WEBVTT - The Enshittifinancial Crisis: Part Three

0:00:03.160 --> 0:00:17.680
<v Speaker 1>Media. Hi, I'm ed Zetran and welcome to Better Offline.

0:00:19.800 --> 0:00:21.959
<v Speaker 2>That's right, folks, were back for the third part of

0:00:22.000 --> 0:00:24.319
<v Speaker 2>our four part series about the terminal stage of the

0:00:24.320 --> 0:00:29.680
<v Speaker 2>incitification process, the in shitification of the financial markets. So far,

0:00:29.720 --> 0:00:32.640
<v Speaker 2>we've talked about the incertification of the stock market, the

0:00:32.640 --> 0:00:35.400
<v Speaker 2>complicity of the analysts and the media, and how venture

0:00:35.400 --> 0:00:37.400
<v Speaker 2>capital will be the Canarian the coal mine for any

0:00:37.440 --> 0:00:40.960
<v Speaker 2>eventual collapse. Venture capital is both the perpetrator and the

0:00:41.040 --> 0:00:43.599
<v Speaker 2>victim of the air bubble, and it's worth exploring how

0:00:43.600 --> 0:00:47.000
<v Speaker 2>it came to this point. Now, hang on the gentle voice, folks.

0:00:47.000 --> 0:00:49.239
<v Speaker 2>We haven't seen values this big and alone time. These

0:00:49.280 --> 0:00:52.120
<v Speaker 2>the biggest numbers we've ever seen. They're simply tremendous open

0:00:52.159 --> 0:00:55.120
<v Speaker 2>airs maybe with eight hundred and thirty billion. Can you

0:00:55.160 --> 0:00:57.360
<v Speaker 2>believe that they lose so much money? But folks, we

0:00:57.360 --> 0:01:00.640
<v Speaker 2>don't mind. Clamisam Morbourn. They call him Klamiel because everybody

0:01:00.680 --> 0:01:02.440
<v Speaker 2>he's going to give them one billion dollars and where

0:01:02.480 --> 0:01:04.280
<v Speaker 2>we love him dead. Centers are going to have the

0:01:04.319 --> 0:01:07.319
<v Speaker 2>biggest deals we've ever seen, even even if we have

0:01:07.400 --> 0:01:10.240
<v Speaker 2>to work with Dario. Sorry to those of you who

0:01:10.280 --> 0:01:12.720
<v Speaker 2>don't like that voice. It was only about thirty seconds,

0:01:13.160 --> 0:01:16.280
<v Speaker 2>you see right Now, AI startups are big exciting news

0:01:16.280 --> 0:01:19.759
<v Speaker 2>for the limited partners funding large language model companies. Things

0:01:19.800 --> 0:01:22.360
<v Speaker 2>feel exciting because the value of the assets under management

0:01:22.560 --> 0:01:26.280
<v Speaker 2>aum are going up, which is nothing dodgy, per se,

0:01:26.319 --> 0:01:28.679
<v Speaker 2>but it's just how vcs value things. And if they're

0:01:28.760 --> 0:01:31.839
<v Speaker 2>valuing AI stocks, well that's how their fees are getting paid,

0:01:31.840 --> 0:01:34.480
<v Speaker 2>because the thing is they've invested in are worth more money,

0:01:34.480 --> 0:01:36.880
<v Speaker 2>and now limited partners are paying them a percentage so

0:01:36.920 --> 0:01:38.679
<v Speaker 2>that they can sit in an apartment that costs too

0:01:38.760 --> 0:01:42.000
<v Speaker 2>much not see friends because they just sit online posting

0:01:42.040 --> 0:01:44.480
<v Speaker 2>for eleven hours a day. Yeah, and then they can

0:01:44.560 --> 0:01:47.160
<v Speaker 2>lose all the money in the future. I guess investing

0:01:47.160 --> 0:01:49.520
<v Speaker 2>early in open Ai allows a bench capitalist or even

0:01:49.560 --> 0:01:52.280
<v Speaker 2>an asset manager like Blackstone, which invested in twenty twenty

0:01:52.280 --> 0:01:54.480
<v Speaker 2>four to say it as a big holding and a

0:01:54.480 --> 0:01:58.720
<v Speaker 2>big increase in its aum AI stocks, make venture capitalists

0:01:58.720 --> 0:02:00.840
<v Speaker 2>who bet on them two years ago, Like genius is

0:02:00.880 --> 0:02:05.680
<v Speaker 2>on paper, you got in early on open Ai, Athropic, Cursor, Cognition, Perplexity,

0:02:05.920 --> 0:02:08.360
<v Speaker 2>or any other company that loves to burn several dollars

0:02:08.360 --> 0:02:11.040
<v Speaker 2>per dollar of revenue. You have a big, beautiful number,

0:02:11.080 --> 0:02:13.440
<v Speaker 2>the biggest you've ever seen, and your limited partners to

0:02:13.480 --> 0:02:16.400
<v Speaker 2>pay you a feed just to manage it. Venture capital

0:02:16.440 --> 0:02:18.919
<v Speaker 2>has not seen valuations like this in a long time,

0:02:19.120 --> 0:02:21.519
<v Speaker 2>and on paper it feels a lot like a lot

0:02:21.520 --> 0:02:24.120
<v Speaker 2>of vcs got in on companies worth billions of dollars

0:02:24.240 --> 0:02:28.160
<v Speaker 2>right on paper, that is, on paper, Cognition is worth

0:02:28.200 --> 0:02:32.239
<v Speaker 2>ten point two billion, Perplexit eighteen billion, Cursor twenty nine

0:02:32.280 --> 0:02:36.079
<v Speaker 2>point three billion, Lovable six point six billion, Coher six

0:02:36.080 --> 0:02:39.720
<v Speaker 2>point eight billion, Repic three billion, Glean seven point two billion.

