WEBVTT - The AI Money Trap, Part Two

0:00:03.160 --> 0:00:03.560
<v Speaker 1>Media.

0:00:05.200 --> 0:00:07.880
<v Speaker 2>Hi, I'm Edzetron, and this is the second part of

0:00:07.920 --> 0:00:22.479
<v Speaker 2>this week's better offline. As ever, go to the episode

0:00:22.520 --> 0:00:25.239
<v Speaker 2>notes by some merch get the challenge coin, subscribe to

0:00:25.239 --> 0:00:28.800
<v Speaker 2>my newsletter if you like it, go premium. Either way,

0:00:28.960 --> 0:00:31.000
<v Speaker 2>you've got this episode for free if you can stand

0:00:31.040 --> 0:00:32.360
<v Speaker 2>the ads, which many of you can.

0:00:32.840 --> 0:00:33.240
<v Speaker 3>Anyway.

0:00:33.320 --> 0:00:35.280
<v Speaker 2>The thrust of this two part is that the AI

0:00:35.400 --> 0:00:37.880
<v Speaker 2>trade is going to, as you've probably guessed by now,

0:00:38.240 --> 0:00:40.599
<v Speaker 2>end badly, and as I'll also explain, I believe we're

0:00:40.640 --> 0:00:43.720
<v Speaker 2>now set up for potentially economic and market wrecking consequences

0:00:44.000 --> 0:00:48.320
<v Speaker 2>of this hysterical investment bubble. Today we're going to go

0:00:48.560 --> 0:00:51.000
<v Speaker 2>beyond just startups and into how the poison of the

0:00:51.040 --> 0:00:53.720
<v Speaker 2>bubble is crept into our economy and how dangerous things

0:00:53.760 --> 0:00:56.640
<v Speaker 2>are getting as a result. In the first installment, we

0:00:56.680 --> 0:00:59.800
<v Speaker 2>talked about Cursor, the AI coding startup that, despite having

0:00:59.800 --> 0:01:03.000
<v Speaker 2>no mote and zero sustainability, has somehow earned a valuation

0:01:03.080 --> 0:01:05.440
<v Speaker 2>of ten billion dollars, and then asked why none of

0:01:05.440 --> 0:01:08.160
<v Speaker 2>these generative AI companies are being acquired, at least not

0:01:08.280 --> 0:01:11.560
<v Speaker 2>in the traditional sense. And no, I'm not including the

0:01:11.560 --> 0:01:14.160
<v Speaker 2>acquisitions where a big company snags the talent and leaves

0:01:14.200 --> 0:01:17.880
<v Speaker 2>barely breathing bodies of startups in their wake. This matters

0:01:17.920 --> 0:01:20.720
<v Speaker 2>because if these AI companies can't get an exit, go public,

0:01:20.800 --> 0:01:23.759
<v Speaker 2>or get bought, it raises serious questions about how any

0:01:23.800 --> 0:01:26.200
<v Speaker 2>of this ends. Cursor is the best example, and the

0:01:26.240 --> 0:01:29.160
<v Speaker 2>fact that Anthropic is hopelessly dependent on Cursor for revenue

0:01:29.319 --> 0:01:31.640
<v Speaker 2>despite doing everything they can to kill it, makes the

0:01:31.800 --> 0:01:36.200
<v Speaker 2>entire situation feel a little bit weird. Cursor's paths to

0:01:36.280 --> 0:01:39.200
<v Speaker 2>viability seem frankly to be non existent, and the collapse

0:01:39.200 --> 0:01:41.320
<v Speaker 2>of Cursor will inevitably result in a vast chunk of

0:01:41.400 --> 0:01:44.560
<v Speaker 2>anthropics revenue is going up in smoke, which again will

0:01:44.600 --> 0:01:47.120
<v Speaker 2>also make its collapse seem inevitable, as how can it

0:01:47.160 --> 0:01:49.960
<v Speaker 2>fundraise that evaluation over its previous one when its biggest

0:01:49.960 --> 0:01:53.400
<v Speaker 2>customer just died? So much of what I'm describing as

0:01:53.440 --> 0:01:57.160
<v Speaker 2>structural and few companies have bigger structural obstacles to survive

0:01:57.200 --> 0:02:00.160
<v Speaker 2>all than open AI. As a reminder, open aia I

0:02:00.160 --> 0:02:02.320
<v Speaker 2>appears to have burned at least ten billion dollars in

0:02:02.360 --> 0:02:05.120
<v Speaker 2>the last two months. It has just raised another eight

0:02:05.120 --> 0:02:07.560
<v Speaker 2>point three billion dollars after raising ten billion dollars in

0:02:07.640 --> 0:02:09.800
<v Speaker 2>June according to The New York Times, and intends to

0:02:09.800 --> 0:02:12.040
<v Speaker 2>receive around twenty two point five billion dollars by the

0:02:12.120 --> 0:02:14.400
<v Speaker 2>end of the year from SoftBank, and that is assuming

0:02:14.440 --> 0:02:16.280
<v Speaker 2>it becomes a for profit entity by the end of

0:02:16.360 --> 0:02:18.680
<v Speaker 2>the year, and if that doesn't happen, the route gets

0:02:18.720 --> 0:02:21.079
<v Speaker 2>cut to twenty billion dollars total, meaning that SoftBank would

0:02:21.080 --> 0:02:22.560
<v Speaker 2>only be on the hook for a further one point

0:02:22.600 --> 0:02:25.880
<v Speaker 2>seven billion dollars. I am repeating myself, and I need

0:02:25.919 --> 0:02:28.520
<v Speaker 2>you to really get this. Open Ai just got ten

0:02:28.560 --> 0:02:31.600
<v Speaker 2>billion dollars in June, just in June, and had to

0:02:31.680 --> 0:02:34.200
<v Speaker 2>raise another eight point three billion dollars in August. And

0:02:34.240 --> 0:02:37.920
<v Speaker 2>that is an unbelievable cash burn, one dwarfing any startup

0:02:37.919 --> 0:02:40.760
<v Speaker 2>in history, raveled only by Xai, makers of Grok the

0:02:40.840 --> 0:02:43.760
<v Speaker 2>races dollar m which loses over a billion dollars a month.

0:02:44.440 --> 0:02:46.720
<v Speaker 2>I should also be clear that if open ai does

0:02:46.760 --> 0:02:48.840
<v Speaker 2>not convert to a for profit there is no path

0:02:48.919 --> 0:02:52.000
<v Speaker 2>forward to continue raising capital. Open Ai must have the

0:02:52.000 --> 0:02:54.720
<v Speaker 2>promise of an IPO. It must go public because a

0:02:54.919 --> 0:02:57.880
<v Speaker 2>valuation of three hundred billion dollars or more. Open Ai

0:02:57.960 --> 0:03:00.360
<v Speaker 2>can no longer be acquired because nobody has that much money.

0:03:00.360 --> 0:03:03.600
<v Speaker 2>And if Let's be real, nobody actually believes open ai

0:03:03.680 --> 0:03:06.120
<v Speaker 2>is worth that much. The only way to prove that

0:03:06.160 --> 0:03:08.320
<v Speaker 2>anybody does is to take open ai public, and that

0:03:08.360 --> 0:03:11.800
<v Speaker 2>will be impossible if it cannot convert. And ironically, SoftBank's

0:03:11.840 --> 0:03:16.079
<v Speaker 2>large and late stage participation makes exits actually harder, as

0:03:16.120 --> 0:03:19.000
<v Speaker 2>early investors will see their holdings diluted as a percentage

0:03:19.000 --> 0:03:21.400
<v Speaker 2>of total equity or whatever the hell we're calling it.

0:03:21.400 --> 0:03:25.400
<v Speaker 2>Because these aren't real equity shares, their profit participation units.

0:03:25.560 --> 0:03:28.440
<v Speaker 3>They're a nonprofit. You can't do stock in them.

0:03:28.800 --> 0:03:31.079
<v Speaker 2>While a normal company could just issue equity and deal

0:03:31.120 --> 0:03:34.040
<v Speaker 2>with the dilution that way, open AI's structure necess states

0:03:34.040 --> 0:03:36.920
<v Speaker 2>the negotiation, where companies can obstruct the entire process if

0:03:36.920 --> 0:03:40.560
<v Speaker 2>they see fit. Speaking of companies doing that, let's talk

0:03:40.600 --> 0:03:43.440
<v Speaker 2>about Microsoft. As I asked in my premium newsletter a

0:03:43.480 --> 0:03:46.280
<v Speaker 2>few weeks ago, what if Microsoft does not want open

0:03:46.320 --> 0:03:49.640
<v Speaker 2>ai to convert. They own all the IP, they own

0:03:49.680 --> 0:03:52.240
<v Speaker 2>all the access to open AI's research, and already run

0:03:52.320 --> 0:03:55.600
<v Speaker 2>most of their infrastructure, while assuming the best case scenario

0:03:55.720 --> 0:03:57.440
<v Speaker 2>that it would end up owning a massive junk of

0:03:57.440 --> 0:03:59.360
<v Speaker 2>the biggest tech stile up of all time. And I'm

0:03:59.400 --> 0:04:02.360
<v Speaker 2>talking about equity not open AI's current profit sharing units.

0:04:02.600 --> 0:04:05.400
<v Speaker 2>Microsoft might also believe that it stands to gain more

0:04:05.600 --> 0:04:08.160
<v Speaker 2>by letting open ai die and assuming its role in

0:04:08.200 --> 0:04:11.960
<v Speaker 2>the AI ecosystem. But let's assume that open ai converts

0:04:12.000 --> 0:04:15.240
<v Speaker 2>and now has to continue raising money at a rate

0:04:15.280 --> 0:04:18.000
<v Speaker 2>that will require it allegedly to only need to raise

0:04:18.040 --> 0:04:21.440
<v Speaker 2>seventeen billion dollars in twenty twenty seven. That number doesn't

0:04:21.440 --> 0:04:23.960
<v Speaker 2>make any sense considering the open ai already had to

0:04:23.960 --> 0:04:26.400
<v Speaker 2>bring forward its eight point three billion dollar fund raised

0:04:26.400 --> 0:04:29.240
<v Speaker 2>by at least three months. But let's stick with that idea.

0:04:29.400 --> 0:04:32.480
<v Speaker 2>Open ai believes it will be profitable somehow by twenty thirty,

0:04:32.520 --> 0:04:34.480
<v Speaker 2>and even if we assume that that means it intends

0:04:34.480 --> 0:04:37.279
<v Speaker 2>to burn over one hundred billion dollars total to get there,

0:04:37.760 --> 0:04:40.680
<v Speaker 2>is the plan to take open ai public and then

0:04:40.760 --> 0:04:45.279
<v Speaker 2>dumper toxic acx onto the public markets. Then it will

0:04:45.320 --> 0:04:48.320
<v Speaker 2>just flounder and convulse and die for everybody to seek.

