being responsible and tightening
• 33:30
  Speaker 2: the belt. And the reality is that was the same
• 33:33
  Speaker 2: person who was just a year ago talking about how
• 33:36
  Speaker 2: you need to be bold about your spending and you
• 33:38
  Speaker 2: need to be innovative. The truth is they will simply fail.
• 33:41
  Speaker 2: follow whatever is happening in the public markets. And so
• 33:46
  Speaker 2: whatever image of Silicon Valley these AI founders have gotten
• 33:50
  Speaker 2: over the past few years, that these guys are awesome
• 33:53
  Speaker 2: and they back us no matter what. And I love
• 33:56
  Speaker 2: these guys because they have conviction in my company. They
• 33:59
  Speaker 2: have conviction in me. You are in for a major
• 34:03
  Speaker 2: slap in the face because these guys do not have
• 34:06
  Speaker 2: conviction in anything. They will follow the price. They will
• 34:10
  Speaker 2: follow whoever the last investor was. And if the music
• 34:13
  Speaker 2: starts to get ugly, you can trust me that they
• 34:15
  Speaker 2: will start to become cruel as well. That's what's going
• 34:18
  Speaker 2: to happen here.
• 34:19
  Speaker 1: Well, let's talk about other potential convictions. Incredible segue there.
• 34:23
  Speaker 1: Let's talk about one, Mr. Jensen Huang, CEO of NVIDIA.
• 34:28
  Speaker 1: I've already had so many emails wondering about my comment here,
• 34:31
  Speaker 1: and the answer is, I talked about this shit in March.
• 34:34
  Speaker 1: March 2026, which was three to five months after the
• 34:37
  Speaker 1: original Megaspeed story. To catch everyone up, this is an
• 34:40
  Speaker 1: incredibly shortened version.
• 34:44
  Speaker 2: By the way, credit to you, because this is a
• 34:48
  Speaker 2: And it's something that journalists really, I mean, they should
• 34:54
  Speaker 2: have been all over. And by the way, I am
• 34:56
  Speaker 2: happy and respect that it was journalists that broke this.
• 35:00
  Speaker 2: And this, in some ways, is a win for investigative
• 35:03
  Speaker 2: journalism because it's so important. But the fact that you
• 35:06
  Speaker 2: were on this is huge. So I just want to
• 35:09
  Speaker 2: give you a credit there.
• 35:11
  Speaker 1: Back in October last year, the New York Times reported,
• 35:14
  Speaker 1: and then in December, Bloomberg reported, a company called Megaspeed,
• 35:17
  Speaker 1: which is a, I'm doing air quotes, Singaporean company, bought
• 35:21
  Speaker 1: $ 4. 8 billion worth of NVIDIA GPUs. But it also turns
• 35:26
  Speaker 1: out that an identical company called Shanghai Hexi, I think
• 35:29
  Speaker 1: it's called, existed in China with the same website as Megaspeed.
• 35:34
  Speaker 1: and also an investor deck that showed, in Megaspeed's case,
• 35:38
  Speaker 1: a bunch of data centers in Southeast Asia, and then
• 35:40
  Speaker 1: a few projected projects with no specific location. Well, it
• 35:44
  Speaker 1: turns out that there is an identical website for Shanghai XC,
• 35:47
  Speaker 1: which shows that those other locations were all in China.
• 35:51
  Speaker 1: And it turns out that the CEO of Megaspeed, one
• 35:55
  Speaker 1: Alice Huang, or Huang Lei, I think is her Chinese name,
• 35:59
  Speaker 1: turns out that she was knocking boots with the one
• 36:02
  Speaker 1: Jensen Huang. no relation, CEO of Nvidia, who was pictured
• 36:07
  Speaker 1: in the Bloomberg article, I think also in the New
• 36:10
  Speaker 1: York Times article last year, next to Jensen Huang, and
• 36:14
  Speaker 1: has been photographed many times for them, and has apparently
• 36:16
  Speaker 1: cavorted with the fella to the point that Mr. Huang
• 36:21
  Speaker 1: has actually admitted to this. He's actually said he apologized
• 36:24
  Speaker 1: to his family, his wife Laurie of 40 years, his employees.
• 36:28
  Speaker 1: And it turns out that Jensen may or may not
• 36:32
  Speaker 1: have been directly helping the Chinese buy NVIDIA GPUs through Megaspeed.
• 36:37
  Speaker 1: Now I am truncating the entire story because Megaspeed is
• 36:40
  Speaker 1: connected to a bunch of other companies The information had
• 36:43
  Speaker 1: the story saying that Megaspeed connected entities bought $ 7. 5 billion
• 36:47
  Speaker 1: worth of GPUs. But the biggest story for me is
• 36:51
  Speaker 1: if you look at all the pictures of Jensen Huang
• 36:53
  Speaker 1: with Alice Wong, again, no relation other than the biblical sense,
• 36:58
  Speaker 1: he looks happy. And so it turns out the biggest
• 37:02
  Speaker 1: part of the story, though, is that Alice Wong is
• 37:04
  Speaker 1: actually a Chinese spy. She's a spy for the Chinese,
• 37:08
  Speaker 1: who have denied this, of course, because, I mean, of
• 37:11
  Speaker 1: course they would. But the U.S. Commerce Department was also
• 37:14
  Speaker 1: looking into Megaspeed, and NVIDIA's general counsel, Tim Teeter, called them,
• 37:21
  Speaker 1: called multiple people there, and scuppered the investigation and got
• 37:24
  Speaker 1: it passed to the Department of Justice. All of this
• 37:27
  Speaker 1: is to say is that it very much looks like
• 37:30
  Speaker 1: the CEO of NVIDIA got targeted and shagged and fell
• 37:34
  Speaker 1: for a Chinese honeypot. And I just want to say,
• 37:37
  Speaker 1: finally some, like Gordon Ramsay, finally some good fucking stories.