0:02:40.440 --> 0:02:43.320
<v Speaker 2>Massive valuations to companies that all basically do products that

0:02:43.360 --> 0:02:45.800
<v Speaker 2>open AI or Anthropic or Amazon or Google or any

0:02:45.880 --> 0:02:48.760
<v Speaker 2>number of Chinese companies are already working to clone. They're

0:02:48.760 --> 0:02:50.799
<v Speaker 2>all losing lots of money too, and have no path

0:02:50.880 --> 0:02:54.760
<v Speaker 2>the profitability of any kind. But right now the numbers

0:02:54.800 --> 0:02:58.160
<v Speaker 2>are simply tremendous. Folks. I've heard venture capitalists tell me

0:02:58.160 --> 0:03:00.320
<v Speaker 2>that there are times when they have to agree to

0:03:00.360 --> 0:03:03.280
<v Speaker 2>invest with little to no information or know that they'll

0:03:03.280 --> 0:03:08.239
<v Speaker 2>lose the opportunity to another sucker, I mean, investor. I've

0:03:08.280 --> 0:03:11.480
<v Speaker 2>heard venture capitalists say they don't have any insight into finances,

0:03:11.639 --> 0:03:14.600
<v Speaker 2>but we love the numbers. Folks, they're simply tremendous. We

0:03:14.760 --> 0:03:17.520
<v Speaker 2>love it. Let me tell you how not tremendous they are.

0:03:17.960 --> 0:03:20.959
<v Speaker 2>I had a source recently suggests that nobody who invested

0:03:21.160 --> 0:03:24.640
<v Speaker 2>pasted open AI's one hundred billion dollar valuation have any

0:03:24.680 --> 0:03:28.400
<v Speaker 2>information rates. And what that means is they literally do

0:03:28.440 --> 0:03:31.280
<v Speaker 2>not know anything about the company. They don't know anything

0:03:31.320 --> 0:03:32.800
<v Speaker 2>other than what the company will tell them. That the

0:03:32.800 --> 0:03:35.440
<v Speaker 2>company doesn't have to tell them shit, which means they

0:03:35.440 --> 0:03:37.680
<v Speaker 2>don't know the revenues, they don't know the costs. They

0:03:37.760 --> 0:03:42.280
<v Speaker 2>only know what new start new ridiculous stat Sarah Friar,

0:03:42.400 --> 0:03:45.680
<v Speaker 2>CFO of open ais, farted out. They posted recently that

0:03:45.760 --> 0:03:49.440
<v Speaker 2>twenty billion dollars annualized revenue. Yeah. I think that's complete nonsense,

0:03:49.480 --> 0:03:51.360
<v Speaker 2>But they want to do that so that you go

0:03:51.440 --> 0:03:54.000
<v Speaker 2>open ai made twenty billion dollars in twenty twenty five.

0:03:54.480 --> 0:03:57.440
<v Speaker 2>Fuck that, I don't even think they made ten. Now,

0:03:57.520 --> 0:04:01.400
<v Speaker 2>venture capitalists would argue, of course that they'd say I'm insane.

0:04:01.200 --> 0:04:04.440
<v Speaker 2>They would say they would think me mad, and they

0:04:04.480 --> 0:04:07.240
<v Speaker 2>would say that the growth is obviously there, while pointing

0:04:07.240 --> 0:04:09.760
<v Speaker 2>to whatever startup has made one hundred million dollars in

0:04:09.880 --> 0:04:13.560
<v Speaker 2>annualized recurring revenue, which is eight point three million dollars

0:04:13.600 --> 0:04:16.560
<v Speaker 2>a month, by the way, all while not discussing the

0:04:16.640 --> 0:04:19.960
<v Speaker 2>underlying operating expenses. The idea I believe is that the

0:04:19.960 --> 0:04:22.839
<v Speaker 2>current state of AI spending is only set to increase

0:04:22.880 --> 0:04:27.479
<v Speaker 2>the next year, and that will somehow lead to fixing

0:04:27.520 --> 0:04:32.120
<v Speaker 2>the margins. I think venture capitalists staunchly refuse to learn

0:04:32.160 --> 0:04:34.760
<v Speaker 2>anything other than investing growth and then profit from growth,

0:04:34.800 --> 0:04:38.520
<v Speaker 2>even if profiting from growth doesn't appear to be happening anymore.

0:04:39.120 --> 0:04:41.760
<v Speaker 2>In reality, venture capital shouldn't have touched LMS with a

0:04:41.800 --> 0:04:45.240
<v Speaker 2>shitty stick because the margins were obviously blatantly bad from

0:04:45.279 --> 0:04:48.480
<v Speaker 2>the very beginning. We knew open ai would lose five

0:04:48.480 --> 0:04:51.200
<v Speaker 2>million dollars in the middle of twenty twenty four. The

0:04:51.240 --> 0:04:54.159
<v Speaker 2>same venture capital climate would have fucking panicked, but instead

0:04:54.240 --> 0:04:57.480
<v Speaker 2>chose to double dribble and kadruple down. I believe that

0:04:57.520 --> 0:05:01.040
<v Speaker 2>massive valuation drawdowns are a certainty for AI companies, and

0:05:01.080 --> 0:05:03.800
<v Speaker 2>by that I mean these companies will end up being

0:05:03.839 --> 0:05:06.479
<v Speaker 2>revalued and the number will be smaller than the big

0:05:06.560 --> 0:05:12.159
<v Speaker 2>number everybody fell in love with. Look, venture capitalists, I

0:05:12.200 --> 0:05:15.320
<v Speaker 2>gotta ask you a question, what happens if open AI dies?