0:04:48.720 --> 0:04:51.640
<v Speaker 2>Kind of sucks. Can you imagine open aiss one? How

0:04:51.680 --> 0:04:54.240
<v Speaker 2>old do you think this company would do when they

0:04:54.360 --> 0:04:57.200
<v Speaker 2>had a true financial audit from a major accounting firm,

0:04:57.320 --> 0:04:59.599
<v Speaker 2>and you can claim they already have one. I simply

0:04:59.600 --> 0:05:02.600
<v Speaker 2>don't belie leave you, this burn does not seem like

0:05:02.640 --> 0:05:05.720
<v Speaker 2>an accountancy firm would be good with them. And if

0:05:05.760 --> 0:05:07.880
<v Speaker 2>you want to know what all that looks like Google

0:05:07.880 --> 0:05:10.279
<v Speaker 2>we Work, which went from tech industry darling to joke

0:05:10.320 --> 0:05:12.200
<v Speaker 2>in a matter of days in part because it was

0:05:12.240 --> 0:05:14.560
<v Speaker 2>forced to disclose how bad things actually were when they

0:05:14.560 --> 0:05:16.880
<v Speaker 2>filed to go public. No really are linked to an

0:05:16.960 --> 0:05:19.200
<v Speaker 2>article I've linked in the spreadsheet called the Strangest and

0:05:19.240 --> 0:05:22.080
<v Speaker 2>most Alarming things in we Work's IPO filing. It's a

0:05:22.080 --> 0:05:24.719
<v Speaker 2>bad sign if your IPO filing is described as strange

0:05:24.720 --> 0:05:28.080
<v Speaker 2>and alarming. Those are not even terms of art, they're

0:05:28.080 --> 0:05:30.880
<v Speaker 2>just worrisome. With that in mind, I feel the same

0:05:30.880 --> 0:05:33.640
<v Speaker 2>way about Anthropic. Nobody is buying this company a one

0:05:33.720 --> 0:05:35.680
<v Speaker 2>hundred and seventy billion dollars, which is how much it's

0:05:35.720 --> 0:05:37.680
<v Speaker 2>raising it right now, and thus the only way to

0:05:37.720 --> 0:05:40.599
<v Speaker 2>access liquidity would be to take Anthropic public and show

0:05:40.600 --> 0:05:42.920
<v Speaker 2>the world how a company that made seventy two million

0:05:42.960 --> 0:05:45.599
<v Speaker 2>dollars in January and more than four hundred million dollars

0:05:45.600 --> 0:05:48.320
<v Speaker 2>in July twenty twenty five and also plans to lose

0:05:48.320 --> 0:05:51.680
<v Speaker 2>three billion dollars or more, and then you show them

0:05:51.720 --> 0:05:54.280
<v Speaker 2>that and let the market sign a fair price, especially

0:05:54.279 --> 0:05:56.440
<v Speaker 2>if the S one includes a bunch of strange and

0:05:56.480 --> 0:06:00.000
<v Speaker 2>alarming stuff that raises questions about whether anthropics value is

0:06:00.000 --> 0:06:06.800
<v Speaker 2>actually justified. I'm I'm stuttering here because everyone's acting like

0:06:06.839 --> 0:06:09.680
<v Speaker 2>these irrational things like yeah, our open ai, which burns

0:06:09.839 --> 0:06:12.520
<v Speaker 2>five billion dollars in twenty twenty four, probably ten or

0:06:12.600 --> 0:06:15.880
<v Speaker 2>more in twenty twenty five. If if not fifteen or twenty,

0:06:16.000 --> 0:06:17.960
<v Speaker 2>like that's fine, Like this is all good.

0:06:17.960 --> 0:06:20.719
<v Speaker 3>They'll just go public one day, will they really?

0:06:20.920 --> 0:06:21.240
<v Speaker 2>Really?

0:06:21.520 --> 0:06:22.200
<v Speaker 3>Will they? No?

0:06:22.360 --> 0:06:25.960
<v Speaker 2>Really, what's the plan? What's the plan here? Because the

0:06:26.040 --> 0:06:28.360
<v Speaker 2>arguments against my work always come down to costs will

0:06:28.400 --> 0:06:31.120
<v Speaker 2>go down and these products will become essential. Outside of

0:06:31.200 --> 0:06:33.520
<v Speaker 2>chat GPT, there's no real proof that these products are

0:06:33.520 --> 0:06:37.080
<v Speaker 2>anything remotely essential, And I'd argue there's very little about

0:06:37.120 --> 0:06:40.120
<v Speaker 2>chat GPT that Microsoft can't provide with rate limits via

0:06:40.200 --> 0:06:42.840
<v Speaker 2>co pilot. I'd also argue that essential is a very

0:06:42.880 --> 0:06:46.280
<v Speaker 2>subjective term, essentially in the sense that some people use

0:06:46.320 --> 0:06:49.080
<v Speaker 2>it as search doesn't mean it's useful for enterprises or

0:06:49.120 --> 0:06:51.480
<v Speaker 2>the majority of people, or could not be replaced by

0:06:51.520 --> 0:06:55.520
<v Speaker 2>anything else. And I guess chat GPT somehow makes one

0:06:55.600 --> 0:06:58.080
<v Speaker 2>billion dollars a month in revenue selling access to premium

0:06:58.160 --> 0:07:01.400
<v Speaker 2>versions of chat GPT, though I'm not one hundred percent

0:07:01.440 --> 0:07:04.760
<v Speaker 2>sure how Assuming it has twenty million customers paying twenty

0:07:04.800 --> 0:07:07.080
<v Speaker 2>bucks a month, that's four hundred million dollars a month.

0:07:07.160 --> 0:07:09.640
<v Speaker 2>Then five million business customers paying an average of one

0:07:09.720 --> 0:07:12.680
<v Speaker 2>hundred dollars a month each, that's nine hundred million dollars.

0:07:12.960 --> 0:07:15.760
<v Speaker 2>And is that is that really indicative? Is the average

0:07:15.760 --> 0:07:19.000
<v Speaker 2>really that good? Are there that many people paying thirty

0:07:19.000 --> 0:07:21.560
<v Speaker 2>five or fifty or two hundred bucks a month? How

0:07:21.560 --> 0:07:24.760
<v Speaker 2>many of those versus one hundred bucks? I mean, most

0:07:24.800 --> 0:07:27.080
<v Speaker 2>don't pay one hundred bucks. It's either thirty five, fifty

0:07:27.120 --> 0:07:29.720
<v Speaker 2>or two hundred. Open ai doesn't break out the actual

0:07:29.720 --> 0:07:31.960
<v Speaker 2>revenues behind these numbers for a reason. I believe that

0:07:32.240 --> 0:07:34.880
<v Speaker 2>reason is they don't look as good. I also, by

0:07:34.920 --> 0:07:37.480
<v Speaker 2>the way, when I was fucking around trying to play

0:07:37.480 --> 0:07:40.880
<v Speaker 2>with GPT five, I got offered a team subscription for

0:07:40.960 --> 0:07:45.200
<v Speaker 2>a dollar for one month. How how's how many people?

0:07:45.400 --> 0:07:47.680
<v Speaker 2>How many of those business customers got a month for

0:07:47.720 --> 0:07:50.920
<v Speaker 2>a dollar? How long does open ai counter customer? What

0:07:51.080 --> 0:07:54.240
<v Speaker 2>is open AI's churn like? And does it really as

0:07:54.240 --> 0:07:56.000
<v Speaker 2>I wrote in my news there, how much money do

0:07:56.040 --> 0:07:58.560
<v Speaker 2>open AI and Anthropic make in the year making more

0:07:58.600 --> 0:08:01.280
<v Speaker 2>than Spotify at one point five billion dollars a month.

0:08:01.720 --> 0:08:03.960
<v Speaker 2>We don't know an open AI, much like Anthropic, has

0:08:03.960 --> 0:08:07.320
<v Speaker 2>never shared actual revenues other than in June when they

0:08:07.360 --> 0:08:10.680
<v Speaker 2>said ten billion annualized revenue, choosing instead to leak to

0:08:10.680 --> 0:08:12.880
<v Speaker 2>the media and hope to wfiscape the actual amounts of

0:08:12.920 --> 0:08:16.240
<v Speaker 2>money being spent on its services. Anyway, long story short,

0:08:16.280 --> 0:08:18.760
<v Speaker 2>these companies are unprofitable with no end in sight, don't

0:08:18.800 --> 0:08:20.960
<v Speaker 2>even make that much money in most cases, and have

0:08:21.040 --> 0:08:23.560
<v Speaker 2>valued more than anybody would ever buy them for. And

0:08:23.600 --> 0:08:25.400
<v Speaker 2>they don't even have that much in the way of

0:08:25.480 --> 0:08:28.280
<v Speaker 2>valuable IP. And on top of that, the two biggest

0:08:28.280 --> 0:08:31.440
<v Speaker 2>players burn billions of dollars more than they make serving

0:08:31.520 --> 0:08:36.400
<v Speaker 2>these companies that also burn billions of dollars. I wish

0:08:36.480 --> 0:08:39.760
<v Speaker 2>someone would make this make sense because it doesn't. But

0:08:40.080 --> 0:08:42.559
<v Speaker 2>ed the government will give them more money forever, They'll

0:08:42.559 --> 0:08:45.400
<v Speaker 2>bail them out. I'm tired of this fucking point. I'm

0:08:45.440 --> 0:08:47.640
<v Speaker 2>so tired of hearing about bailouce. I'm so tired of

0:08:47.679 --> 0:08:50.440
<v Speaker 2>people saying they'll bail them out. I get that you're scared,

0:08:50.480 --> 0:08:53.080
<v Speaker 2>I get that things are grim, things are absolutely fucking grim,

0:08:53.240 --> 0:08:56.320
<v Speaker 2>But engage with reality here. You can't bail this out.

0:08:56.720 --> 0:08:59.240
<v Speaker 2>You can't bail it out even if this were going

0:08:59.320 --> 0:09:02.280
<v Speaker 2>to happen, will not. Who would they give the money

0:09:02.320 --> 0:09:02.840
<v Speaker 2>to and.

0:09:02.760 --> 0:09:05.800
<v Speaker 3>For how long would they give it to? All AI startups?

0:09:06.200 --> 0:09:08.880
<v Speaker 2>Is every startup going to get a paycheck protection program

0:09:08.920 --> 0:09:11.839
<v Speaker 2>for generative AI? How would that play out in rural,

0:09:11.880 --> 0:09:14.400
<v Speaker 2>red districts where big tech has never been popular, which

0:09:14.440 --> 0:09:16.720
<v Speaker 2>are being hit with both massive cuts to welfare as

0:09:16.760 --> 0:09:18.520
<v Speaker 2>well as the shock waves of betrayed war that has

0:09:18.520 --> 0:09:22.240
<v Speaker 2>made American agricultural exports like feedstocks, which previously went to

0:09:22.280 --> 0:09:26.840
<v Speaker 2>China by shiploads, less appealing worldwide. So they just, in

0:09:26.880 --> 0:09:29.160
<v Speaker 2>this scenario, by the way, bail out open AI then

0:09:29.360 --> 0:09:31.679
<v Speaker 2>stuff it full of government contracts to the tune of

0:09:32.120 --> 0:09:35.720
<v Speaker 2>what fifteen billion dollars a year? Right just in this example,

0:09:36.280 --> 0:09:38.080
<v Speaker 2>just to be clear, that's the low end of what

0:09:38.120 --> 0:09:39.679
<v Speaker 2>this would take to do. And they'll have to keep

0:09:39.720 --> 0:09:42.040
<v Speaker 2>doing it forever until Sam Wultman can build enough data

0:09:42.080 --> 0:09:45.720
<v Speaker 2>centers to what keep burning billions of dollars. There's no

0:09:45.880 --> 0:09:49.240
<v Speaker 2>plan to make this profitable. Say this happens, it won't

0:09:49.240 --> 0:09:54.000
<v Speaker 2>be Sam, Now what America has a bullshit generative AI

0:09:54.080 --> 0:09:56.920
<v Speaker 2>company attached to it, attached to the state that doesn't

0:09:56.920 --> 0:09:59.280
<v Speaker 2>really innovate. It doesn't really matter in any meaningful way

0:09:59.320 --> 0:10:01.840
<v Speaker 2>except that it owns a bunch of data centers, which

0:10:01.840 --> 0:10:04.520
<v Speaker 2>it doesn't yet. By the way, Microsoft owns all their infrastructure,

0:10:04.520 --> 0:10:07.080
<v Speaker 2>the core stuff is barely happening. Fucking hell, I just

0:10:07.120 --> 0:10:09.400
<v Speaker 2>don't think this happens. I think it's a silly idea,

0:10:09.440 --> 0:10:11.720
<v Speaker 2>and the most likely situation would be that Microsoft would

0:10:11.800 --> 0:10:16.080
<v Speaker 2>unhinge its daw and swallow open AI in its customer's whole. Hey, hey, hey,

0:10:16.120 --> 0:10:18.560
<v Speaker 2>did you know, by the way, the Microsoft's data center

0:10:18.600 --> 0:10:20.880
<v Speaker 2>construction is down year every year, and it's basically signed

0:10:20.920 --> 0:10:22.200
<v Speaker 2>no new data center leases.

0:10:22.520 --> 0:10:24.000
<v Speaker 3>I wonder why it isn't building these new.