• 37:41
  Speaker 1: I'm so bored of these boring assholes with their boring lives.
• 37:45
  Speaker 1: Chinese sex spies. We don't have to talk about the
• 37:48
  Speaker 1: singularity anymore. No more RSI. We're talking about sex. But
• 37:53
  Speaker 1: the thing is, this is a remarkable story, but just
• 37:56
  Speaker 1: to be clear, here's the tip of the iceberg. There's
• 37:59
  Speaker 1: so much more to come.
• 38:15
  Speaker 2: It is unbelievable. And my initial reaction to this news
• 38:19
  Speaker 2: is just, What an incredible story the AI boom has been.
• 38:23
  Speaker 2: I mean, there has been so much.
• 38:24
  Speaker 1: Thrilling.
• 38:25
  Speaker 2: Just thrilling drama. I mean, we could talk about, remember
• 38:29
  Speaker 2: when Sam Altman got fired from OpenAI?
• 38:31
  Speaker 1: Oh, yeah.
• 38:32
  Speaker 2: And then suddenly he was ousted and then he was
• 38:33
  Speaker 2: brought back in. I mean, that was an incredible story.
• 38:37
  Speaker 2: We had DeepSeek. We had... We had situational awareness being
• 38:42
  Speaker 2: blown up.
• 38:43
  Speaker 1: Twice now.
• 38:44
  Speaker 2: Twice now. I mean, there have just been so many
• 38:46
  Speaker 2: incredible plot points in the AI story. And now we
• 38:50
  Speaker 2: have Chinese sex spies honeypotting the CEO of what was
• 38:57
  Speaker 2: recently the most valuable company in the world. And of course,
• 39:01
  Speaker 2: no more, because it has, as I said, shed more
• 39:04
  Speaker 2: than $ 1. 5 trillion in market value over the past few weeks.
• 39:10
  Speaker 2: The whole thing is just incredible. This company, Megaspeed, clearly
• 39:18
  Speaker 2: they were doing something illegal, specifically smuggling AI chips from
• 39:23
  Speaker 2: Singapore to China. This woman, Alice Huang, she was a
• 39:28
  Speaker 2: TV reporter for Chinese state media And then in 2023,
• 39:31
  Speaker 2: she joins this company, Megaspeed, suddenly becomes CEO, suddenly becomes
• 39:36
  Speaker 2: one of the largest buyers of NVIDIA chips in the world,
• 39:39
  Speaker 2: bought nearly $ 5 billion in chips in just a few years.
• 39:43
  Speaker 2: And then at the same time, the company sells itself
• 39:46
  Speaker 2: to a Singaporean shell company, and then also operates out
• 39:50
  Speaker 2: of all of these empty offices. So clearly the implication
• 39:53
  Speaker 2: is they were smuggling to China. I think one of
• 39:55
  Speaker 2: the large questions is how much did Jensen Huang know
• 39:59
  Speaker 2: about this? Because if he was aware of what was
• 40:04
  Speaker 2: going on, then this is, I mean, he's essentially a fugitive.
• 40:10
  Speaker 2: I mean, this is one of the largest criminal scandals
• 40:15
  Speaker 2: in recent memory, because this is a national security issue.
• 40:19
  Speaker 2: with trillions and trillions of dollars on the line. And
• 40:21
  Speaker 2: this is the guy who was influencing export policy in America.
• 40:26
  Speaker 2: And by the way, this is the problem with Trump
• 40:29
  Speaker 2: and the problem with the crony capitalism that he has
• 40:33
  Speaker 2: made so common in the business world in America over
• 40:36
  Speaker 2: the past few years, which is the way that he
• 40:39
  Speaker 2: operates US policy is he transacts based on who gave
• 40:44
  Speaker 2: him a call lost. That's basically how it works. If
• 40:48
  Speaker 2: you ring Donald Trump, you call him on the phone,
• 40:51
  Speaker 2: you say, hey, you butter him up, you say some
• 40:53
  Speaker 2: nice things, you will get what you want. That's how
• 40:56
  Speaker 2: this works. And that is exactly what Jensen Huang has
• 40:58
  Speaker 2: been doing over the past several years, which is why
• 41:01
  Speaker 2: Trump eased the chip restrictions to China. Because Jensen Huang,
• 41:05
  Speaker 2: he's a nice guy. I like that guy. And he's
• 41:07
  Speaker 2: rich and he's the CEO of Nvidia and he's cool.
• 41:10
  Speaker 2: And so that's how we're going to run our policy.
• 41:12
  Speaker 2: And here we are, it's backfiring.