0:05:15.680 --> 0:05:18.400
<v Speaker 2>Do you think this will make investors interested in funding

0:05:18.480 --> 0:05:21.440
<v Speaker 2>or acquiring other AI startups. How much longer are we

0:05:21.480 --> 0:05:24.120
<v Speaker 2>going to do this? When will Venture Capital realize it's

0:05:24.120 --> 0:05:27.279
<v Speaker 2>setting itself up for a disaster And what exactly is

0:05:27.320 --> 0:05:31.360
<v Speaker 2>the plan? Open AI and anthropic will cuck the lake's

0:05:31.440 --> 0:05:34.640
<v Speaker 2>dry like an Nvidia GPU named after Nancy Reagan? How

0:05:34.720 --> 0:05:36.560
<v Speaker 2>is this meant to continue and what will be left

0:05:36.600 --> 0:05:39.719
<v Speaker 2>when it does? The answer is simple, there won't be

0:05:39.800 --> 0:05:42.760
<v Speaker 2>any money for venture capital for a while. Those AI

0:05:42.839 --> 0:05:45.000
<v Speaker 2>holdings are going to be worth at best fifty percent

0:05:45.200 --> 0:05:48.400
<v Speaker 2>if they retain any value at all. Once one of

0:05:48.400 --> 0:05:51.440
<v Speaker 2>these startups die, Once there's a major casualty, a panic

0:05:51.480 --> 0:05:54.200
<v Speaker 2>will ensue, sending venture capitalists scrambling to get their holdings

0:05:54.240 --> 0:05:57.479
<v Speaker 2>acquired by which I mean sold to someone else, until

0:05:57.480 --> 0:06:00.960
<v Speaker 2>there's little or no investor interest left. Why would limited

0:06:01.000 --> 0:06:04.279
<v Speaker 2>partners ever trust venture capital after this? Why would anybody?

0:06:04.600 --> 0:06:06.760
<v Speaker 2>Because based on the past four years, it doesn't appear

0:06:06.760 --> 0:06:09.279
<v Speaker 2>that venture capital is actually good at investing money. It

0:06:09.360 --> 0:06:11.520
<v Speaker 2>just got lucky year after year until there were a

0:06:11.520 --> 0:06:13.600
<v Speaker 2>few ideas that they could sell for one hundreds of

0:06:13.600 --> 0:06:17.760
<v Speaker 2>billions or billions of dollars, And now we're out of them,

0:06:18.080 --> 0:06:20.880
<v Speaker 2>we're out. I think we're out. I think there's probably

0:06:21.000 --> 0:06:23.800
<v Speaker 2>still some, but I think the era of the unicorn

0:06:23.880 --> 0:06:27.000
<v Speaker 2>might be dead. For a while, venture capital believed it

0:06:27.040 --> 0:06:30.039
<v Speaker 2>knew better as it turned its back on basic business fundamentals,

0:06:30.040 --> 0:06:34.360
<v Speaker 2>starting with Clubhouse, Crypto, the Metaverse, and now Generative AI.

0:06:34.760 --> 0:06:37.000
<v Speaker 2>Yet they're far from the only fuck wits on the

0:06:37.000 --> 0:06:40.800
<v Speaker 2>Dickhead Express. Because yes, I am going to talk about

0:06:40.920 --> 0:06:44.599
<v Speaker 2>data centers. Per Bloomberg, there were at least one hundred

0:06:44.600 --> 0:06:46.920
<v Speaker 2>and seventy eight point five billion dollars in data center

0:06:46.960 --> 0:06:49.120
<v Speaker 2>credit deals and in the US in twenty twenty five,

0:06:49.480 --> 0:06:51.880
<v Speaker 2>rivaling the two hundred and fifteen point four billion dollars

0:06:51.880 --> 0:06:54.400
<v Speaker 2>invested in US venture capital in twenty twenty four and

0:06:54.440 --> 0:06:56.559
<v Speaker 2>the one hundred and ninety seven point two billion dollars

0:06:56.640 --> 0:06:59.359
<v Speaker 2>invested in US venture capital through August seventh to twenty

0:06:59.360 --> 0:07:02.680
<v Speaker 2>twenty five, and over one hundred billion dollars more than

0:07:02.720 --> 0:07:05.360
<v Speaker 2>the sixty point sixty nine billion dollars of data center

0:07:05.440 --> 0:07:09.680
<v Speaker 2>deals done in twenty twenty four. That's that's really worrying.

0:07:10.560 --> 0:07:13.440
<v Speaker 2>It's like more than more than more than one hundred

0:07:13.440 --> 0:07:15.480
<v Speaker 2>percent more. I'm very worried, and I'm going to tell

0:07:15.520 --> 0:07:19.160
<v Speaker 2>you why, using a company called core Weave, core Weave

0:07:19.280 --> 0:07:22.760
<v Speaker 2>that I've been actively warning about, warning people about since March.

0:07:22.800 --> 0:07:26.280
<v Speaker 2>And I get a little crazy around corwaf because I

0:07:26.760 --> 0:07:30.360
<v Speaker 2>have had the core Weave conversation a lot, and I've

0:07:30.360 --> 0:07:33.000
<v Speaker 2>seen people straight up just steal the title of my

0:07:33.120 --> 0:07:36.160
<v Speaker 2>articles as they come to this realization six to seven

0:07:36.160 --> 0:07:39.480
<v Speaker 2>months later. But cor Weave is a mess, and corwy

0:07:39.680 --> 0:07:42.880
<v Speaker 2>was obviously always a mess. And everyone has been saying

0:07:42.880 --> 0:07:47.080
<v Speaker 2>Corwave's fine because they went public. Core Weave is not fine.

0:07:48.000 --> 0:07:52.960
<v Speaker 2>All right, Let's talk about core Weave. Corwave is something

0:07:52.960 --> 0:07:56.440
<v Speaker 2>called a neocloud. It's a company that builds data centers

0:07:56.560 --> 0:07:59.119
<v Speaker 2>by renting out, and they build data centers they film

0:07:59.160 --> 0:08:01.200
<v Speaker 2>full of GPUs and they rent them to people. They're

0:08:01.240 --> 0:08:03.760
<v Speaker 2>in something called AI Compume, And as I explained a

0:08:03.760 --> 0:08:06.240
<v Speaker 2>few months ago, they do so by building data centers

0:08:06.240 --> 0:08:10.840
<v Speaker 2>backed by endless debt. And I quote myself saying, up

0:08:10.840 --> 0:08:13.520
<v Speaker 2>a neocloud is expensive. Even if the company in question