0:10:23.960 --> 0:10:26.920
<v Speaker 2>Data centers or open AI could be anything. By the way,

0:10:26.960 --> 0:10:30.480
<v Speaker 2>Stargate isn't saving them either. As I've written recently, Stargate

0:10:30.480 --> 0:10:32.800
<v Speaker 2>doesn't actually exist beyond the media hype it generated, and

0:10:32.880 --> 0:10:37.199
<v Speaker 2>every single person who wrote big articles about Stargate being

0:10:37.280 --> 0:10:40.800
<v Speaker 2>real and Stargate being connected to Abilene Texas should be

0:10:40.800 --> 0:10:44.920
<v Speaker 2>fucking ashamed of themselves. It's deeply unprofessional anyway. That also

0:10:45.040 --> 0:10:49.920
<v Speaker 2>counts the hint of the involvement. There are several articles

0:10:49.920 --> 0:10:53.920
<v Speaker 2>that hint without directly saying that Abilene Texas was part

0:10:53.960 --> 0:10:57.120
<v Speaker 2>of the soft bank stargate thing. You can't wriggle out

0:10:57.120 --> 0:10:59.560
<v Speaker 2>of this one when the time comes or will be

0:10:59.600 --> 0:11:01.679
<v Speaker 2>ad by which I mean, I'm going to be pointing

0:11:01.679 --> 0:11:05.079
<v Speaker 2>at the articles aggressively. We have other stuff to do anyway.

0:11:05.160 --> 0:11:07.240
<v Speaker 2>By the way, does the government do this for everybody?

0:11:07.240 --> 0:11:10.280
<v Speaker 2>This very important question is everyone getting bailed out? Is

0:11:10.320 --> 0:11:13.360
<v Speaker 2>it just anthropical out and open AI? What about Google

0:11:13.440 --> 0:11:15.320
<v Speaker 2>and Microsoft? They're going to be burning and come on

0:11:15.440 --> 0:11:17.040
<v Speaker 2>a bunch of money. Why do they get the money

0:11:17.080 --> 0:11:19.080
<v Speaker 2>no one else? While their industries are going to turn

0:11:19.120 --> 0:11:20.840
<v Speaker 2>around and go wait wait, wait, wait, wait wait. Why

0:11:20.840 --> 0:11:23.240
<v Speaker 2>do I have to fucking make money when what's the

0:11:23.240 --> 0:11:25.760
<v Speaker 2>point is? Because AI is important, You're going to get

0:11:25.760 --> 0:11:27.920
<v Speaker 2>a bunch of fraud if this happens, it's not going

0:11:27.960 --> 0:11:30.840
<v Speaker 2>to Just to be clear, and what is the limit?

0:11:30.880 --> 0:11:33.959
<v Speaker 2>By the way, do they subsidize or compute for companies

0:11:34.000 --> 0:11:36.760
<v Speaker 2>like Cursor to what end? Where's the limit? How does

0:11:36.800 --> 0:11:38.640
<v Speaker 2>it end? What is the bailout? What are you bailing out?

0:11:38.640 --> 0:11:41.400
<v Speaker 2>Where does the bailout go? These are all questions that

0:11:41.400 --> 0:11:46.080
<v Speaker 2>would get answered. You can. You can sit there and say, oh,

0:11:46.200 --> 0:11:48.560
<v Speaker 2>the government's corrupt. Oh I feel this way and that way.

0:11:48.800 --> 0:11:51.959
<v Speaker 2>Here's the thing top, which is the thing that was

0:11:52.040 --> 0:11:54.199
<v Speaker 2>used to bail out banks and such and hedge ones,

0:11:54.200 --> 0:11:57.200
<v Speaker 2>and during that whole situation at least made more sense

0:11:57.240 --> 0:12:00.160
<v Speaker 2>because if a bank fails, people lose actual money. This

0:12:00.160 --> 0:12:02.680
<v Speaker 2>stuff fails, what do people lose? What is the service

0:12:02.679 --> 0:12:04.600
<v Speaker 2>that people lose access to? Chat GPT?

0:12:05.480 --> 0:12:06.000
<v Speaker 3>That's it.

0:12:06.520 --> 0:12:08.920
<v Speaker 2>I don't even think that Donald Trump could explain what

0:12:09.040 --> 0:12:12.240
<v Speaker 2>chat GPT is. And even then, he al really has

0:12:12.280 --> 0:12:14.200
<v Speaker 2>people to give him money, like these people don't have

0:12:14.240 --> 0:12:16.560
<v Speaker 2>anything to offer other than just this vague sense of

0:12:16.640 --> 0:12:20.240
<v Speaker 2>AI is important. I just don't buy it. And I

0:12:20.280 --> 0:12:23.120
<v Speaker 2>think that people if you're worried to get hopeful.

0:12:23.400 --> 0:12:23.760
<v Speaker 3>I don't know.

0:12:24.160 --> 0:12:27.080
<v Speaker 2>I'm hopeful that something's going to explode within this terrible

0:12:27.160 --> 0:12:31.160
<v Speaker 2>fucking industry. And I'm kind of hopeful because I think

0:12:31.200 --> 0:12:34.439
<v Speaker 2>it's time to ask a basic question, which is what

0:12:34.520 --> 0:12:38.080
<v Speaker 2>if generative AI just is not profitable? And I think

0:12:38.120 --> 0:12:40.160
<v Speaker 2>this is a question that we have to seriously consider

0:12:40.160 --> 0:12:43.760
<v Speaker 2>at this point because it's ramifications are significant. If I'm honest,

0:12:43.800 --> 0:12:45.760
<v Speaker 2>I think the future of large language models will be

0:12:45.840 --> 0:12:50.160
<v Speaker 2>largely client side on egregiously expensive personal setups or enthusiasts,

0:12:50.200 --> 0:12:53.440
<v Speaker 2>and in a handful of niche enterprise roles. Large language

0:12:53.440 --> 0:12:56.360
<v Speaker 2>models do not scale profitably, and their functionality is not

0:12:56.400 --> 0:12:59.040
<v Speaker 2>significant enough to justify the costs of running them. By

0:12:59.080 --> 0:13:01.800
<v Speaker 2>immediately applying old economics the idea that you would pay

0:13:01.840 --> 0:13:05.160
<v Speaker 2>a monthly fee to have relatively unlimited access, companies like

0:13:05.200 --> 0:13:07.840
<v Speaker 2>open ai and Anthropic immediately trained users to use their

0:13:07.880 --> 0:13:10.280
<v Speaker 2>products in a way that was antithetical to their costs.

0:13:10.840 --> 0:13:13.360
<v Speaker 2>Then again, had these models been served in that way,

0:13:13.720 --> 0:13:15.560
<v Speaker 2>had they been mindful of the cost and charged what

0:13:15.600 --> 0:13:17.840
<v Speaker 2>they actually were, there would likely have been no way

0:13:17.880 --> 0:13:20.240
<v Speaker 2>to get this far. If open ai is making billions

0:13:20.280 --> 0:13:22.600
<v Speaker 2>of dollars a month, or a billion dollars a month,

0:13:22.679 --> 0:13:25.040
<v Speaker 2>or whatever the fuck it is, it's possibly losing that

0:13:25.160 --> 0:13:27.680
<v Speaker 2>much or more after revenue, and that's the money it

0:13:27.679 --> 0:13:29.640
<v Speaker 2>can get selling the product in a form that can

0:13:29.679 --> 0:13:33.080
<v Speaker 2>never turn profitable itself. If open ai charged in line

0:13:33.080 --> 0:13:35.960
<v Speaker 2>with its actual costs, would even be able to justify

0:13:36.000 --> 0:13:38.720
<v Speaker 2>a free version of chat GPT outside of a few

0:13:38.760 --> 0:13:41.800
<v Speaker 2>requests a month. The revenue you see today is what

0:13:41.800 --> 0:13:44.160
<v Speaker 2>people are willing to pay for a product that loses money,

0:13:44.280 --> 0:13:46.400
<v Speaker 2>and I cannot imagine they would pay as much as

0:13:46.400 --> 0:13:49.520
<v Speaker 2>companies in question charge their costs. If I'm wrong, Curse

0:13:49.640 --> 0:13:51.760
<v Speaker 2>will be just fine. And that's assuming that Cursor's current

0:13:51.800 --> 0:13:54.520
<v Speaker 2>hobbled form is even profitable, which it has not said

0:13:54.559 --> 0:13:57.280
<v Speaker 2>it is. So you've got an entire industry of companies

0:13:57.320 --> 0:13:59.199
<v Speaker 2>that struggle to do anything other than lose a lot

0:13:59.200 --> 0:14:02.400
<v Speaker 2>of money. Great, and now we have a massive expansion

0:14:02.440 --> 0:14:04.760
<v Speaker 2>of data centers the likes of which we've never seen

0:14:04.960 --> 0:14:07.360
<v Speaker 2>or to capture demand for a product that nobody makes

0:14:07.440 --> 0:14:10.480
<v Speaker 2>much money selling. This naturally leads to an important question,

0:14:10.920 --> 0:14:13.960
<v Speaker 2>how do these people building data centers actually make money.

0:14:24.640 --> 0:14:26.760
<v Speaker 2>At the start of August, the Wall Street Journal published

0:14:26.760 --> 0:14:28.360
<v Speaker 2>one of the most worrying facts I've seen in the

0:14:28.440 --> 0:14:31.800
<v Speaker 2>last two years. The Journal quoted investor Paul Kodrowski is

0:14:31.840 --> 0:14:34.520
<v Speaker 2>saying that, and I quote as a percentage of gross

0:14:34.520 --> 0:14:37.840
<v Speaker 2>domestic products, spending on AI infrastructure has already exceeded spending

0:14:37.840 --> 0:14:40.560
<v Speaker 2>on telecom and internet infrastructure from the dot com boom,

0:14:40.600 --> 0:14:44.720
<v Speaker 2>and it's still growing. Kudrowski described this insane capex spending

0:14:44.760 --> 0:14:48.160
<v Speaker 2>as a sort of private sector stimulus program. The Journal

0:14:48.200 --> 0:14:51.560
<v Speaker 2>also quotes analyst Neil Dhuta as saying that CAPEX spending

0:14:51.600 --> 0:14:54.280
<v Speaker 2>for AI contributed more to growth in the US economy

0:14:54.320 --> 0:14:57.800
<v Speaker 2>in the past two quarters than all consumers spending. The

0:14:57.880 --> 0:15:00.440
<v Speaker 2>massive build out of data centers and the associate sheated

0:15:00.440 --> 0:15:03.160
<v Speaker 2>physical gear like chips, service some raw materials to building

0:15:03.160 --> 0:15:06.240
<v Speaker 2>them has become a massive dominant economic force, building capacity

0:15:06.280 --> 0:15:07.440
<v Speaker 2>for an industry that is yet.

0:15:07.280 --> 0:15:08.720
<v Speaker 3>To prove it can make real money.