• 41:14
  Speaker 1: But the other thing is, is When you look at
• 41:17
  Speaker 1: the timing of all of this, so Jensen Huang donated,
• 41:21
  Speaker 1: or NVIDIA, I forget which one it was, invested $ 25
• 41:25
  Speaker 1: million in the Trump ballroom. And Trump had been informed
• 41:29
  Speaker 1: sometime in October, like, hey, Mr. Trump, sir, you big,
• 41:33
  Speaker 1: beautiful man. It turns out that Jensen Huang might actually
• 41:37
  Speaker 1: be possibly relating sexually with a Chinese spy. And Trump
• 41:42
  Speaker 1: was like, oh, it's okay. We don't need to worry
• 41:45
  Speaker 1: about that.
• 41:46
  Speaker 2: Yes.
• 41:46
  Speaker 1: And they just kept, they moved on from it. The
• 41:49
  Speaker 1: State Department, or the Commerce Department even, has been monitoring
• 41:53
  Speaker 1: Alice Wong since 2024. But what's crazy is NVIDIA's general counsel,
• 41:59
  Speaker 1: because Alice Wong was arrested by the Singaporean police sometime
• 42:04
  Speaker 1: in 2024. And she was detained and then let go. Apparently, Tim, Timmy, Timmy,
• 42:11
  Speaker 1: the general counsel over there was hounding the state department, sorry,
• 42:16
  Speaker 1: I keep saying the wrong one, Commerce Department, asking about
• 42:19
  Speaker 1: that detainment and whether they'd had anything on it, and
• 42:22
  Speaker 1: then basically muzzled the investigation. But the thing is, there
• 42:26
  Speaker 1: is this- Which.
• 42:27
  Speaker 2: Is something we've seen multiple times in this administration, by
• 42:29
  Speaker 2: the way.
• 42:30
  Speaker 1: Exactly.
• 42:30
  Speaker 2: This is not the first instance of an investigation and
• 42:33
  Speaker 2: criminal behavior being muzzled, and then the officials running those
• 42:37
  Speaker 2: investigations and pursuing them being ousted mysteriously. I mean, the
• 42:41
  Speaker 2: level of cover-up, I mean, we can also talk about
• 42:44
  Speaker 2: the Epstein files as well, which is in a similar vein,
• 42:47
  Speaker 2: but the amount of corruption and crime that has occurred
• 42:51
  Speaker 2: over the past several years and that we haven't gotten
• 42:53
  Speaker 2: to see justice or retribution for because they were in
• 42:57
  Speaker 2: bed with the police, because the police were either defunded, fired,
• 43:02
  Speaker 2: or essentially told, you're not allowed to pursue this thing,
• 43:05
  Speaker 2: is mind-blowing. And this is yet another example, and what
• 43:09
  Speaker 2: do you know, It applies to the AI boom and
• 43:12
  Speaker 2: the semiconductor trade. So I just want to just point out, like,
• 43:15
  Speaker 2: this is a thing that's happened multiple times.
• 43:19
  Speaker 1: But the other thing is, as well, is what this means.
• 43:22
  Speaker 1: So there was a short report from Culper in May,
• 43:25
  Speaker 1: who's getting a lot of credit, as they should, because
• 43:27
  Speaker 1: they dug into this really deep and found, for example,
• 43:31
  Speaker 1: the Megaspeed's biggest client is Alibaba. And also Alibaba gave
• 43:37
  Speaker 1: them an indeterminately large loan. to buy GPUs, and that's
• 43:41
  Speaker 1: how Alice Wong met Jensen Wong. But the other thing
• 43:45
  Speaker 1: that Culper said is, based on their numbers, that 20%
• 43:51
  Speaker 1: or more of Nvidia's fiscal year 2026 revenue came from China.
• 43:56
  Speaker 1: That's over $ 43 billion. This is not just Megaspeed. There
• 44:02
  Speaker 1: is obviously a much bigger thing going on here, And
• 44:06
  Speaker 1: credit to just Dario, one of my friends of the show,
• 44:08
  Speaker 1: who in 2023 had an incredible thread of the preferred
• 44:14
  Speaker 1: partners of NVIDIA and the elite partners. One of them
• 44:17
  Speaker 1: was a Russian murder shed. It was like when you
• 44:20
  Speaker 1: go and look at the address, it's literally just what
• 44:23
  Speaker 1: looks like a place you get murdered in a horror film.
• 44:27
  Speaker 1: One of them was just an empty Taiwanese office above
• 44:31
  Speaker 1: a wine shop.
• 44:32
  Speaker 2: Yep.
• 44:33
  Speaker 1: It's so obvious that something is going on here and
• 44:37
  Speaker 1: it's really big. And like 20% of Nvidia's earnings from
• 44:42
  Speaker 1: last year could have come from crime. Like from actual,
• 44:48
  Speaker 1: and here's the thing. I don't know.
• 44:49
  Speaker 2: More than that, by the way, because what we're learning
• 44:52
  Speaker 2: about this is that this money wasn't coming from China.
• 44:55
  Speaker 2: It was coming from Singapore, but ultimately then being funneled
• 44:59
  Speaker 2: over to China. This was essentially a workaround. during the
• 45:03
  Speaker 2: period where there were chip restrictions in order to sell
• 45:07
  Speaker 2: chips to China. And so what it basically tells you
• 45:09
  Speaker 2: is that there might actually be a lot more than
• 45:13
  Speaker 2: that 20% that are being sold to China. It might
• 45:16
  Speaker 2: be that there are shell companies that are operating all
• 45:19
  Speaker 2: over the world and then smuggling those chips to China.