0:08:13.560 --> 0:08:15.680
<v Speaker 2>already has data centers, as core We've did with its

0:08:15.680 --> 0:08:20.200
<v Speaker 2>scryptocurrency mining operation, AI requires completely new data center infrastructure

0:08:20.200 --> 0:08:22.880
<v Speaker 2>to house and call the GPUs, and those GPUs also

0:08:22.960 --> 0:08:26.360
<v Speaker 2>need paying for, and then there's the other stuff like power, water,

0:08:26.440 --> 0:08:28.760
<v Speaker 2>and the other bits of the computer, the CPU, the motherboard,

0:08:28.800 --> 0:08:30.280
<v Speaker 2>the memory of the storage, and the housing and the

0:08:30.320 --> 0:08:33.319
<v Speaker 2>power supplying all that good stuff. As a result, these

0:08:33.360 --> 0:08:35.680
<v Speaker 2>neoclouds are forced to raise billions of dollars in debt,

0:08:35.800 --> 0:08:39.160
<v Speaker 2>which they collateralize using the GPUs they already have, along

0:08:39.200 --> 0:08:44.079
<v Speaker 2>with contracts from customers which they use to buy more GPUs. Corwei,

0:08:44.160 --> 0:08:46.720
<v Speaker 2>for example, has twenty five billion dollars in debt on

0:08:46.760 --> 0:08:49.319
<v Speaker 2>an estimated five point three five billion dollars in revenue,

0:08:49.440 --> 0:08:52.200
<v Speaker 2>losing hundreds of billions of dollars in a quarter. You

0:08:52.280 --> 0:08:57.160
<v Speaker 2>know who also invests in these neoclouds in Video, the

0:08:57.200 --> 0:09:00.160
<v Speaker 2>people who make the GPUs in VideA is also one

0:09:00.160 --> 0:09:02.920
<v Speaker 2>of Corwey's largest customers, accounting for fifty percent of its

0:09:02.960 --> 0:09:05.079
<v Speaker 2>revenue in twenty twenty four, and just signed a deal

0:09:05.120 --> 0:09:07.480
<v Speaker 2>to buy six point three billion dollars of any capacity

0:09:07.600 --> 0:09:09.800
<v Speaker 2>the care we can't otherwise sell to someone else through

0:09:09.800 --> 0:09:12.640
<v Speaker 2>twenty thirty two, an extension of a one point three

0:09:12.720 --> 0:09:15.360
<v Speaker 2>billion dollar deal from twenty twenty three reported by The

0:09:15.360 --> 0:09:18.520
<v Speaker 2>Information oh And in Video was the anchor investor of

0:09:18.559 --> 0:09:21.640
<v Speaker 2>Corweav's IPO, putting in two hundred and fifty million dollars.

0:09:21.960 --> 0:09:25.160
<v Speaker 2>Corwave is also one of the largest providers AI compute

0:09:25.200 --> 0:09:28.040
<v Speaker 2>in the world, and its business model is indicative of

0:09:28.080 --> 0:09:31.200
<v Speaker 2>how most data center companies make money. In fact, i'd

0:09:31.280 --> 0:09:33.600
<v Speaker 2>argue that they are probably doing better because they have

0:09:33.640 --> 0:09:36.560
<v Speaker 2>the backing of everyone. Now you may think that's a

0:09:36.559 --> 0:09:38.720
<v Speaker 2>too big to fail thing. Not really, It just means

0:09:38.720 --> 0:09:43.439
<v Speaker 2>that they're kind of the village bicycle of AI compute grossness. Aside,

0:09:43.800 --> 0:09:46.760
<v Speaker 2>let me explain how this works. First, core we've signed

0:09:46.800 --> 0:09:49.360
<v Speaker 2>contracts such as it's fourteen billion dollar deal with Meto

0:09:49.440 --> 0:09:51.400
<v Speaker 2>or it's twenty two point five billion dollar deal with

0:09:51.440 --> 0:09:54.480
<v Speaker 2>open Ai before it has the physical infrastructure to actually

0:09:54.520 --> 0:09:57.800
<v Speaker 2>service them. It then raises debt using this contract as

0:09:57.800 --> 0:10:01.680
<v Speaker 2>collateral orders GPUs from VideA. Those take three months to

0:10:01.840 --> 0:10:04.600
<v Speaker 2>arrive and then another three months to install, at which

0:10:04.600 --> 0:10:07.880
<v Speaker 2>point monthly client payments begin. And I should also add

0:10:07.920 --> 0:10:10.120
<v Speaker 2>that that's kind of taken out of deferred revenue that

0:10:10.160 --> 0:10:13.839
<v Speaker 2>the customers already put down. To really simplify this, data

0:10:13.840 --> 0:10:16.800
<v Speaker 2>center developers are raising money months up to a year

0:10:16.840 --> 0:10:19.360
<v Speaker 2>before they ever expect to make a penny of revenue,

0:10:19.800 --> 0:10:23.319
<v Speaker 2>not profit. In fact, I can find no consistent answer

0:10:23.520 --> 0:10:25.480
<v Speaker 2>to how long a data center takes to build, and

0:10:25.520 --> 0:10:28.080
<v Speaker 2>the answer here is pretty important because that's how the

0:10:28.080 --> 0:10:29.800
<v Speaker 2>money is going to get made from the data center's

0:10:29.800 --> 0:10:33.520
<v Speaker 2>for building fucking everywhere. You may also notice that monthly

0:10:33.559 --> 0:10:35.840
<v Speaker 2>payments in the chart that I linked, which I should

0:10:35.880 --> 0:10:38.959
<v Speaker 2>have mentioned, begin at six to thirty months, a curious

0:10:38.960 --> 0:10:41.800
<v Speaker 2>and broad blob of time. You see, data centers are

0:10:41.800 --> 0:10:44.280
<v Speaker 2>extremely difficult to build, and the concept of an AI

0:10:44.400 --> 0:10:46.640
<v Speaker 2>data center is barely a few years old, with the

0:10:46.679 --> 0:10:49.800
<v Speaker 2>concepts of hundreds of megawats in one data center campus

0:10:49.920 --> 0:10:53.680
<v Speaker 2>entirely made up of aigpus barely two years old jpus,

0:10:54.120 --> 0:10:57.120
<v Speaker 2>which means basically everybody building one is doing so for

0:10:57.160 --> 0:11:00.440
<v Speaker 2>the first time, and even experienced developers are running into problems.