0:15:09.200 --> 0:15:11.920
<v Speaker 2>And no Microsoft talking about it's a zero revenue it's

0:15:12.000 --> 0:15:13.880
<v Speaker 2>last quarterly earnings for the first time is not the

0:15:13.920 --> 0:15:17.200
<v Speaker 2>same thing as it stopped explicitly stating its AI revenue

0:15:17.240 --> 0:15:19.760
<v Speaker 2>in January when it was thirteen billion dollars annualized, so

0:15:19.800 --> 0:15:23.440
<v Speaker 2>about a billion a month. And by the way, that's revenue,

0:15:23.480 --> 0:15:27.280
<v Speaker 2>not profit. Just to remind you, I need to every

0:15:27.320 --> 0:15:31.160
<v Speaker 2>time remind anyone who ever mentions AI revenue that that's

0:15:31.280 --> 0:15:36.080
<v Speaker 2>not profit anyway. Aikapex allegedly, though I've heard some questions

0:15:36.120 --> 0:15:38.480
<v Speaker 2>about this figure, accounts to one point two percent of

0:15:38.480 --> 0:15:40.480
<v Speaker 2>the US GDP in the first half of the year,

0:15:40.640 --> 0:15:42.720
<v Speaker 2>and it accounted for more than half of the to

0:15:42.840 --> 0:15:45.400
<v Speaker 2>quote the Wall Street Journal, already sluggish one point two

0:15:45.440 --> 0:15:47.920
<v Speaker 2>percent growth of the US economy in the first half

0:15:47.960 --> 0:15:50.720
<v Speaker 2>of the year. Another Wall Street Journal piece published a

0:15:50.760 --> 0:15:53.400
<v Speaker 2>few days later discussed how data center development is souring

0:15:53.440 --> 0:15:55.800
<v Speaker 2>the free cash flow for big tech, turning them from

0:15:55.800 --> 0:15:58.440
<v Speaker 2>the kind of asset light businesses that the markets love

0:15:58.600 --> 0:16:01.520
<v Speaker 2>into entities burdened by physics, real estate and their associated

0:16:01.520 --> 0:16:05.040
<v Speaker 2>costs and I quote. For years, investors loved those companies

0:16:05.080 --> 0:16:07.280
<v Speaker 2>because they were asset lighting. They earned their profits or

0:16:07.320 --> 0:16:10.800
<v Speaker 2>on intangible assets such as intellectual property, software, and digital

0:16:10.800 --> 0:16:13.640
<v Speaker 2>platforms with network effects. This can be seen as a

0:16:13.640 --> 0:16:16.320
<v Speaker 2>metric called free cash flow, roughly defined as cash flow

0:16:16.360 --> 0:16:20.720
<v Speaker 2>from operations minus capital expenditures. This is arguably the purest

0:16:20.760 --> 0:16:25.200
<v Speaker 2>measure of business's underlying cash generating potential. From twenty sixteen

0:16:25.240 --> 0:16:27.800
<v Speaker 2>through twenty twenty three, free cash flow and net earnings

0:16:27.800 --> 0:16:32.280
<v Speaker 2>of Alphabet, Amazon, Meta, and Microsoft grew roughly in tandem,

0:16:32.400 --> 0:16:34.760
<v Speaker 2>but since twenty twenty three the two have diverged. The

0:16:34.800 --> 0:16:37.040
<v Speaker 2>four companies can by net income is up seventy three

0:16:37.080 --> 0:16:39.400
<v Speaker 2>percent to ninety one billion in the second quarter from

0:16:39.400 --> 0:16:42.080
<v Speaker 2>two years earlier, while free cash flow is down thirty

0:16:42.080 --> 0:16:44.880
<v Speaker 2>percent to forty billion. According to fact Sheet Data, Apple

0:16:45.000 --> 0:16:48.920
<v Speaker 2>a relative pika on capital spending has also seen free

0:16:48.960 --> 0:16:52.560
<v Speaker 2>cash flow lag behind. These numbers are all very scary,

0:16:52.600 --> 0:16:54.680
<v Speaker 2>and I mean that sincerely, but they also failed to

0:16:54.720 --> 0:16:58.600
<v Speaker 2>explain why how much was actually spent on AI capex

0:16:58.640 --> 0:17:00.840
<v Speaker 2>in the US. One would think two different articles on

0:17:00.880 --> 0:17:03.560
<v Speaker 2>the subject would include that number versus a single quarter's worth.

0:17:03.720 --> 0:17:07.000
<v Speaker 2>But from my estimates, I expect capital expenditures from the

0:17:07.080 --> 0:17:09.600
<v Speaker 2>Magnificent seven alone to crest two hundred billion dollars in

0:17:09.640 --> 0:17:12.440
<v Speaker 2>the first half of twenty twenty five, with Axios estimating

0:17:12.480 --> 0:17:15.879
<v Speaker 2>they'd spend around four hundred billion total this year. Most

0:17:15.960 --> 0:17:18.160
<v Speaker 2>articles are drafting off of a blog from Paul Kodrowski,

0:17:18.200 --> 0:17:20.560
<v Speaker 2>who estimates that total AI capex would be somewhere in

0:17:20.560 --> 0:17:22.560
<v Speaker 2>the region of five hundred and twenty billion dollars for

0:17:22.600 --> 0:17:25.600
<v Speaker 2>the total in the year, which felt conservative for me,

0:17:25.680 --> 0:17:28.760
<v Speaker 2>so I did the smart thing and asked him. Kudrowski

0:17:28.880 --> 0:17:31.000
<v Speaker 2>noted that these numbers focus entirely on the four big

0:17:31.040 --> 0:17:34.119
<v Speaker 2>spenders Microsoft, Google, Meta, and Amazon, and his own estimated

0:17:34.160 --> 0:17:36.520
<v Speaker 2>three hundred and twelve billion dollar capex and one point

0:17:36.560 --> 0:17:38.920
<v Speaker 2>two percent number came from the assumption that the US

0:17:39.000 --> 0:17:41.240
<v Speaker 2>GDP in twenty twenty five will be around twenty eight

0:17:41.240 --> 0:17:44.480
<v Speaker 2>trillion dollars, which is lower than other forecasts which put

0:17:44.520 --> 0:17:48.520
<v Speaker 2>it around thirty trillion. Even Kujoski, in his own words,

0:17:48.560 --> 0:17:50.760
<v Speaker 2>was trying to be conservative using public data and then

0:17:50.760 --> 0:17:54.040
<v Speaker 2>building his analysis from there. I personally believe his estimate

0:17:54.080 --> 0:17:56.040
<v Speaker 2>is too conservative because it does in factor in the

0:17:56.080 --> 0:17:58.840
<v Speaker 2>capital expenditures Maracle, which along with Crusoe is building the

0:17:58.960 --> 0:18:02.200
<v Speaker 2>vast abilenei Exis data center for Open AI or any

0:18:02.240 --> 0:18:05.200
<v Speaker 2>other private data center developers thinking cash into ar capex.

0:18:05.600 --> 0:18:08.280
<v Speaker 2>When I asked him to elaborate, he estimated that AI

0:18:08.400 --> 0:18:12.359
<v Speaker 2>spending all in was around half of three percent Q

0:18:12.520 --> 0:18:15.640
<v Speaker 2>two real GDP growth. He added that it could be

0:18:15.880 --> 0:18:19.919
<v Speaker 2>half of US GDP four year GDP growth if.

0:18:19.760 --> 0:18:22.200
<v Speaker 3>Certain conditions are out. That's so cool.

0:18:22.720 --> 0:18:25.120
<v Speaker 2>Half of the US economy's growth came from building data

0:18:25.160 --> 0:18:27.399
<v Speaker 2>centers for generative AI, which is the combined revenue of

0:18:27.400 --> 0:18:29.600
<v Speaker 2>a little more than the fucking smartwatch industry did in

0:18:29.640 --> 0:18:32.680
<v Speaker 2>twenty twenty four. Another troubling point is that big tech

0:18:32.720 --> 0:18:34.879
<v Speaker 2>doesn't just buy data centers and uses them, but in

0:18:34.880 --> 0:18:37.400
<v Speaker 2>many cases, pace for a construction company to build them,

0:18:37.560 --> 0:18:39.679
<v Speaker 2>fills them full of GPUs, then leases them from a

0:18:39.720 --> 0:18:42.679
<v Speaker 2>company that runs them, meaning that they themselves don't have

0:18:42.720 --> 0:18:45.239
<v Speaker 2>to personally staff up and maintain them. This creates an

0:18:45.240 --> 0:18:47.679
<v Speaker 2>economic boom for construction companies in the short term, as

0:18:47.720 --> 0:18:51.360
<v Speaker 2>well as lucrative contracts for ongoing support as long as

0:18:51.400 --> 0:18:54.560
<v Speaker 2>the company in question still wants them. While Microsoft or

0:18:54.560 --> 0:18:57.399
<v Speaker 2>Amazon might use a data center and indeed act as

0:18:57.440 --> 0:18:59.800
<v Speaker 2>if they own it, ultimately somebody else is holding the

0:18:59.800 --> 0:19:03.719
<v Speaker 2>bag and the ultimate responsibility for said data center. One

0:19:03.800 --> 0:19:06.480
<v Speaker 2>such company is QTS, the data center developer that leases

0:19:06.480 --> 0:19:08.560
<v Speaker 2>both to both Amazon and Meta according to The New

0:19:08.640 --> 0:19:11.560
<v Speaker 2>York Times, which was acquired by Blackstone in twenty twenty

0:19:11.600 --> 0:19:14.439
<v Speaker 2>one for ten billion dollars. Since then, Blackstone has used

0:19:14.480 --> 0:19:18.040
<v Speaker 2>commercially mortgage backed securities I know it's so great to

0:19:18.119 --> 0:19:21.320
<v Speaker 2>raise over eight point seven billion dollars since since then

0:19:21.359 --> 0:19:24.240
<v Speaker 2>to sink into qts's expansion, and as of mid July

0:19:24.359 --> 0:19:27.080
<v Speaker 2>said it'd be investing twenty five billion dollars in AI

0:19:27.160 --> 0:19:30.600
<v Speaker 2>data centers and energy. Blackstone, according to The New York Times,

0:19:30.600 --> 0:19:33.320
<v Speaker 2>see strong demand from tech companies who apparently are willing

0:19:33.359 --> 0:19:36.080
<v Speaker 2>to sign what they describe as airtight leases for fifteen

0:19:36.080 --> 0:19:38.640
<v Speaker 2>to twenty years to rent out data center space. As

0:19:38.640 --> 0:19:42.080
<v Speaker 2>a reminder that Microsoft has over one thousand lawyers in

0:19:42.119 --> 0:19:45.439
<v Speaker 2>the company anyway. Yet the Times also names another problem,

0:19:45.480 --> 0:19:48.560
<v Speaker 2>the unanswered question of how these private equity firms actually

0:19:48.600 --> 0:19:52.480
<v Speaker 2>exit these situations. Blackstone, KKR and other asset management firms

0:19:52.480 --> 0:19:55.119
<v Speaker 2>do not buy companies with the intention of siphoning off revenue,

0:19:55.160 --> 0:19:57.560
<v Speaker 2>but to pump them and sell them to another company

0:19:57.680 --> 0:20:00.280
<v Speaker 2>or dump them onto the public markets. Of course, much

0:20:00.320 --> 0:20:03.000
<v Speaker 2>like AI startups, it isn't obvious who would buy qts,

0:20:03.000 --> 0:20:04.600
<v Speaker 2>and what I imagine would be a twenty five or

0:20:04.640 --> 0:20:08.000
<v Speaker 2>thirty billion dollar valuation mean that black Stone would, as

0:20:08.000 --> 0:20:11.040
<v Speaker 2>I've mentioned, have to take them public. Similarly, KKR is

0:20:11.040 --> 0:20:14.359
<v Speaker 2>supposed fifteen billion dollar partnership with investment firm Energy Capital

0:20:14.400 --> 0:20:17.240
<v Speaker 2>Partners to build data centers and their associated utilities does

0:20:17.240 --> 0:20:19.280
<v Speaker 2>not appear to have much of an exit plan either.

0:20:20.040 --> 0:20:23.040
<v Speaker 2>And let's not forget Oracle Open AI and Cruso's abominable

0:20:23.080 --> 0:20:25.800
<v Speaker 2>mess in Abilene, Texas, where Oracle is paying for forty

0:20:25.800 --> 0:20:29.359
<v Speaker 2>billion dollars worth of GPUs and Crusoe is spending fifteen

0:20:29.359 --> 0:20:32.159
<v Speaker 2>dollars billion dollars raised from Blue Man, Owl Capital and

0:20:32.200 --> 0:20:35.240
<v Speaker 2>primary digital infrastructure. Though I think JP Morgan is involved

0:20:35.240 --> 0:20:37.719
<v Speaker 2>too to build data centers for open Ai, a company

0:20:37.760 --> 0:20:39.359
<v Speaker 2>that loses billion dollars.

0:20:39.400 --> 0:20:41.320
<v Speaker 3>If yeah, why would they do that?