• 45:23
  Speaker 2: So that's part of the problem here too, is like,
• 45:27
  Speaker 2: we actually don't know how reliant Nvidia truly is on China,
• 45:32
  Speaker 2: because this case shows you that they have been obfuscating
• 45:38
  Speaker 2: and smuggling their business into that region.
• 45:42
  Speaker 1: And the other thing is, it's very clear, because I'm
• 45:46
  Speaker 1: a curious little critter and This episode is going out
• 45:49
  Speaker 1: the day after my blog that talks about this, but
• 45:52
  Speaker 1: it turns out, do you remember last year when the
• 45:54
  Speaker 1: co-founder of Supermicro, along with several other people, was arrested?
• 45:58
  Speaker 1: He was arrested for smuggling hundreds of millions of dollars.
• 46:03
  Speaker 1: They said $ 2. 5 billion worth, but if you looked, only
• 46:06
  Speaker 1: a few hundred million dollars got over it, over there.
• 46:10
  Speaker 1: But apparently... Wally Liao, co-founder of Supermicro, who, by the way,
• 46:14
  Speaker 1: had been fired from the company due to a 2018
• 46:17
  Speaker 1: accounting scandal. He was working to help sell chips to
• 46:22
  Speaker 1: a company in China that was also a subsidiary of Megaspeed.
• 46:30
  Speaker 1: So it's all connected. And I love this. It's so
• 46:33
  Speaker 1: exciting to be part of a conspiracy that's real. 100%.
• 46:36
  Speaker 1: Because you can really spiral out. 100%.
• 46:40
  Speaker 2: This is the thing with conspiracy theories. People write them
• 46:43
  Speaker 2: off as conspiracy theories, and then they turn out to
• 46:45
  Speaker 2: be maybe not completely correct, but mostly correct.
• 46:49
  Speaker 1: Yeah.
• 46:50
  Speaker 2: I mean, this is what I loved about the Epstein files.
• 46:52
  Speaker 2: It's like, oh, there's a secret cabal of billionaires who
• 46:54
  Speaker 2: are having sex with children in a pizza restaurant. Okay,
• 46:58
  Speaker 2: the last part's not correct, but yeah, they are having
• 47:00
  Speaker 2: sex with children. They're just doing it over on an
• 47:03
  Speaker 2: island and not in the basement of a restaurant in D.C.,
• 47:07
  Speaker 2: I mean, these conspiracy theories are grounded in generally in
• 47:12
  Speaker 2: some level of truth. And they might not be entirely correct,
• 47:16
  Speaker 2: but in some cases they are at least directionally correct.
• 47:19
  Speaker 2: And yeah, this is another example of that playing out.
• 47:23
  Speaker 1: But that actually, it's good you brought up Epstein because
• 47:27
  Speaker 1: people like, oh, it wouldn't be this obvious. Jensen wouldn't
• 47:30
  Speaker 1: be this sloppy. He wouldn't be cavorting with this woman
• 47:33
  Speaker 1: photographed everywhere.
• 47:34
  Speaker 2: Because he's drunk on power and success.
• 47:37
  Speaker 1: But also if you look at the Epstein files, they're like, Hey, Jeffrey,
• 47:41
  Speaker 1: how's it going? You want to do the weird pedophile pie? Absolutely,
• 47:44
  Speaker 1: I'll be there. I can't make it this time. Did
• 47:47
  Speaker 1: you see this drawing I did of all of it?
• 47:50
  Speaker 2: Exactly.
• 47:51
  Speaker 1: And no one went to fucking prison. Not a single
• 47:53
  Speaker 1: fucking one of them.
• 47:55
  Speaker 2: Hopefully that changes.
• 47:56
  Speaker 1: No, that should change. But the thing is, I think
• 47:59
  Speaker 1: Jensen might have to go to prison. If he doesn't,
• 48:04
  Speaker 1: I don't know why. Because he appears to have directly...
• 48:10
  Speaker 1: aggressively shagged his way into a conspiracy himself. He appears to,
• 48:15
  Speaker 1: he is clearly involved in this on a material and
• 48:19
  Speaker 1: physical level in a way that is clearly bigger than
• 48:24
  Speaker 1: just Megaspeed. It's clearly Megaspleed, Megaspleed, Megaspeed plus a hundred
• 48:29
  Speaker 1: other little subsidiaries. And also, I don't know, now I've
• 48:34
  Speaker 1: seen this and now I've seen my murder shed and
• 48:38
  Speaker 1: Is Russia involved in this? Are chips getting over to Russia?
• 48:41
  Speaker 1: That's a question that we need answered too, because this
• 48:44
  Speaker 1: might not just be China. This might be something completely much,
• 48:50
  Speaker 1: much larger. I don't know. But at the same time,
• 48:53
  Speaker 1: Nvidia's earnings are, as people are listening to this today,
• 48:57
  Speaker 1: and if they are not fucking perfect, and the guidance
• 49:02
  Speaker 1: is not big and sexy and beautiful, I don't know
• 49:05
  Speaker 1: what Nvidia... I think Nvidia could be The thing that
• 49:08
  Speaker 1: finally flushes the toilet on this. Now, I think they
• 49:11
  Speaker 1: beat and raise because, I don't know, as we'll be discussing,
• 49:14
  Speaker 1: they found all sorts of creative ways to get in revenue.