0:11:01.720 --> 0:11:05.839
<v Speaker 2>For example, core Scientific, Corweave's weird partner organization it tried

0:11:05.880 --> 0:11:08.440
<v Speaker 2>and failed to buy that is truly different despite the

0:11:08.440 --> 0:11:10.960
<v Speaker 2>similar name, has been trying to convert it's Denton, Texas

0:11:11.000 --> 0:11:13.960
<v Speaker 2>scriptocurrency mining data center into an AI data center since

0:11:14.000 --> 0:11:17.080
<v Speaker 2>November twenty twenty four, specifically so that Corewave can rent

0:11:17.080 --> 0:11:21.080
<v Speaker 2>it to Microsoft for Open AI. This hasn't gone well,

0:11:21.320 --> 0:11:23.959
<v Speaker 2>and this is when things get weird. The Wall Street

0:11:24.000 --> 0:11:26.360
<v Speaker 2>Journal reported a few weeks ago that Denton had been

0:11:26.400 --> 0:11:29.160
<v Speaker 2>racked with several months of delays thanks to rainstorms preventing

0:11:29.200 --> 0:11:34.200
<v Speaker 2>contractors and pooring contry contrete. I'm going with it. It's contrete.

0:11:34.240 --> 0:11:39.000
<v Speaker 2>Now we're calling it contry concrete. I originally, within this

0:11:39.080 --> 0:11:41.360
<v Speaker 2>script wrote down that this was a possible thing. I

0:11:41.480 --> 0:11:45.840
<v Speaker 2>thought that this was normal. However, I should add something,

0:11:46.240 --> 0:11:49.920
<v Speaker 2>this is this may actually be bollocks. So I'm a psycho,

0:11:50.040 --> 0:11:53.360
<v Speaker 2>and I went and I looked up whether. I looked

0:11:53.440 --> 0:11:56.960
<v Speaker 2>up weather in Abilene, Texas, and it turns out that

0:11:57.040 --> 0:12:01.439
<v Speaker 2>there were like eleven days with rain total, and there

0:12:01.520 --> 0:12:04.200
<v Speaker 2>was only I think on June third and September twenty second,

0:12:04.240 --> 0:12:07.120
<v Speaker 2>there were rainstorms or thunderstorms of any kind. I think

0:12:07.160 --> 0:12:09.000
<v Speaker 2>that there are only two days where rain was over

0:12:09.120 --> 0:12:13.520
<v Speaker 2>point one inches. The people I've talked to about this

0:12:13.640 --> 0:12:15.960
<v Speaker 2>deal are saying it wouldn't be weather, it would actually

0:12:16.040 --> 0:12:19.760
<v Speaker 2>just be money. Nevertheless, those are things that are stopping

0:12:19.760 --> 0:12:22.319
<v Speaker 2>them building a two hundred and sixty megawat data center,

0:12:22.760 --> 0:12:25.560
<v Speaker 2>and that's fucking nothing compared to the Giga. What's the

0:12:25.600 --> 0:12:28.959
<v Speaker 2>open Ai claims to want to build. And what this means,

0:12:29.000 --> 0:12:31.960
<v Speaker 2>by the way, is that call we've can't actually get

0:12:31.960 --> 0:12:34.920
<v Speaker 2>paid by open Ai because per its contract, customers don't

0:12:34.920 --> 0:12:37.559
<v Speaker 2>have to stop paying until the computer is actually available.

0:12:38.640 --> 0:12:40.920
<v Speaker 2>This is a very important deal to know for literally

0:12:40.960 --> 0:12:43.679
<v Speaker 2>any data center development you've ever heard of or seen.

0:12:55.160 --> 0:12:57.800
<v Speaker 2>As of cour we've SQ three twenty twenty five earnings,

0:12:57.880 --> 0:13:00.000
<v Speaker 2>they're sitting on one point one billion dollars into thirst

0:13:00.200 --> 0:13:03.040
<v Speaker 2>revenue that's income for service is not yet rendered are

0:13:03.080 --> 0:13:05.200
<v Speaker 2>from nine hundred and fifty one million in Q two

0:13:05.240 --> 0:13:08.160
<v Speaker 2>and four hundred and thirty six million in Q one

0:13:08.240 --> 0:13:11.080
<v Speaker 2>twenty twenty five. This means deposits have been made, but

0:13:11.120 --> 0:13:14.920
<v Speaker 2>the contract is yet to be serviced. Now, I'm also

0:13:15.040 --> 0:13:16.839
<v Speaker 2>a curious little critter, so I went and found the

0:13:16.880 --> 0:13:19.040
<v Speaker 2>nine hundred and twenty one page two point six billion

0:13:19.080 --> 0:13:23.840
<v Speaker 2>dollar DDTL delayed draw Term Loan three point zero loan

0:13:23.840 --> 0:13:28.280
<v Speaker 2>agreement between core Weave and banks including Morgan Stanley, Mitsubishi,

0:13:28.400 --> 0:13:31.080
<v Speaker 2>UFG and Gold and Sachs, and in doing so I

0:13:31.160 --> 0:13:34.360
<v Speaker 2>learned something well. First of all, open ai appears to

0:13:34.360 --> 0:13:36.960
<v Speaker 2>have net three sixty payment terms from Corwave, meaning that

0:13:37.000 --> 0:13:40.880
<v Speaker 2>it can literally pay a year from invoice. Second, Corweve

0:13:41.000 --> 0:13:44.760
<v Speaker 2>is required to maintain something called a contract realization ratio

0:13:44.760 --> 0:13:47.240
<v Speaker 2>of point eighty five times, meaning that core Weave has