0:20:41.520 --> 0:20:44.479
<v Speaker 2>Right? Why why would they do that? Well, it's simple,

0:20:44.520 --> 0:20:47.840
<v Speaker 2>you roube, you moron, you buffoon. It's so that open

0:20:47.840 --> 0:20:50.439
<v Speaker 2>Ai can allegedly, starting in twenty twenty eight, pay Oracle

0:20:50.520 --> 0:20:53.320
<v Speaker 2>thirty billion dollars a year for compute. And I am

0:20:53.359 --> 0:20:57.280
<v Speaker 2>being fully serious to be clear, open Ai, by my estimates,

0:20:57.280 --> 0:21:00.000
<v Speaker 2>has made only five point twenty six billion dollars this year,

0:21:00.480 --> 0:21:03.200
<v Speaker 2>and we'll have trouble meeting their twelve point seven billion

0:21:03.240 --> 0:21:05.800
<v Speaker 2>dollar revenue projection for twenty twenty five and will likely

0:21:05.840 --> 0:21:09.240
<v Speaker 2>lose more than ten billion dollars doing so. Oracle will,

0:21:09.280 --> 0:21:12.560
<v Speaker 2>according to the information, oh Crusoe one billion dollars in

0:21:12.600 --> 0:21:15.399
<v Speaker 2>payments across a fifteen year span of release. How does

0:21:15.440 --> 0:21:17.960
<v Speaker 2>Crusoe afford to pay back it's fifteen billion dollars in lungs?

0:21:18.160 --> 0:21:18.560
<v Speaker 3>Beats me.

0:21:19.080 --> 0:21:21.560
<v Speaker 2>The information says that Crusoe is raising a billion dollars

0:21:21.600 --> 0:21:24.199
<v Speaker 2>to take on the cloud giants by earning construction management

0:21:24.240 --> 0:21:26.320
<v Speaker 2>fees and rent and you can sell it's stake in

0:21:26.359 --> 0:21:29.480
<v Speaker 2>the project upon reaching certain completion milestones, while also building

0:21:29.520 --> 0:21:32.520
<v Speaker 2>its own Ai Compute, making the assumption that the demand

0:21:32.560 --> 0:21:35.960
<v Speaker 2>is there outside of hyperscalers. Then there's corp Eve, my

0:21:36.080 --> 0:21:38.960
<v Speaker 2>least favor company in the world. As I discussed a

0:21:38.960 --> 0:21:41.880
<v Speaker 2>few months ago, core Wave is burned by obscene debt

0:21:41.920 --> 0:21:44.439
<v Speaker 2>and a horrifying cash burn and has seen its stock spikee

0:21:44.600 --> 0:21:46.159
<v Speaker 2>to a high of one hundred and eighty three on

0:21:46.280 --> 0:21:48.800
<v Speaker 2>June twentieth, twenty twenty five, which has led to an

0:21:48.800 --> 0:21:51.520
<v Speaker 2>all stock attempt to acquire develop a core scientific for

0:21:51.640 --> 0:21:54.880
<v Speaker 2>nine billion dollars. And they're doing that with all stock,

0:21:54.960 --> 0:21:57.000
<v Speaker 2>and now that's falling apart because core Wave, as my

0:21:57.080 --> 0:21:58.680
<v Speaker 2>last check, and who knows where it will be by

0:21:58.680 --> 0:22:02.080
<v Speaker 2>the time this episode runs, is below one hundred bucks.

0:22:02.080 --> 0:22:04.520
<v Speaker 2>Because the IPO lockiles has gone.

0:22:05.080 --> 0:22:05.800
<v Speaker 3>All the people that.

0:22:05.800 --> 0:22:08.280
<v Speaker 2>Invested seem to be getting out. We'll see where they

0:22:08.320 --> 0:22:11.280
<v Speaker 2>go from here. Cor we've by the way, has since

0:22:11.320 --> 0:22:14.080
<v Speaker 2>gone public and had to borrow billions of dollars to

0:22:14.119 --> 0:22:17.399
<v Speaker 2>fund their obscene capital expenditures. And they are also by

0:22:17.440 --> 0:22:19.399
<v Speaker 2>the way, going to have to handle in October at

0:22:19.400 --> 0:22:22.240
<v Speaker 2>the beginning of their DDTL two point zero, their biggest loan,

0:22:22.840 --> 0:22:25.080
<v Speaker 2>and that's also when open Ai is allegedly going to

0:22:25.119 --> 0:22:29.040
<v Speaker 2>start paying them for their compute and by the way,

0:22:28.840 --> 0:22:32.080
<v Speaker 2>they lose hundreds of millions. They lose so much money

0:22:32.080 --> 0:22:36.560
<v Speaker 2>every quarter. They have no part to profitability. Ah, every

0:22:36.600 --> 0:22:39.440
<v Speaker 2>time I think about core Weave feel crazy and to

0:22:39.560 --> 0:22:41.360
<v Speaker 2>be honest, I'm gonna be honest when to put out

0:22:41.359 --> 0:22:44.400
<v Speaker 2>that blog, and people sent me some very unfair emails

0:22:44.560 --> 0:22:47.040
<v Speaker 2>about that. They said, there's no way that corl We've

0:22:47.040 --> 0:22:49.440
<v Speaker 2>can run out of money. If Corwe've runs out of money,

0:22:49.440 --> 0:22:51.240
<v Speaker 2>by the way, and they collapse, which I think is

0:22:51.320 --> 0:22:54.520
<v Speaker 2>very possible. I'm taking some fucking names on this one.

0:22:54.520 --> 0:22:56.840
<v Speaker 2>A lot of people very unfair to miss. The ztrum

0:22:57.240 --> 0:22:59.800
<v Speaker 2>corwave I had is also pretty much reliant on Microsoft

0:22:59.800 --> 0:23:02.760
<v Speaker 2>as one of its primary customers. While this relationship has

0:23:02.760 --> 0:23:05.840
<v Speaker 2>been fairly smooth so far as far as we know,

0:23:06.240 --> 0:23:08.640
<v Speaker 2>this dependence also presents the next eistential threat to core

0:23:08.680 --> 0:23:10.040
<v Speaker 2>we even is part of the reason why I'm so

0:23:10.160 --> 0:23:14.119
<v Speaker 2>pessimistic about their survival. Microsoft has its own infrastructure and

0:23:14.119 --> 0:23:16.520
<v Speaker 2>has every incentive to cut out middlemen when it's able

0:23:16.560 --> 0:23:19.440
<v Speaker 2>to meet supply with the demand itsself owns or lease

0:23:19.480 --> 0:23:22.760
<v Speaker 2>it's rather than subcontracts out simply because middlemen add costs

0:23:22.760 --> 0:23:23.720
<v Speaker 2>and shrink margins.

0:23:23.920 --> 0:23:25.480
<v Speaker 3>If Microsoft walks, what's left?

0:23:25.600 --> 0:23:29.640
<v Speaker 2>How does it service its its ongoing obligations and the debt.

0:23:29.960 --> 0:23:32.200
<v Speaker 2>In all of these cases, data center developers seem to

0:23:32.240 --> 0:23:35.040
<v Speaker 2>have very few options as to making actual money. We

0:23:35.119 --> 0:23:37.760
<v Speaker 2>have companies spending billions of dollars to vastly expend their

0:23:37.840 --> 0:23:40.199
<v Speaker 2>data center footprint, but very little evidence that doing so

0:23:40.320 --> 0:23:43.280
<v Speaker 2>results in revenue, let alone some sort of payoff event.

0:23:43.600 --> 0:23:46.560
<v Speaker 2>And similarly, the actual capital expenditures they're making are much

0:23:46.560 --> 0:23:50.159
<v Speaker 2>smaller than those of big tech. Digital Realty Trust, one

0:23:50.200 --> 0:23:52.760
<v Speaker 2>of the largest developers with over three hundred data centers

0:23:52.800 --> 0:23:55.280
<v Speaker 2>worldwide and five point five billion dollars in revenue in

0:23:55.280 --> 0:23:58.080
<v Speaker 2>twenty twenty four, only spend three point five billion dollars

0:23:58.080 --> 0:24:00.960
<v Speaker 2>in Capex last quarter, and Equinix eight point seven billion

0:24:00.960 --> 0:24:03.159
<v Speaker 2>dollars of revenue in twenty twenty four total, which has

0:24:03.200 --> 0:24:06.040
<v Speaker 2>two hundred and seventy data centers, but Capex at three

0:24:06.040 --> 0:24:09.080
<v Speaker 2>point five billion dollars for a quarter. Two NTT Global

0:24:09.160 --> 0:24:11.199
<v Speaker 2>Data Centers, which has over one hundred and sixty of

0:24:11.240 --> 0:24:14.080
<v Speaker 2>the fuckers, has dedicated ten billion dollars in Capex through

0:24:14.119 --> 0:24:16.040
<v Speaker 2>twenty twenty seven to build out more of them.

0:24:16.520 --> 0:24:17.280
<v Speaker 3>Yeah, in many of.

0:24:17.240 --> 0:24:19.520
<v Speaker 2>These cases, it's because these companies are, to quote a

0:24:19.560 --> 0:24:22.960
<v Speaker 2>source of mind, functionally obsolete for this cycle because legacy

0:24:23.000 --> 0:24:25.280
<v Speaker 2>data centers are not plug and play ready for GPUs

0:24:25.320 --> 0:24:28.480
<v Speaker 2>to slot into. And remember GPUs and the way they

0:24:28.480 --> 0:24:31.479
<v Speaker 2>work with GENERATIVEAI is just atypical. They burn them, they

0:24:31.560 --> 0:24:34.760
<v Speaker 2>run them hot constantly, They're running full fucking till the

0:24:34.800 --> 0:24:38.000
<v Speaker 2>whole fucking time. It's absolutely ridiculous, leading to a bunch

0:24:38.000 --> 0:24:41.600
<v Speaker 2>of replacements. But it just to be specific, requires a

0:24:41.640 --> 0:24:45.920
<v Speaker 2>bunch of specialized cooling, much more energy. And yeah, there's

0:24:46.040 --> 0:24:48.080
<v Speaker 2>just it's not a thing where you just take out

0:24:48.119 --> 0:24:50.960
<v Speaker 2>one computer and you put in another. And also any

0:24:51.000 --> 0:24:53.239
<v Speaker 2>investment in CAPEX by these companies would have to be

0:24:53.480 --> 0:24:56.879
<v Speaker 2>for both GPUs and said retrofitting. By the way, massive

0:24:56.880 --> 0:24:59.600
<v Speaker 2>amounts of money. And this means that the money flowing

0:24:59.640 --> 0:25:02.400
<v Speaker 2>into a ID data centers, it's predominantly going to neo

0:25:02.440 --> 0:25:05.360
<v Speaker 2>clouds like Corweven Crusoe, and all seems to flow back

0:25:05.359 --> 0:25:07.399
<v Speaker 2>to private equity firms that never thought about where the

0:25:07.400 --> 0:25:10.560
<v Speaker 2>cash out might be. Blackstone led corby seven point five

0:25:10.600 --> 0:25:13.040
<v Speaker 2>billion dollar a line with Magnetal Capital, who, by the way,

0:25:13.080 --> 0:25:15.680
<v Speaker 2>was involved with a bunch of subprime mortgages, and Crusoe

0:25:15.720 --> 0:25:18.200
<v Speaker 2>signed a deal a week ago with infrastructure firm and

0:25:18.200 --> 0:25:21.160
<v Speaker 2>they're also owned by Blackstone called tall Grass, and they're

0:25:21.160 --> 0:25:23.920
<v Speaker 2>building a data center in Wyoming, all of which seems

0:25:23.960 --> 0:25:25.760
<v Speaker 2>very good for Blackstone, unless you think, how does it

0:25:25.800 --> 0:25:28.919
<v Speaker 2>actually make money here? As private equity firms generally not

0:25:29.080 --> 0:25:32.080
<v Speaker 2>like to hold assets longer than five years, and the

0:25:32.119 --> 0:25:35.840
<v Speaker 2>timer is running. Even if it did, it's capital expenditures

0:25:35.840 --> 0:25:37.879
<v Speaker 2>that are dropping the bucket in the grand scheme of things.