• 49:17
  Speaker 1: But there's a certain degree here of, when? as when
• 49:21
  Speaker 1: does the music stop, but it's when does Wile E.
• 49:23
  Speaker 1: Coyote look down? Because that really is it. We've got
• 49:26
  Speaker 1: so much different shit going wrong. So many different things
• 49:31
  Speaker 1: are poking the bubble with sharp objects. It feels inevitable
• 49:37
  Speaker 1: that something else topples it.
• 49:39
  Speaker 2: Yeah, I think the question in terms of will Jensen
• 49:43
  Speaker 2: Huang go to prison, he certainly could. The question is,
• 49:48
  Speaker 2: how much did he know and what can be proven
• 49:51
  Speaker 2: about what he knew? It'll be sort of like the
• 49:53
  Speaker 2: Lindsay Clancy of AI, like how much agency over this
• 49:58
  Speaker 2: did he have or was he truly honeypotted and he
• 50:01
  Speaker 2: had no idea what was going on? I kind of
• 50:03
  Speaker 2: like to think that he's a smart guy. He probably
• 50:08
  Speaker 2: knew something. I mean, this is a Chinese national who
• 50:12
  Speaker 2: he decided to have a relationship with who suddenly came
• 50:16
  Speaker 2: into billions of dollars randomly after being a TV journalist
• 50:20
  Speaker 2: for Chinese state media. Like, you'd think that he would
• 50:23
  Speaker 2: have a few warning signals start to flash. So I
• 50:28
  Speaker 2: kind of think he probably knew something, but I don't know,
• 50:32
  Speaker 2: at the same time, your point about, you know, the
• 50:35
  Speaker 2: Epstein emails and the fact that these guys were so sloppy.
• 50:39
  Speaker 2: And so, I mean, that was what was so fascinating
• 50:41
  Speaker 2: about these emails. Like they're literally just texting each other
• 50:44
  Speaker 2: about this stuff. All of it's out in the open.
• 50:47
  Speaker 2: And I think the reality is when you achieve something,
• 50:51
  Speaker 2: That level of power. And when you are surrounded by
• 50:54
  Speaker 2: as many yes-men as these people surround themselves with, you
• 50:57
  Speaker 2: do get very, very sloppy. And you start to think
• 51:01
  Speaker 2: you're invincible. You start to think everything's going to be fine.
• 51:03
  Speaker 2: I think the same thing about Trump. I mean, the
• 51:05
  Speaker 2: fact that he goes out the day before the election
• 51:08
  Speaker 2: and starts talking about the reflecting pool and thinks that
• 51:10
  Speaker 2: that's going to be enough to change Americans' minds after
• 51:13
  Speaker 2: he completely crippled American workers around the country with Iran
• 51:17
  Speaker 2: and tariffs. I mean, some of the worst economic policy
• 51:20
  Speaker 2: decisions in the history of America. And he thinks it's
• 51:23
  Speaker 2: all fine because he's surrounded himself in this bubble of
• 51:27
  Speaker 2: obsequience and acceptance. And it's the same for a lot
• 51:31
  Speaker 2: of these guys. So this to me is another example
• 51:34
  Speaker 2: of that. Any regular person would have realized This is
• 51:38
  Speaker 2: probably a bad idea. I probably shouldn't have a romantic
• 51:41
  Speaker 2: relationship with this person who is clearly some sort of
• 51:45
  Speaker 2: Chinese intelligence asset. That's probably not the right move. But
• 51:50
  Speaker 2: when you're surrounded by all these people saying, no, Jensen
• 51:52
  Speaker 2: Huang is the greatest guy ever. He's created a fortune.
• 51:55
  Speaker 2: He's an excellent human. He's awesome. And I mean, if
• 52:00
  Speaker 2: that's your life experience, then... I start to understand why
• 52:03
  Speaker 2: you could become as delusional as these people have become.
• 52:06
  Speaker 2: But now reality is catching up to them. Now we're
• 52:10
  Speaker 2: seeing what happens when people open their eyes.
• 52:13
  Speaker 1: Well, I think that I am serious. As is disclosed
• 52:17
  Speaker 1: in my blog, I think Jensen liked her. I have
• 52:20
  Speaker 1: looked at no less than 50 pictures of Jensen Huang
• 52:23
  Speaker 1: in the last 48 hours. And I've compared them to
• 52:26
  Speaker 1: the photos of him with Alice Huang. No relation. And
• 52:30
  Speaker 1: he is goofy. He looks happy. He looks relaxed. In
• 52:37
  Speaker 1: all the other ones, he has this cold sternness, but
• 52:39
  Speaker 1: in this one, he looks.
• 52:40
  Speaker 2: Kind of like- It's almost surprise.
• 52:43
  Speaker 1: Yeah, exactly.
• 52:44
  Speaker 2: He looks shocked at how things worked out.
• 52:46
  Speaker 1: Here's the thing. I found this photo of her, though,
• 52:49
  Speaker 1: the real damning one, is she's got this big smile
• 52:52
  Speaker 1: on her face, and then she has this loose drop
• 52:54
  Speaker 1: of a hand, arm completely around the neck. As I
• 52:57
  Speaker 1: said in the blog, ladies, you see another woman with
• 53:00
  Speaker 1: her hand like that on your boy's shoulder, watch out.