0:13:47.280 --> 0:13:49.640
<v Speaker 2>to make at least eighty five cents of every expected

0:13:49.679 --> 0:13:52.360
<v Speaker 2>dollar or it is in default of its loan. This

0:13:52.480 --> 0:13:54.280
<v Speaker 2>is important to note because it means that if say

0:13:54.280 --> 0:13:56.200
<v Speaker 2>open Ai decides not to pay up in a year,

0:13:56.440 --> 0:13:59.480
<v Speaker 2>corwave will be up shit creek without a paddle. Now,

0:13:59.520 --> 0:14:02.280
<v Speaker 2>I apolog that suggests that core Weave is an already

0:14:02.360 --> 0:14:04.960
<v Speaker 2>up shit creek, or that it might have a paddle

0:14:05.000 --> 0:14:08.200
<v Speaker 2>of some sort buried inside in Vidia's latest earnings on

0:14:08.240 --> 0:14:11.920
<v Speaker 2>page seventeen, there's a little clue at home, and I

0:14:12.040 --> 0:14:15.760
<v Speaker 2>quote in the third quarter of fiscal year twenty twenty six,

0:14:15.800 --> 0:14:18.199
<v Speaker 2>we entered into an agreement to guarantee of partner's facility

0:14:18.280 --> 0:14:20.960
<v Speaker 2>lease obligations in the event of their default. The agreement

0:14:21.040 --> 0:14:24.200
<v Speaker 2>allows our partner to secure a limited availability facility lease

0:14:24.320 --> 0:14:28.000
<v Speaker 2>backed by our credit profile in exchange for issuing US warrants.

0:14:28.160 --> 0:14:30.680
<v Speaker 2>The maximum gross exposure is eight hundred and sixty million,

0:14:30.880 --> 0:14:32.880
<v Speaker 2>which is reduced as the partner makes payments to the

0:14:32.920 --> 0:14:35.480
<v Speaker 2>leaser of a five years. The partner has placed four

0:14:35.520 --> 0:14:38.200
<v Speaker 2>hundred and seventy million dollars in escrow and executed an

0:14:38.200 --> 0:14:41.000
<v Speaker 2>agreement to sell the data center cloud capacity, mitigating our

0:14:41.040 --> 0:14:44.920
<v Speaker 2>default risk credit where credit is due. Eaglide analysts Just

0:14:45.040 --> 0:14:47.760
<v Speaker 2>Dario caught this in November, but in core weaves condensed

0:14:47.760 --> 0:14:50.200
<v Speaker 2>to consolidated balance sheets, there sits a four hundred and

0:14:50.240 --> 0:14:53.000
<v Speaker 2>seventy seven point five million dollars line item under restricted

0:14:53.000 --> 0:14:56.400
<v Speaker 2>cash and cash equivalence non current, though this might not

0:14:56.520 --> 0:15:00.320
<v Speaker 2>be in Vidia's escrow. This number shifted from six hundred

0:15:00.320 --> 0:15:02.640
<v Speaker 2>and seventeen in Q one to three hundred and forty

0:15:02.680 --> 0:15:05.160
<v Speaker 2>million in Q two. It lines up all too precisely,

0:15:05.880 --> 0:15:10.800
<v Speaker 2>and who else would in video be guaranteeing In any case,

0:15:10.840 --> 0:15:12.760
<v Speaker 2>corweav is likely to get the best deals in data

0:15:12.760 --> 0:15:15.840
<v Speaker 2>centered debt outside of Oracle. It has top tier financiers

0:15:15.880 --> 0:15:17.880
<v Speaker 2>who I'll get to in a little bit, the fallbacking

0:15:18.280 --> 0:15:20.720
<v Speaker 2>in Video who is both an investor customer and apparent

0:15:20.760 --> 0:15:23.640
<v Speaker 2>financial backstop, and also Core with as a customer of

0:15:23.680 --> 0:15:27.360
<v Speaker 2>theirs and the ability to raise debt quickly. Corweb's deals

0:15:27.360 --> 0:15:30.160
<v Speaker 2>are likely indicative of how data center financing takes place,

0:15:30.160 --> 0:15:34.680
<v Speaker 2>and those top tier financiers it's they may basically been

0:15:34.720 --> 0:15:38.160
<v Speaker 2>in every single deal. It's actually it's actually really worrying

0:15:38.200 --> 0:15:40.680
<v Speaker 2>when you, I mean, you lay it all out, which

0:15:40.720 --> 0:15:43.400
<v Speaker 2>is exactly what I'm going to do. So who's actually

0:15:43.400 --> 0:15:45.720
<v Speaker 2>paying for this shit? I went and dug through a

0:15:45.760 --> 0:15:48.640
<v Speaker 2>pile of twenty six prominent data center loan deals, including

0:15:48.680 --> 0:15:51.880
<v Speaker 2>the proposed thirty eight billion dollar debt package that Oracle

0:15:51.920 --> 0:15:55.760
<v Speaker 2>Advantage data Center partners are raising for Stargate Shackelford and Wisconsin,

0:15:55.840 --> 0:15:59.880
<v Speaker 2>Stargate Ablene, New Mexico, and of course soft Banks fifty

0:16:00.040 --> 0:16:02.120
<v Speaker 2>billion dollar bridge loan, which I included for a reason

0:16:02.120 --> 0:16:04.760
<v Speaker 2>that will become obvious shortly, and multiple core wave loans.

0:16:04.760 --> 0:16:07.080
<v Speaker 2>I've found a few commonalities, and forgive me, I'm about

0:16:07.080 --> 0:16:08.640
<v Speaker 2>to say a lot of names and numbers. You might

0:16:08.680 --> 0:16:10.560
<v Speaker 2>want to pull up the newsletter to read along with

0:16:10.600 --> 0:16:14.640
<v Speaker 2>it like a sing along. Now, Blue Owl, Blue ol,

0:16:14.680 --> 0:16:17.080
<v Speaker 2>We love blue ol folks. They were present in every

0:16:17.080 --> 0:16:20.120
<v Speaker 2>single Stargate deal other than the thirty eight billion dollar

0:16:20.160 --> 0:16:23.000
<v Speaker 2>deal that's being raised by Advantage that is yet to close.