0:25:38.040 --> 0:25:40.960
<v Speaker 2>Assuming Crusoe burns as the information suggests, it will as

0:25:41.000 --> 0:25:43.760
<v Speaker 2>much as four billion dollars in twenty twenty five. Corweave

0:25:43.800 --> 0:25:46.560
<v Speaker 2>spends as much as twenty billion dollars. Digital Realty Crust

0:25:46.600 --> 0:25:49.679
<v Speaker 2>spends as much as fourteen billion dollars. Global data centers

0:25:49.720 --> 0:25:52.600
<v Speaker 2>three point three three billion, that's ten billion over three years.

0:25:52.760 --> 0:25:56.400
<v Speaker 2>Equiniz spends fourteen billion. That's about fifty five point three

0:25:56.480 --> 0:25:58.880
<v Speaker 2>three billion dollars in AI CAPEK spent in twenty twenty

0:25:58.880 --> 0:26:01.639
<v Speaker 2>five from the largest developer of data centers in the world.

0:26:01.840 --> 0:26:04.119
<v Speaker 2>For some context, one hundred and two billion dollars were

0:26:04.160 --> 0:26:07.439
<v Speaker 2>spent by Meta Alphabet, Microsoft and Amazon in the last quarter.

0:26:09.000 --> 0:26:11.720
<v Speaker 2>Private equity may face the same problem as AI startups.

0:26:11.720 --> 0:26:14.680
<v Speaker 2>In the end, though, there is no real exit strategy

0:26:14.720 --> 0:26:17.800
<v Speaker 2>for any of these investments. Furthermore, the supposed AI capex

0:26:17.840 --> 0:26:20.320
<v Speaker 2>boom that's driving the US economy is not, as reported,

0:26:20.600 --> 0:26:24.439
<v Speaker 2>driven by the massive interest in driving AI infrastructure up

0:26:24.440 --> 0:26:25.720
<v Speaker 2>for a variety of customers.

0:26:25.960 --> 0:26:29.280
<v Speaker 3>The reality is simple. The majority of AI capex.

0:26:28.920 --> 0:26:31.440
<v Speaker 2>Is from big tech, which is a massive systemic weakness

0:26:31.440 --> 0:26:49.280
<v Speaker 2>for our economy. Now, while some might say that AI

0:26:49.359 --> 0:26:51.919
<v Speaker 2>capex has swallowed the US economy, I think it's more

0:26:51.960 --> 0:26:54.280
<v Speaker 2>appropriate to say that big tech capex has swallowed the

0:26:54.400 --> 0:26:57.080
<v Speaker 2>US economy. I also want to be clear that the economy,

0:26:57.119 --> 0:26:59.800
<v Speaker 2>which is the overall state of the country's production and

0:26:59.840 --> 0:27:02.679
<v Speaker 2>can asumption of stuff and the flow of money between

0:27:02.680 --> 0:27:05.280
<v Speaker 2>participants in said economy, and the markets as in the

0:27:05.280 --> 0:27:07.679
<v Speaker 2>stock market, well, those are two very different things. But

0:27:07.720 --> 0:27:10.520
<v Speaker 2>the calculations from Paul Kodrowski and others have now allowed

0:27:10.600 --> 0:27:13.280
<v Speaker 2>us to see where one might hit the other. You see,

0:27:13.320 --> 0:27:16.880
<v Speaker 2>the markets do not actually represent reality. While Microsoft, Amazon,

0:27:16.880 --> 0:27:18.520
<v Speaker 2>Google and Meta might want you to think there's a

0:27:18.600 --> 0:27:20.720
<v Speaker 2>ton of money in AI, their growth is mostly from

0:27:20.720 --> 0:27:23.960
<v Speaker 2>selling further iterations and contracts of their existing stuff, or

0:27:24.000 --> 0:27:26.480
<v Speaker 2>in the case of Meta further increasing its ad revenue.

0:27:26.680 --> 0:27:28.760
<v Speaker 2>The economy is where things are actually bought and sold,

0:27:28.800 --> 0:27:31.280
<v Speaker 2>representing the economic effects of both the things that happen

0:27:31.560 --> 0:27:34.040
<v Speaker 2>to build out the AI boom or whatever you call it,

0:27:34.280 --> 0:27:36.919
<v Speaker 2>and selling access to the services and the models themselves.

0:27:37.560 --> 0:27:39.920
<v Speaker 2>I recognize this is simplistic, but I'm laying it out

0:27:39.920 --> 0:27:43.680
<v Speaker 2>for a reason. As I discussed in the three part

0:27:43.720 --> 0:27:46.080
<v Speaker 2>Hater's Guide to the AI Bubble in video, is the

0:27:46.080 --> 0:27:48.919
<v Speaker 2>weak point in the US stock market, representing ninety percent

0:27:48.920 --> 0:27:51.000
<v Speaker 2>of the value of the Magnificent Seven, which in turn

0:27:51.040 --> 0:27:53.159
<v Speaker 2>makes up about thirty five percent of the value of

0:27:53.160 --> 0:27:56.440
<v Speaker 2>the U S stock market. The associated Magnificent Seven stocks

0:27:56.560 --> 0:27:58.920
<v Speaker 2>have seen a huge boom through their own growth, which

0:27:58.960 --> 0:28:01.800
<v Speaker 2>has been mistakenly in incorrectly attributed to revenue from AI,

0:28:01.840 --> 0:28:04.359
<v Speaker 2>which I've laid out previously is about forty billion dollars

0:28:04.760 --> 0:28:06.680
<v Speaker 2>of revenue in what in the last year or two,

0:28:07.040 --> 0:28:10.080
<v Speaker 2>maybe a little bit more. Nevertheless, the markets can continue

0:28:10.080 --> 0:28:12.040
<v Speaker 2>to be irrational because all they care about is number

0:28:12.080 --> 0:28:14.199
<v Speaker 2>going up. As the value of a stock is often

0:28:14.440 --> 0:28:17.680
<v Speaker 2>times disconnected from the value of the company itself, instead

0:28:18.080 --> 0:28:21.600
<v Speaker 2>associated with its kind of propensity for growth. It's fishy,

0:28:21.640 --> 0:28:24.520
<v Speaker 2>it sucks, and it's going to end badly. GDP and

0:28:24.560 --> 0:28:26.560
<v Speaker 2>other measurements of the economy aren't really something you can

0:28:26.640 --> 0:28:29.280
<v Speaker 2>fudge quite as easily, nor can you say a bunch

0:28:29.320 --> 0:28:31.320
<v Speaker 2>of fancy words to make people feel better in the

0:28:31.320 --> 0:28:33.160
<v Speaker 2>event that growth stores are the clients.

0:28:33.480 --> 0:28:36.320
<v Speaker 3>This leads me to my principal worry that AI.

0:28:36.119 --> 0:28:39.120
<v Speaker 2>Capex is actually a term for the expenditures of four companies,

0:28:39.440 --> 0:28:43.360
<v Speaker 2>namely Microsoft, Amazon, Google, and Meta, with Nvidia's GPU sales

0:28:43.360 --> 0:28:46.480
<v Speaker 2>being part of that capex too. What we include you

0:28:46.640 --> 0:28:49.560
<v Speaker 2>can if you want, include others like Oracle, XAI and

0:28:49.640 --> 0:28:53.400
<v Speaker 2>various neoclouds like Corweven, Crusoe, or who. According to DA

0:28:53.520 --> 0:28:56.280
<v Speaker 2>Davidson's gil Luria will account for about ten percent of

0:28:56.320 --> 0:28:59.400
<v Speaker 2>GPU sales from Nvideo in twenty twenty five. The reality

0:28:59.480 --> 0:29:02.600
<v Speaker 2>is that whatever our economic forces being driven by AI investment,

0:29:02.840 --> 0:29:05.840
<v Speaker 2>it's really just four companies building and leasing data centers

0:29:05.880 --> 0:29:08.440
<v Speaker 2>to burn on generatorvai a product that makes it relatively

0:29:08.480 --> 0:29:11.080
<v Speaker 2>small amount of money before losing a great deal more. Also,

0:29:11.120 --> 0:29:14.360
<v Speaker 2>want to be clear, Anthropic runs on Amazon and Google,

0:29:15.000 --> 0:29:18.400
<v Speaker 2>Open AI runs on Microsoft. These companies don't really own

0:29:18.440 --> 0:29:21.280
<v Speaker 2>their own infrastructure. That's what Open AI is allegedly building

0:29:21.320 --> 0:29:25.160
<v Speaker 2>with Oracle. Now forty two percent of Nvidia's revenue comes

0:29:25.160 --> 0:29:28.040
<v Speaker 2>from Magnificent seven per Laura Brand at Yahoo Finance, Big

0:29:28.080 --> 0:29:29.840
<v Speaker 2>up to Laura one of the best out there, which

0:29:29.920 --> 0:29:32.320
<v Speaker 2>naturally means that big tech is the lynchpin of investment

0:29:32.360 --> 0:29:35.320
<v Speaker 2>in data centers. I'll put it far more simply. If

0:29:35.360 --> 0:29:37.920
<v Speaker 2>AI capex represents such a large part of our GDP

0:29:38.040 --> 0:29:40.960
<v Speaker 2>and economic growth, our economy does on some level rest

0:29:40.960 --> 0:29:43.440
<v Speaker 2>on the back of Microsoft, Google, Meta, and Amazon and

0:29:43.480 --> 0:29:46.920
<v Speaker 2>their continued investment in AI. What should worry everybody is

0:29:46.920 --> 0:29:49.400
<v Speaker 2>that Microsoft, which makes up eighteen point nine percent of

0:29:49.520 --> 0:29:52.880
<v Speaker 2>Nvidia's revenue, assigned basically no leases in the last twelve months,

0:29:52.880 --> 0:29:55.920
<v Speaker 2>and it's committed data center construction and lamp purchases are

0:29:55.920 --> 0:29:59.560
<v Speaker 2>down year over year. Also, their data center capex is

0:29:59.600 --> 0:30:02.280
<v Speaker 2>not all to be AI, and all of these data

0:30:02.280 --> 0:30:05.840
<v Speaker 2>centers won't necessarily be used for that. In fact, nothing

0:30:05.880 --> 0:30:09.480
<v Speaker 2>they're doing suggests that they're super into it.

0:30:08.480 --> 0:30:10.200
<v Speaker 3>It's not good.

0:30:10.880 --> 0:30:13.000
<v Speaker 2>And while its CAPEX may not have dipped yet, in

0:30:13.040 --> 0:30:15.320
<v Speaker 2>part because the chip heavy nature of generative AI means

0:30:15.360 --> 0:30:19.680
<v Speaker 2>that CAPEX isn't exclusively dominated by property or construction. It's

0:30:19.720 --> 0:30:22.600
<v Speaker 2>now obvious that if it does, there will be direct

0:30:22.640 --> 0:30:24.880
<v Speaker 2>effects on both the US economy and the stock market,

0:30:25.040 --> 0:30:27.240
<v Speaker 2>as Microsoft is part of what amounts to a stimulus

0:30:27.240 --> 0:30:31.000
<v Speaker 2>packaged propping up America's economic growth. And not to repeat

0:30:31.000 --> 0:30:32.680
<v Speaker 2>the point too much, but Big Tech is yet to

0:30:32.680 --> 0:30:35.960
<v Speaker 2>actually turn anything resembling a profit on these data centers

0:30:36.000 --> 0:30:38.400
<v Speaker 2>and isn't making much revenue at all out of generative AI.

0:30:39.120 --> 0:30:41.040
<v Speaker 2>How exactly does this end?

0:30:41.240 --> 0:30:42.080
<v Speaker 3>What is the plan?

0:30:42.240 --> 0:30:44.680
<v Speaker 2>Here's Big Tech going to spend hundreds of billions of

0:30:44.720 --> 0:30:47.640
<v Speaker 2>dollars a year in CAPEX on generative AI in perpetuity?