• 53:05
  Speaker 2: Red flag.
• 53:06
  Speaker 1: She's working for the Chinese. No, that's not what I'm saying.
• 53:10
  Speaker 1: She's working with every woman who does this as a
• 53:12
  Speaker 1: Chinese agent.
• 53:13
  Speaker 2: Well, I do wonder how many Chinese spies are out there.
• 53:16
  Speaker 1: I am actually going to make a point, though. On Twitter,
• 53:20
  Speaker 1: I have seen something fucking vile, which is, oh, now
• 53:24
  Speaker 1: I've seen the Jensen thing. Check this out. And it's
• 53:26
  Speaker 1: just every tech executive with an Asian wife. And it's
• 53:29
  Speaker 1: like fucking racist. That's what that is. It's like, oh,
• 53:32
  Speaker 1: an Asian married a white guy. They're a spy. It's like, no,
• 53:36
  Speaker 1: go fuck yourself. You fucking racists. Right. Also Larry Ellison,
• 53:40
  Speaker 1: just from what I hear, likes Asian woman. That's a
• 53:42
  Speaker 1: quote from a venture capitalist I know. So like, I,
• 53:45
  Speaker 1: you know, what, what Larry would like to do is
• 53:47
  Speaker 1: up to him. But also it's just, this is not
• 53:50
  Speaker 1: that this, what happened to Jensen is very clearly the,
• 53:55
  Speaker 1: he is on the road all the time. And also
• 53:59
  Speaker 1: very greedy and wants to pump Nvidia's earnings. And met
• 54:03
  Speaker 1: a woman who could help him do both. And also
• 54:06
  Speaker 1: clearly fell for her. Clearly fell for this kind of like, oh,
• 54:09
  Speaker 1: this sophisticated woman who understands. God, the conversations must have
• 54:14
  Speaker 1: been fucking boring. Oh, I got to sell so many GPUs.
• 54:18
  Speaker 1: And she's like, yeah, yeah, you should ship them to Malaysia. Yeah.
• 54:22
  Speaker 1: You should ship those GPUs to Malaysia, Jensen. Or, of course,
• 54:27
  Speaker 1: he fully knew. Or perhaps he was trying to impress her.
• 54:31
  Speaker 1: That's also a good one. Because this is the thing.
• 54:34
  Speaker 1: He's fucked up in a way that so many guys
• 54:36
  Speaker 1: fuck up. They blow up their lives on a romance
• 54:39
  Speaker 1: scam in Southeast Asia. I mean, that's like 50% of
• 54:43
  Speaker 1: Facebook's revenue.
• 54:44
  Speaker 2: Isn't it? This is sort of his equivalent of some
• 54:47
  Speaker 2: sort of OnlyFans scam. It's just that now, instead of
• 54:51
  Speaker 2: couple of hundred dollars on the line, it's trillions of
• 54:53
  Speaker 2: dollars on the line. But the question that'll be really interesting,
• 54:57
  Speaker 2: the question will be really interesting is what is the
• 54:59
  Speaker 2: board going to do about this? Because a lot of
• 55:02
  Speaker 2: people believe that Nvidia isn't Nvidia without Jensen Huang. And
• 55:06
  Speaker 2: by the way, their governance structure is pretty shady because
• 55:09
  Speaker 2: they do not have a board chair because they've always
• 55:12
  Speaker 2: just assumed that Jensen Huang can handle everything.
• 55:14
  Speaker 1: What does that mean in practicality though?
• 55:18
  Speaker 2: It means that you don't have a balance of power
• 55:20
  Speaker 2: between the CEO versus the chairman to set the agenda.
• 55:24
  Speaker 2: So essentially, Jensen Huang is the de facto chairman. He
• 55:28
  Speaker 2: has a lot of power of the company. Now, he
• 55:30
  Speaker 2: doesn't have the level of voting power that you see
• 55:33
  Speaker 2: with people like Mark Zuckerberg, who set up these dual-class
• 55:36
  Speaker 2: shareholder structures so that they could determine and essentially never
• 55:41
  Speaker 2: be ousted from the company because they control more than 50%
• 55:43
  Speaker 2: of the votes. He has around 3.5%. But still... there
• 55:47
  Speaker 2: is no chair and there is no succession plan either,
• 55:50
  Speaker 2: because I think they have operated up until this point
• 55:53
  Speaker 2: under the assumption that Jensen Huang will kind of live
• 55:56
  Speaker 2: forever and he's the greatest and no problems whatsoever. And
• 56:01
  Speaker 2: now they've run into this problem where obviously he might
• 56:04
  Speaker 2: have seriously compromised national security, but at the same time,
• 56:09
  Speaker 2: he's also just exhibited extremely poor judgment, like just at
• 56:13
  Speaker 2: a very simple level. He fell into this romantic relationship
• 56:17
  Speaker 2: with a woman who appears to be a spy. And
• 56:20
  Speaker 2: that really just shakes your confidence in him as a leader,
• 56:23
  Speaker 2: as a manager, as an operator. And the question I
• 56:26
  Speaker 2: think will be, does the board think that this is disqualifying?
• 56:30
  Speaker 2: And I think it's a decent chance that they might.
• 56:33
  Speaker 2: I think it's a decent chance he will be out.