0:16:23.960 --> 0:16:26.040
<v Speaker 2>Blue Hour was also involved in a one point three

0:16:26.040 --> 0:16:28.680
<v Speaker 2>billion dollar Australian data set and a debt package by

0:16:28.760 --> 0:16:33.960
<v Speaker 2>Virtue of owning Stack infrastructure. Remember that name MUFG MUG

0:16:34.400 --> 0:16:37.280
<v Speaker 2>also Mitsubishi UFJ Financial Group is their full name, but

0:16:37.320 --> 0:16:40.160
<v Speaker 2>I like saying. MUG was present in seventeen out of

0:16:40.160 --> 0:16:43.359
<v Speaker 2>twenty six of the deals, including three separate core weafinancing,

0:16:43.560 --> 0:16:47.000
<v Speaker 2>Stargate New Mexico an eighteen billion dollar deal, the alleged

0:16:47.040 --> 0:16:50.600
<v Speaker 2>Stargate thirty eight billion dollar deal, SoftBank's bridge loan which

0:16:50.640 --> 0:16:52.720
<v Speaker 2>they had to raise by the way, to fund open AI,

0:16:52.960 --> 0:16:55.520
<v Speaker 2>and a five billion dollar green loan package for Vantage

0:16:55.600 --> 0:16:59.240
<v Speaker 2>Data Centers, who you might have just heard me say. JP.

0:16:59.360 --> 0:17:02.080
<v Speaker 2>Morgan Chase was involved in eight deals, and they were

0:17:02.120 --> 0:17:04.600
<v Speaker 2>involved in some of the largest. Two cory was October

0:17:04.600 --> 0:17:08.560
<v Speaker 2>twenty twenty four financing their third delayed draw Tone loan

0:17:08.920 --> 0:17:12.359
<v Speaker 2>and November financing as well the funding behind Stargate, Apple

0:17:12.400 --> 0:17:14.879
<v Speaker 2>in the thirty eight billion dollar Oracle deal, and Blue

0:17:14.880 --> 0:17:18.480
<v Speaker 2>Owl's acquisition of IPI Partners in twenty twenty four. And yes,

0:17:18.520 --> 0:17:20.960
<v Speaker 2>they were also part of soft banks bridge loan. Now,

0:17:21.040 --> 0:17:24.000
<v Speaker 2>let's not forget Deutsche Bank, who've never done anything dodgy. Ever,

0:17:24.160 --> 0:17:27.000
<v Speaker 2>do not google what Deutsche Bank has done with banks

0:17:27.040 --> 0:17:30.320
<v Speaker 2>or particular fellas Deutsche Bank, they were involved in soft

0:17:30.320 --> 0:17:34.000
<v Speaker 2>Bank's Bridge loan and three smaller deals data center in Seoul,

0:17:34.520 --> 0:17:37.480
<v Speaker 2>Cory's twenty twenty four debt couise, November financing, and a

0:17:37.560 --> 0:17:40.080
<v Speaker 2>data center deal in Latin America, as well as a

0:17:40.119 --> 0:17:42.320
<v Speaker 2>six hundred and ten million dollar data center project in

0:17:42.400 --> 0:17:44.840
<v Speaker 2>Virginia and a billion euro of data center project in

0:17:44.880 --> 0:17:49.479
<v Speaker 2>Germany that was invested in within Video BNP Perribus seven deals.

0:17:49.600 --> 0:17:53.199
<v Speaker 2>Cory was delayed draw Tone loan three Stargate New Mexico,

0:17:53.320 --> 0:17:57.040
<v Speaker 2>Stargate Wisconsin and Texas IPI partners of Blue Owl, that

0:17:57.200 --> 0:17:59.920
<v Speaker 2>data center deal in Seoul, and another data center in Chile.

0:18:01.119 --> 0:18:04.000
<v Speaker 2>Morgan Stanley. You remember Morgan Stanley, don't you? While they

0:18:04.000 --> 0:18:07.040
<v Speaker 2>were in eight of these fucking things. Core's October twenty

0:18:07.080 --> 0:18:11.280
<v Speaker 2>twenty four loan loan, three November loans stargeting New Mexico

0:18:11.440 --> 0:18:14.040
<v Speaker 2>Stargate with Advantage. They're thirty eight billion dollar one I

0:18:14.160 --> 0:18:17.960
<v Speaker 2>just talked about, and one with EQT Edge Connects where

0:18:18.000 --> 0:18:20.160
<v Speaker 2>they I don't I don't need to go into detO.

0:18:20.280 --> 0:18:22.320
<v Speaker 2>It's another fucking data center. Oh and they were also

0:18:22.359 --> 0:18:25.880
<v Speaker 2>in SoftBank's Bridge loan, which they needed to to find

0:18:25.920 --> 0:18:28.159
<v Speaker 2>open AI. Now here's another name that you're going to

0:18:28.200 --> 0:18:32.520
<v Speaker 2>hear in the future, SNBC, Sumi Tomo, Mitsui Banking Corporation,

0:18:32.640 --> 0:18:35.720
<v Speaker 2>and I must say the Japanese nyler name of fucking company. Anyway,

0:18:35.960 --> 0:18:39.240
<v Speaker 2>they were in seven deals, all notable cor Weeve's DDTL

0:18:39.320 --> 0:18:42.399
<v Speaker 2>three point zero and November Financing loans, Stargate, New Mexico,

0:18:42.480 --> 0:18:45.960
<v Speaker 2>Stargate Texas, Wisconsin. That's that thirty eight billion dollar one

0:18:46.040 --> 0:18:49.600
<v Speaker 2>A data center in Rowan, Maryland, also involving mug TD

0:18:49.680 --> 0:18:52.359
<v Speaker 2>Securities and HSBC, as well as data center deals in

0:18:52.440 --> 0:18:54.960
<v Speaker 2>Chile and Latin America you've already heard about. Oh and