0:30:47.840 --> 0:30:50.120
<v Speaker 2>Will they continue to buy more and more in video

0:30:50.200 --> 0:30:52.360
<v Speaker 2>chips as they do so, because it's always got to

0:30:52.400 --> 0:30:54.240
<v Speaker 2>be more. It can't be the same. They must get

0:30:54.280 --> 0:30:57.680
<v Speaker 2>more otherwise Nvidia doesn't grow At some point. Surely these

0:30:57.680 --> 0:31:01.000
<v Speaker 2>companies are built in enough data centers. Surely at some

0:31:01.160 --> 0:31:03.040
<v Speaker 2>point they will run out of space to put these

0:31:03.080 --> 0:31:05.640
<v Speaker 2>GPUs in. Is the plan to by then make so

0:31:05.720 --> 0:31:08.360
<v Speaker 2>much money from MAI that it won't matter? What does

0:31:08.480 --> 0:31:10.720
<v Speaker 2>Nvidia do at that point? How does the US economy

0:31:10.760 --> 0:31:12.720
<v Speaker 2>rebound from the loss of activity that follows?

0:31:12.920 --> 0:31:14.960
<v Speaker 3>I don't fucking know. And it's worrying me.

0:31:16.640 --> 0:31:19.400
<v Speaker 2>As I've said again and again, the genera IFAI bubble

0:31:19.520 --> 0:31:22.880
<v Speaker 2>is and always has been fundamentally irrational and inherently gothic,

0:31:22.960 --> 0:31:25.840
<v Speaker 2>playing in the ruins, patterns, and pathways of previous tech booms,

0:31:25.840 --> 0:31:28.840
<v Speaker 2>despite this one having little or no resemblance to them.

0:31:29.520 --> 0:31:31.520
<v Speaker 2>Though the tech industry loves to talk about building a

0:31:31.560 --> 0:31:34.960
<v Speaker 2>glorious future, its presence is one steeped in rituals of

0:31:35.000 --> 0:31:37.560
<v Speaker 2>death and decay, with the virtues of value creation and

0:31:37.560 --> 0:31:40.200
<v Speaker 2>productivity take a back seat to burning billions of dollars

0:31:40.200 --> 0:31:42.520
<v Speaker 2>and lying to the public again and again and again.

0:31:43.000 --> 0:31:45.080
<v Speaker 2>The way in which the media has participated in these

0:31:45.120 --> 0:31:49.200
<v Speaker 2>lives is disgusting. Venture capital, still drunk off the fumes

0:31:49.200 --> 0:31:52.120
<v Speaker 2>of twenty twenty one, keeps running the same old planebook,

0:31:52.600 --> 0:31:54.680
<v Speaker 2>shove as much money into a company as possible in

0:31:54.680 --> 0:31:56.360
<v Speaker 2>the hopes you can dump it onto an acquirer of

0:31:56.360 --> 0:31:58.840
<v Speaker 2>the public markets, only to get high on their own supply,

0:31:59.120 --> 0:32:01.920
<v Speaker 2>pushing valuations to the point that there's no possible liquidity

0:32:01.960 --> 0:32:04.360
<v Speaker 2>event for the majority of big private AI companies as

0:32:04.360 --> 0:32:07.960
<v Speaker 2>a result of their overstuffed valuations, burdens and business models,

0:32:08.040 --> 0:32:11.320
<v Speaker 2>and a lack of any real intellectual property, and like

0:32:11.360 --> 0:32:13.600
<v Speaker 2>the rest of the AI and bubble, Silicon Valley's only

0:32:13.680 --> 0:32:16.880
<v Speaker 2>liquidity path out of the bubble is big tech itself.

0:32:17.320 --> 0:32:20.400
<v Speaker 2>Without Google, character Dot AI, and WINDSORFS founders would likely

0:32:20.440 --> 0:32:22.680
<v Speaker 2>have been left for dead, and the same ghosts for Inflection,

0:32:22.800 --> 0:32:25.680
<v Speaker 2>And I'd even argue Scale AI, who's fourteen point three

0:32:25.720 --> 0:32:29.360
<v Speaker 2>billion dollar investment doing air quotes there for Meta effectively

0:32:29.360 --> 0:32:33.040
<v Speaker 2>decapitated the company, removing its CEO, Alexander Wang, leaving the

0:32:33.040 --> 0:32:36.080
<v Speaker 2>rest of the company to die playing are fourteen percent

0:32:36.080 --> 0:32:38.400
<v Speaker 2>of its staff on five hundred contractors mere weeks after

0:32:38.440 --> 0:32:42.680
<v Speaker 2>its CEO and investors cashed in. In fact, Generative AI

0:32:42.840 --> 0:32:45.680
<v Speaker 2>is turning out to be a fever dream entirely made up.

0:32:45.680 --> 0:32:46.400
<v Speaker 3>Big tech.

0:32:46.800 --> 0:32:48.800
<v Speaker 2>OpenAI would be dead if it wasn't for the massive

0:32:48.840 --> 0:32:51.800
<v Speaker 2>infrastructure provided by Microsoft at cost in returns for its

0:32:51.880 --> 0:32:54.320
<v Speaker 2>rights to its IP research and the ability to sell

0:32:54.360 --> 0:32:56.440
<v Speaker 2>its models, on top of tens of billions of dollars

0:32:56.480 --> 0:32:59.960
<v Speaker 2>of venture capital thrown into its billion dollar cache. Incinerator

0:33:00.600 --> 0:33:03.000
<v Speaker 2>Anthropic would be dead if both Google and Amazon, the

0:33:03.080 --> 0:33:06.240
<v Speaker 2>latter of which provides most of its infrastructure, hadn't invested

0:33:06.280 --> 0:33:08.200
<v Speaker 2>billions in keeping it alive. So that it can burn

0:33:08.280 --> 0:33:10.800
<v Speaker 2>three billion dollars or more in twenty twenty five while

0:33:10.840 --> 0:33:13.720
<v Speaker 2>fucking over its enterprise customers and rate limiting the rest.

0:33:14.560 --> 0:33:17.719
<v Speaker 2>The generative AI industry is, at its core unnatural. It

0:33:17.760 --> 0:33:20.680
<v Speaker 2>does not make significant revenue compared to its unbelievable costs,

0:33:20.720 --> 0:33:24.080
<v Speaker 2>nor does it have much revenue potential itself. It requires,

0:33:24.160 --> 0:33:27.880
<v Speaker 2>unlike just about every other software revolution, an unbelievable amount

0:33:27.920 --> 0:33:30.880
<v Speaker 2>of physical infrastructure to run because nobody but big tech

0:33:30.880 --> 0:33:33.880
<v Speaker 2>can afford to build the infrastructure necessary, creates very little

0:33:33.920 --> 0:33:37.520
<v Speaker 2>opportunity for competition or efficiency. As the markets are in

0:33:37.560 --> 0:33:39.680
<v Speaker 2>the throes of the growth of or costs rot economy,

0:33:39.840 --> 0:33:42.160
<v Speaker 2>they failed to keep big tech in line, conflating Big

0:33:42.200 --> 0:33:45.000
<v Speaker 2>text ability to grow with growth driven as a result

0:33:45.000 --> 0:33:48.720
<v Speaker 2>of the capital expenditures. Sensible reasonable markets would notice the

0:33:48.760 --> 0:33:51.800
<v Speaker 2>decay of free cash flow or ridiculousness of big text

0:33:51.800 --> 0:33:54.680
<v Speaker 2>capex bonanza, but instead they clap and squeal on oink

0:33:54.720 --> 0:33:57.520
<v Speaker 2>whenever Satching the Della jingles his keys. And I got

0:33:57.520 --> 0:33:59.200
<v Speaker 2>told the other day by someone that Satch in a

0:33:59.280 --> 0:34:01.360
<v Speaker 2>Della is not a is the citiot, that he's secretly

0:34:01.400 --> 0:34:03.880
<v Speaker 2>this big tech guy. And here's here's what I'd like

0:34:03.920 --> 0:34:07.040
<v Speaker 2>to say to them, it's fucking stupid. It's fucking dumb.

0:34:07.120 --> 0:34:10.680
<v Speaker 2>If this guy's actually technical, he's just a craven liar.

0:34:11.000 --> 0:34:14.400
<v Speaker 2>I actually think he's a bozo. Anyway, put in that side,

0:34:14.640 --> 0:34:18.560
<v Speaker 2>what's missing is any real value generation. Again, I tell you,

0:34:18.600 --> 0:34:21.120
<v Speaker 2>put aside any feelings you may have about generative AI

0:34:21.160 --> 0:34:24.320
<v Speaker 2>itself and focus on the actual economic results of this bubble.

0:34:24.600 --> 0:34:27.120
<v Speaker 2>How much revenue is there? Why is there no profit?

0:34:27.520 --> 0:34:29.640
<v Speaker 2>Why are there no exits? Why is big tech, which

0:34:29.640 --> 0:34:32.120
<v Speaker 2>has sunk hundreds of billions of dollars into generative AI,

0:34:32.440 --> 0:34:35.200
<v Speaker 2>not talk about the money they're making from it? Why

0:34:35.320 --> 0:34:38.359
<v Speaker 2>for three straight years have we been asked just wait

0:34:38.400 --> 0:34:40.360
<v Speaker 2>and see and for how long are we going to

0:34:40.400 --> 0:34:42.560
<v Speaker 2>have to wait? And when are we going to see it?

0:34:43.239 --> 0:34:46.200
<v Speaker 2>What's incredible is that the inherently compute intensive nature of

0:34:46.280 --> 0:34:50.160
<v Speaker 2>generative AI basically requires the construction of these facilities without

0:34:50.200 --> 0:34:52.680
<v Speaker 2>actually representing whether they are contributing to the revenues of

0:34:52.680 --> 0:34:55.719
<v Speaker 2>the companies that operate the models, like Anthropic or Open AI,

0:34:55.920 --> 0:34:58.560
<v Speaker 2>or any other business that builds on top of them.

0:34:58.880 --> 0:35:01.080
<v Speaker 2>As the models get more complet and hungary, more data

0:35:01.080 --> 0:35:03.800
<v Speaker 2>centers get built, which Hyperscaler's book as long term revenue,

0:35:03.840 --> 0:35:06.920
<v Speaker 2>even though it's subsidized by Hyperscalers or funded by VC money.

0:35:07.160 --> 0:35:10.240
<v Speaker 2>This in turn stimulates even more CAPEX spending, and without

0:35:10.280 --> 0:35:13.920
<v Speaker 2>having to answer any basic questions about longevity or market

0:35:13.960 --> 0:35:16.480
<v Speaker 2>fit longevity longevity.

0:35:16.040 --> 0:35:17.319
<v Speaker 3>I don't know yet.

0:35:17.360 --> 0:35:20.680
<v Speaker 2>The worst part of this financial farce is that we've

0:35:20.760 --> 0:35:23.560
<v Speaker 2>now got a built in economic breaking point in the

0:35:23.600 --> 0:35:27.120
<v Speaker 2>form of capex from AI. At some point, Kapex has

0:35:27.160 --> 0:35:29.239
<v Speaker 2>the slow if not because the lack of revenues are

0:35:29.280 --> 0:35:31.840
<v Speaker 2>massive costs associated, but because we live in a world

0:35:31.880 --> 0:35:35.840
<v Speaker 2>with finite space. And when said CAPEX slow happens, so

0:35:35.960 --> 0:35:38.640
<v Speaker 2>will the purchase of Nvidio GPUs, which in turn has

0:35:38.680 --> 0:35:42.480
<v Speaker 2>proven by Kadrowski and others slow America's economic growth. And

0:35:42.520 --> 0:35:44.480
<v Speaker 2>that growth is pretty much based on the whims of

0:35:44.520 --> 0:35:47.640
<v Speaker 2>four companies, which is an incredibly risky and scary proposition.

0:35:48.080 --> 0:35:50.120
<v Speaker 2>I haven't even dug into the wealth of private credit

0:35:50.120 --> 0:35:52.640
<v Speaker 2>deals that underpin build outs from private AI neo clouds

0:35:52.719 --> 0:35:55.239
<v Speaker 2>like core Wave, Cruise, So, Nebuis, and Lambda, in part

0:35:55.239 --> 0:35:57.440
<v Speaker 2>because their economic significance is so much smaller than the

0:35:57.440 --> 0:36:00.359
<v Speaker 2>big text ugly meanland the sprawl, and because I don't

0:36:00.400 --> 0:36:04.200
<v Speaker 2>like saying their names very much. To quote Paul Kodrowski,

0:36:05.160 --> 0:36:08.239
<v Speaker 2>we live in a historically anomalous moment. Regardless of what

0:36:08.320 --> 0:36:10.759
<v Speaker 2>one thinks about the merits of AI or explosive data

0:36:10.760 --> 0:36:14.000
<v Speaker 2>center expansion, the scale and pace of capital deployment into

0:36:14.000 --> 0:36:18.520
<v Speaker 2>a rapidly depreciating technology is remarkable. These are not railroads.