• 56:35
  Speaker 1: The question is, how much did the rest of NVIDIA know? Again,
• 56:39
  Speaker 1: the general counsel, Tim Teeter of NVIDIA. was hounding the
• 56:45
  Speaker 1: commerce department. Got it right on that time, then fucked
• 56:47
  Speaker 1: up the word commerce. And he was hounding them specifically
• 56:51
  Speaker 1: about stuff about Alice Wong being detained in 2024. It's
• 56:56
  Speaker 1: obvious that people at the company knew. And I should
• 57:01
  Speaker 1: also say, the California State Bar should have Timmy up
• 57:05
  Speaker 1: before him. Because Tata was saying, now here's the The disclaimer,
• 57:11
  Speaker 1: the FT did not cite where he made these comments.
• 57:13
  Speaker 1: He was saying that the Alice Wong affair was salacious
• 57:16
  Speaker 1: rumors days ago. But apparently they weren't rumors. They were
• 57:20
  Speaker 1: fucking true, Tim. Now, the reason I mention that is
• 57:23
  Speaker 1: California in 2019, 2018 added a rule about making false statements. So,
• 57:32
  Speaker 1: I don't know. If he said that anywhere that can
• 57:34
  Speaker 1: actually be quoted, I feel like that's a violation of his... He's, well, oath?
• 57:39
  Speaker 1: I don't know what they say. Someone's going to email
• 57:41
  Speaker 1: me and be like, it's not an oath when you
• 57:42
  Speaker 1: become a lawyer.
• 57:43
  Speaker 2: But nevertheless, I think- He should be subpoenaed. He should
• 57:47
  Speaker 2: be brought in front of Congress.
• 57:48
  Speaker 1: He absolutely should.
• 57:49
  Speaker 2: I think- Jensen Huang is going to be brought in
• 57:50
  Speaker 2: front of Congress, but this guy should be brought in
• 57:52
  Speaker 2: front as well. And they should do it publicly. Along
• 57:53
  Speaker 2: with a lot of people. They should do it publicly.
• 57:55
  Speaker 2: I mean, there is a debate in the democratic policy
• 58:00
  Speaker 2: right now about whether we should pursue retribution for all
• 58:03
  Speaker 2: of the crimes and all of this stuff my answer
• 58:05
  Speaker 2: is fuck yes we should we.
• 58:07
  Speaker 1: Should we should have i don't know if you knew
• 58:10
  Speaker 1: around 2021 when biden was getting sworn in a lot
• 58:13
  Speaker 1: of the q anon people were doing that that's gonna
• 58:15
  Speaker 1: be 24 7 trials Now, to be clear, that didn't
• 58:20
  Speaker 1: happen and shouldn't have happened. It should here.
• 58:23
  Speaker 2: Yeah. There are clearly crimes.
• 58:25
  Speaker 1: We should have the trial channel. And they should let
• 58:28
  Speaker 1: me be the one that interviews Sam Altman. Because I've
• 58:32
  Speaker 1: got some questions.
• 58:33
  Speaker 2: But in all seriousness, they should- It would generate a
• 58:35
  Speaker 2: lot of ad revenue.
• 58:36
  Speaker 1: Oh, my God. The US government would make billions of
• 58:39
  Speaker 1: dollars in ad revenue on my 11th question to Sam Altman.
• 58:42
  Speaker 2: It would generate a lot more money than the tariffs generated.
• 58:44
  Speaker 2: I can tell you that much.
• 58:45
  Speaker 1: Here's the thing. with that is we do need accountability,
• 58:49
  Speaker 1: but also we need to be able to see what's
• 58:53
  Speaker 1: going on. 100%. The public not only deserves to know,
• 58:59
  Speaker 1: but the public and the people voted in this way
• 59:02
  Speaker 1: as well, we need accountability for these people. We need
• 59:05
  Speaker 1: them to both be embarrassed, but also for to be
• 59:10
  Speaker 1: honest about what they're doing.
• 59:11
  Speaker 2: 100%. The idea that you wouldn't pursue justice, that you
• 59:16
  Speaker 2: wouldn't prosecute what clearly amounts to apparent crimes, and lots
• 59:22
  Speaker 2: of very serious crimes, insider trading, securities fraud, federal bribery,
• 59:28
  Speaker 2: abusive office. I mean, I could go on. The idea
• 59:32
  Speaker 2: that you wouldn't pursue justice out of fear of being
• 59:35
  Speaker 2: seen as political is a fundamental failure of the justice system.
• 59:40
  Speaker 2: It is both unjust and then also at the same time,
• 59:44
  Speaker 2: incredibly cowardly because you're not doing the right thing because
• 59:48
  Speaker 2: you're afraid of what people might think. You're afraid of
• 59:52
  Speaker 2: how people might vote in the next election. And I'm honestly...
• 59:55
  Speaker 2: I'm just sick of the cowardice. And we've talked about
• 59:58
  Speaker 2: this on the show before. I'm sick of the cowardice
• 1:00:00
  Speaker 2: in the government. I'm sick of the cowardice in business.
• 1:00:02
  Speaker 2: All of these people who don't seem to have any
• 1:00:05
  Speaker 2: spines whatsoever... And I want to see them actually stand
• 1:00:08
  Speaker 2: up and say, Hey, this is wrong. This was a crime.
• 1:00:12
  Speaker 2: We're not going to just punish you because for political reasons,
• 1:00:16
  Speaker 2: we're going to punish you if you broke the law.
• 1:00:18
  Speaker 2: So now we're going to put the justice system to work.
• 1:00:20
  Speaker 2: The whole point is to determine whether someone was guilty
• 1:00:24
  Speaker 2: or not. And if you're guilty, then yeah, you're going
• 1:00:26
  Speaker 2: to go to jail. You're going to face the penalty.
• 1:00:28
  Speaker 2: You're going to pay the price. And so the idea
• 1:00:30
  Speaker 2: that we're not going to do that because that's lawfare.
• 1:00:34
  Speaker 2: is just completely ridiculous. There is a difference between lawfare
• 1:00:38
  Speaker 2: and the law. And this is simply the law.
• 1:00:41
  Speaker 1: And to wrap things up as well, the same needs
• 1:00:44
  Speaker 1: to go for these executives. Because a really great way
• 1:00:48
  Speaker 1: to make sure this doesn't happen again is to make
• 1:00:51
  Speaker 1: it really fucking suck. Yeah. To make it genuine, to
• 1:00:55
  Speaker 1: create a social hazard along with any potential crimes you get.
• 1:01:00
  Speaker 2: Yes.
• 1:01:01
  Speaker 1: Like, knocked onto you. Because... Here's the thing. Jensen Huang
• 1:01:06
  Speaker 1: has obviously flagrantly helped Chinese smuggling. And Wally Liao, the
• 1:01:11
  Speaker 1: co-founder of Supermicro, has obviously gone to jail for doing this.
• 1:01:16
  Speaker 1: And I refuse to believe, by the way, that Charles Lang,
• 1:01:20
  Speaker 1: the CEO of Supermicro, had no idea that this was happening.
• 1:01:24
  Speaker 2: Right.
• 1:01:25
  Speaker 1: And this is the magical thing. This is the magical
• 1:01:28
  Speaker 1: thing about this. We now have an opportunity to skewer
• 1:01:31
  Speaker 1: these fucking people. To have them, we need a televised
• 1:01:35
  Speaker 1: trial for them, or even just a Senate panel. We
• 1:01:38
  Speaker 1: should have clips like we had with Enron. We should
• 1:01:42
  Speaker 1: have that. And this is not just my bloodlust, which
• 1:01:45
  Speaker 1: is there. I want it. I want to watch this.
• 1:01:48
  Speaker 1: I want to clip this. I want to add music.
• 1:01:50
  Speaker 1: I want to add the the, remember the, what was it?
• 1:01:52
  Speaker 2: What's the soundtrack going to be for this?
• 1:01:54
  Speaker 1: Well, here's the thing. If it's Jensen Huang admitting something
• 1:01:57
  Speaker 1: on camera, it's got to be, what was it? The Nigerian,
• 1:02:00
  Speaker 1: was it Nigerian funeral music? I got so much shit
• 1:02:05
  Speaker 1: ready for Sam Orman. I got a JoJo's Bizarre Adventure
• 1:02:07
  Speaker 1: clip or two. Like I've got so much to work
• 1:02:09
  Speaker 1: with here, but guess what? If you as a would
• 1:02:13
  Speaker 1: be criminal in the future, see these videos and interviews
• 1:02:17
  Speaker 1: and you go, I don't want to get involved in that.
• 1:02:18
  Speaker 1: They're really going to make me really going to make
• 1:02:20
  Speaker 1: me embarrassed that's far more powerful than any sanctions or
• 1:02:24
  Speaker 1: firings or anything it's when you're because when you are
• 1:02:27
  Speaker 1: this rich all you have is your name really yeah
• 1:02:30
  Speaker 1: that's right all right i'm gonna wrap up on that
• 1:02:33
  Speaker 1: wonderful note ed thank you so much for joining.
• 1:02:36
  Speaker 2: Me thank you always love having you on my show
• 1:02:38
  Speaker 2: too really.
• 1:02:39
  Speaker 1: Hell yeah and everyone you'll have your monologue on friday
• 1:02:42
  Speaker 1: of course you've been listening to better offline.
• 1:02:53
  Speaker 3: Thank you for listening to Better Offline. The editor and
• 1:02:55
  Speaker 3: composer of the Better Offline theme song is Matt Osowski.
• 1:02:58
  Speaker 3: You can check out more of his music and audio
• 1:03:00
  Speaker 3: projects at mattosowski.com.
• 1:03:03
  Speaker 1: M-A-T-T-O-S-O-W-S-K-I dot com.
• 1:03:08
  Speaker 3: You can email me at ez at betteroffline.com or visit
• 1:03:11
  Speaker 3: betteroffline.com to find more podcast links and, of course, my newsletter.
• 1:03:15
  Speaker 3: I also really recommend you go to chat.wheresyoured.at to visit
• 1:03:18
  Speaker 3: the Discord and go to.
• 1:03:20
  Speaker 1: R slash betteroffline to check out our Reddit. Thank you
• 1:03:23
  Speaker 1: so much for listening. Better Offline is a production of
• 1:03:27
  Speaker 1: Cool Zone Media.
• 1:03:28
  Speaker 3: For more from Cool Zone Media, visit our website, coolzonemedia.com,
• 1:03:32
  Speaker 3: or check us out on the iHeartRadio app, Apple Podcasts,
• 1:03:35
  Speaker 3: or wherever you get your podcasts.