0:36:18.760 --> 0:36:22.600
<v Speaker 2>We aren't building century long infrastructure. AI data centers are

0:36:22.640 --> 0:36:27.360
<v Speaker 2>short lived, asset intensive facilities, writing declining cost technology curves,

0:36:27.400 --> 0:36:32.080
<v Speaker 2>requiring frequent hardware replacement to preserve margins, to which I say,

0:36:32.120 --> 0:36:35.000
<v Speaker 2>what margin's for? But nevertheless, you can't bail this out

0:36:35.040 --> 0:36:38.600
<v Speaker 2>because there is nothing to bail out. Microsoft, Meta, Amazon,

0:36:38.640 --> 0:36:40.600
<v Speaker 2>and Google have plenty of money and have proven they

0:36:40.600 --> 0:36:43.360
<v Speaker 2>can spend it in video is already doing everything it

0:36:43.400 --> 0:36:46.319
<v Speaker 2>can to justify people spending more on their GPUs. There's

0:36:46.360 --> 0:36:48.719
<v Speaker 2>little more it can do here other than soak up

0:36:48.760 --> 0:36:52.239
<v Speaker 2>the growth before the party ends. That CAPEX reduction will

0:36:52.239 --> 0:36:55.080
<v Speaker 2>bring with it a reduction in expenditures on those GPUs,

0:36:55.320 --> 0:36:57.680
<v Speaker 2>which will take a chunk out of the US stock market.

0:36:57.840 --> 0:37:00.360
<v Speaker 2>Although the stock market isn't the economy, the two are

0:37:00.440 --> 0:37:03.440
<v Speaker 2>inherently linked, and the popping of the AI bubble will

0:37:03.440 --> 0:37:06.879
<v Speaker 2>have down downstream ramifications, just that the dot com bubble did.

0:37:06.880 --> 0:37:10.000
<v Speaker 2>With the wider economy, expect to see an acceleration in

0:37:10.080 --> 0:37:12.600
<v Speaker 2>layoffs and offshoring in pop driven by a need for

0:37:12.640 --> 0:37:15.080
<v Speaker 2>tech companies to show for the first time in living memory,

0:37:15.320 --> 0:37:18.720
<v Speaker 2>fiscal restraint. For cities where tech is a major sector

0:37:18.719 --> 0:37:21.480
<v Speaker 2>of the economy thinks Seattle and San Francisco, there will

0:37:21.480 --> 0:37:23.799
<v Speaker 2>be knock on effects to those companies and the individuals

0:37:23.800 --> 0:37:27.560
<v Speaker 2>that support the tech sector, like restaurants, construction companies, building apartments,

0:37:27.800 --> 0:37:31.040
<v Speaker 2>uber drivers, and so on. We'll see a drying up

0:37:31.040 --> 0:37:33.960
<v Speaker 2>of VC funding, which is kind of already starting. Pension

0:37:34.000 --> 0:37:35.759
<v Speaker 2>funds will take a hit, which will affect how much

0:37:35.800 --> 0:37:37.320
<v Speaker 2>people have to spend in retirement.

0:37:37.760 --> 0:37:39.000
<v Speaker 3>It'll be grim.

0:37:39.280 --> 0:37:41.120
<v Speaker 2>Worse than that is the fact that these companies will

0:37:41.160 --> 0:37:44.000
<v Speaker 2>be by definition, non performing assets and ones that inflicted

0:37:44.040 --> 0:37:47.160
<v Speaker 2>an opportunity cost that will be impossible to calculate. While

0:37:47.160 --> 0:37:49.920
<v Speaker 2>a house, once built and sold, technically falls into that category,

0:37:50.160 --> 0:37:54.360
<v Speaker 2>doesn't add to any economic productivity. Obviously, people at least

0:37:54.400 --> 0:37:57.439
<v Speaker 2>need somewhere to live. Shelter is an essential component of life.

0:37:57.480 --> 0:37:59.400
<v Speaker 2>You can live without a data center the size of

0:37:59.440 --> 0:38:02.839
<v Speaker 2>Manhattan what would happen if companies like Microsoft and Meta

0:38:02.880 --> 0:38:05.440
<v Speaker 2>instead spent the money on things that actually drug productivity

0:38:05.520 --> 0:38:08.640
<v Speaker 2>or created a valuable competitive business that drove economic activity.

0:38:08.680 --> 0:38:11.160
<v Speaker 2>For example, Hell, even if they just gave everyone a

0:38:11.160 --> 0:38:13.440
<v Speaker 2>ten percent raise, it would likely be better for the

0:38:13.480 --> 0:38:15.840
<v Speaker 2>economy than this if we're factoring in things like consumer

0:38:15.880 --> 0:38:21.800
<v Speaker 2>spending instead. It's just waste, ugly, pointless waste. In summary,

0:38:21.840 --> 0:38:24.680
<v Speaker 2>we're already facing the prospect of a recession, and though

0:38:24.719 --> 0:38:28.080
<v Speaker 2>I'm not an economist, I can imagine being a particularly

0:38:28.160 --> 0:38:30.719
<v Speaker 2>nasty one. Given that the Magnificent seven account for forty

0:38:30.760 --> 0:38:33.280
<v Speaker 2>seven point eight seven percent of the Rustle one thousand

0:38:33.400 --> 0:38:36.719
<v Speaker 2>indexes returns in twenty twenty four. Even if big Tech

0:38:36.800 --> 0:38:39.759
<v Speaker 2>somehow makes this crap profitable, they won't. It's hard to

0:38:39.800 --> 0:38:42.560
<v Speaker 2>imagine that they'll be able to counterbalance any CAPEX reduction

0:38:42.640 --> 0:38:44.480
<v Speaker 2>with revenue, because there doesn't seem to be that much

0:38:44.520 --> 0:38:47.440
<v Speaker 2>revenue in generative AI to begin with. This is what

0:38:47.520 --> 0:38:49.799
<v Speaker 2>happens when you allow the rot economy to run wild,

0:38:49.920 --> 0:38:52.200
<v Speaker 2>building the stock market and tech industry on growth over

0:38:52.280 --> 0:38:55.560
<v Speaker 2>everything else. This is what happens when the tech media

0:38:55.600 --> 0:38:58.320
<v Speaker 2>repeatedly fails to hold the powerful to account catering to

0:38:58.360 --> 0:39:01.799
<v Speaker 2>their narratives and making excuses from are abominable billion dollar

0:39:01.880 --> 0:39:05.520
<v Speaker 2>losses and mediocre, questionably useful products. Well, for all you

0:39:05.560 --> 0:39:08.720
<v Speaker 2>want about the so called agentic era or anyonized revenues

0:39:08.719 --> 0:39:10.920
<v Speaker 2>that make you hot under the collar, I see no

0:39:11.040 --> 0:39:13.720
<v Speaker 2>reason for celebration about an industry with no exit plans

0:39:13.719 --> 0:39:15.920
<v Speaker 2>and needless capital expenditures that appeared to be one of

0:39:15.920 --> 0:39:19.040
<v Speaker 2>the few things keeping the American economy growing for effectively

0:39:19.080 --> 0:39:23.680
<v Speaker 2>no reason. I've been writing about the tech industry's obsession

0:39:23.719 --> 0:39:26.319
<v Speaker 2>with General IVAI for two years, and I've really never

0:39:26.360 --> 0:39:29.799
<v Speaker 2>felt more grim before. This was an economic uncertainty, a

0:39:29.800 --> 0:39:33.000
<v Speaker 2>way that our markets might contract, the big tech might

0:39:33.040 --> 0:39:35.440
<v Speaker 2>take a big haircunt, that a big a bunch of

0:39:35.440 --> 0:39:37.520
<v Speaker 2>money might have been wasted. But otherwise the world could

0:39:37.560 --> 0:39:42.359
<v Speaker 2>keep turning, right, it felt before Kutrowski said that thing,

0:39:42.360 --> 0:39:45.000
<v Speaker 2>and this was all sitting in public as well. I

0:39:45.000 --> 0:39:47.200
<v Speaker 2>don't know how I missed it. It felt until then

0:39:47.239 --> 0:39:49.000
<v Speaker 2>that it was like, Oh, this will really fuck up

0:39:49.040 --> 0:39:51.080
<v Speaker 2>big tech. It will hit their stocks. It'll be bad

0:39:51.120 --> 0:39:55.120
<v Speaker 2>for them now because the economy has slowed so much. Otherwise,

0:39:56.000 --> 0:39:58.640
<v Speaker 2>this data center, Capex and these four companies is holding

0:39:58.640 --> 0:40:02.080
<v Speaker 2>everything up and it won't lost. There's just no way

0:40:02.120 --> 0:40:04.680
<v Speaker 2>it can last forever, and there's not really anything to

0:40:04.840 --> 0:40:08.439
<v Speaker 2>bulk it up or and Microsoft's already pulling back.

0:40:08.960 --> 0:40:09.799
<v Speaker 3>How does this end?

0:40:09.840 --> 0:40:11.960
<v Speaker 2>I keep asking that the original title of this was

0:40:12.000 --> 0:40:14.000
<v Speaker 2>called how does this end? And I really don't know

0:40:14.040 --> 0:40:17.520
<v Speaker 2>how this ends. It feels as if everything is the

0:40:17.560 --> 0:40:20.480
<v Speaker 2>learning for disaster, and I really feel there's nothing that

0:40:20.480 --> 0:40:21.440
<v Speaker 2>can be done to avert in.

0:40:22.200 --> 0:40:25.000
<v Speaker 3>I will be here to tell you what's happening. Is

0:40:25.000 --> 0:40:25.600
<v Speaker 3>it happens?

0:40:26.200 --> 0:40:27.880
<v Speaker 2>And your third part for this week is going to

0:40:27.920 --> 0:40:30.360
<v Speaker 2>be about GPT five, which I'm going to be honest

0:40:31.120 --> 0:40:34.360
<v Speaker 2>doesn't really feel me full of hope or promise either.

0:40:42.719 --> 0:40:44.400
<v Speaker 2>Thank you for listening to Better Offline.

0:40:44.520 --> 0:40:46.960
<v Speaker 1>The editor and composer of the Better Offline theme song

0:40:47.040 --> 0:40:49.680
<v Speaker 1>is Matasowski. You can check out more of his music

0:40:49.719 --> 0:40:53.359
<v Speaker 1>and audio projects at Mattasowski dot com, m A T

0:40:53.360 --> 0:40:57.799
<v Speaker 1>T O S O W s ki dot com. You

0:40:57.840 --> 0:41:00.520
<v Speaker 1>can email me an easy at Better Offline dot or

0:41:00.560 --> 0:41:03.080
<v Speaker 1>visit Better Offline dot com to find more podcast links

0:41:03.080 --> 0:41:06.239
<v Speaker 1>and of course, my newsletter. I also really recommend you

0:41:06.239 --> 0:41:08.480
<v Speaker 1>go to chat dot where'soead dot at to visit the

0:41:08.520 --> 0:41:10.560
<v Speaker 1>Discord and go to our slash.

0:41:10.239 --> 0:41:13.400
<v Speaker 2>Better Offline to check out our reddit. Thank you so

0:41:13.480 --> 0:41:14.240
<v Speaker 2>much for listening.

0:41:15.080 --> 0:41:17.560
<v Speaker 3>Better Offline is a production of cool Zone Media.

0:41:17.680 --> 0:41:20.520
<v Speaker 2>For more from cool Zone Media, visit our website cool

0:41:20.600 --> 0:41:23.919
<v Speaker 2>Zonemedia dot com, or check us out on the iHeartRadio app,

0:41:24.000 --> 0:41:26.680
<v Speaker 2>Apple Podcasts, or wherever you get your podcasts.