0:18:55.000 --> 0:18:58.240
<v Speaker 2>guess what Soft Banks Bridge loan which they used to

0:18:58.280 --> 0:19:01.720
<v Speaker 2>invest in open Ai. There are a lot more data

0:19:01.720 --> 0:19:04.480
<v Speaker 2>centered deals than these, but I wanted to tell you

0:19:04.560 --> 0:19:08.520
<v Speaker 2>exactly how centralized they are. I also really need you

0:19:08.560 --> 0:19:11.480
<v Speaker 2>to know how worrying that is, because literally every part

0:19:11.560 --> 0:19:14.800
<v Speaker 2>of this pie is being funded by like eight to

0:19:14.880 --> 0:19:18.800
<v Speaker 2>nine banks. If open ai is raising money for a date, well,

0:19:18.920 --> 0:19:21.520
<v Speaker 2>they wouldn't raise it would be Oracle or core Weave

0:19:21.600 --> 0:19:23.439
<v Speaker 2>or have you raising for a data center. One of

0:19:23.440 --> 0:19:25.840
<v Speaker 2>these people is in it or companies. They're not people.

0:19:26.800 --> 0:19:30.679
<v Speaker 2>It's really worrying. It's worrying how centralized this is, and

0:19:30.720 --> 0:19:33.800
<v Speaker 2>the largest deals, such as the thirty eight billion Stargate Texas,

0:19:33.880 --> 0:19:37.000
<v Speaker 2>Wisconsin deal or the eighteen billion dollar Stargate New Mexico

0:19:37.080 --> 0:19:40.120
<v Speaker 2>deal both involved Gold and Sachs b and Peter Ribas,

0:19:40.200 --> 0:19:43.800
<v Speaker 2>SMBC and fuf MUFG, and all four of those companies

0:19:43.840 --> 0:19:46.600
<v Speaker 2>have at some point funded core Weave. In fact, everybody

0:19:46.600 --> 0:19:51.680
<v Speaker 2>appears to a funded core Weave at some point. City Bank, Credit, Agricole, Society, General,

0:19:52.000 --> 0:19:55.560
<v Speaker 2>Wells Fargo, Carlile, Blackstone, black Rock, Barkley's, Magnatar, and Jeffrey

0:19:55.600 --> 0:19:58.399
<v Speaker 2>is the name of few, and if you've been watching

0:19:58.560 --> 0:20:02.679
<v Speaker 2>the news recently, I have heard that term term word

0:20:02.880 --> 0:20:06.560
<v Speaker 2>jeffries to refer to the company that invested in first

0:20:06.600 --> 0:20:25.199
<v Speaker 2>brands and their exposure anyway. Of the forty banks and

0:20:25.240 --> 0:20:28.840
<v Speaker 2>financial institutions I research twenty four have at some point

0:20:29.040 --> 0:20:32.879
<v Speaker 2>loan to or organized dead for Corwave. Of those institutions, Blackstone,

0:20:32.920 --> 0:20:36.560
<v Speaker 2>Deutsche Bank, JP, Morgan Chase, Morgan, Stanley, MUFG, and Wells

0:20:36.600 --> 0:20:39.720
<v Speaker 2>Fargo have done so multiple times. Core Weave is a

0:20:39.720 --> 0:20:43.520
<v Speaker 2>deeply unprofitable company, saddled with incredible debt and deteriorating margins.

0:20:43.600 --> 0:20:46.160
<v Speaker 2>With one of the largest clients paying net three sixty

0:20:46.320 --> 0:20:49.480
<v Speaker 2>and as I've said, it is arguably the best financed

0:20:49.560 --> 0:20:51.520
<v Speaker 2>data center company in the world with the best chances

0:20:51.560 --> 0:20:56.760
<v Speaker 2>of survival. What I'm getting at is that most data

0:20:56.800 --> 0:20:58.960
<v Speaker 2>centered deals are likely much worse than the terms that

0:20:59.000 --> 0:21:01.920
<v Speaker 2>corwav faces and are likely financed in a similar way

0:21:02.080 --> 0:21:04.880
<v Speaker 2>whereas where a client is signed for data center capacity

0:21:04.920 --> 0:21:07.400
<v Speaker 2>that doesn't exist, such as when nebuists raised four point

0:21:07.400 --> 0:21:10.199
<v Speaker 2>three billion dollars through a share sale and convertible notes

0:21:10.359 --> 0:21:13.280
<v Speaker 2>read loans to handle its seventeen point four billion dollar

0:21:13.400 --> 0:21:17.720
<v Speaker 2>data center contract with Microsoft and yees what Dobin Sachs

0:21:17.760 --> 0:21:19.960
<v Speaker 2>acted as the lead underwriter on the deal, with assistance

0:21:20.000 --> 0:21:24.000
<v Speaker 2>from Bank of America, City Group, Morgan Stanley, all three

0:21:24.000 --> 0:21:27.439
<v Speaker 2>of which have invested in core Weave. AI. Data centers

0:21:27.480 --> 0:21:29.960
<v Speaker 2>are expensive, required debt due to the massive cost of

0:21:30.000 --> 0:21:32.560
<v Speaker 2>construction in the cost of GPUs, and all take at

0:21:32.640 --> 0:21:35.280
<v Speaker 2>least a year, if not two, to start generating a

0:21:35.359 --> 0:21:38.359
<v Speaker 2>single dollar of revenue, at which point they also begin

0:21:38.400 --> 0:21:40.840
<v Speaker 2>losing money because it seems that renting out GPUs is

0:21:40.840 --> 0:21:44.480
<v Speaker 2>really unprofitable. Didn't think, no, no one, no one fucking

0:21:44.520 --> 0:21:47.679
<v Speaker 2>think to check that one Every single major bank and

0:21:47.720 --> 0:21:50.560
<v Speaker 2>financial institution has piled hundreds of millions, if not billions

0:21:50.560 --> 0:21:53.199
<v Speaker 2>of dollars into building data centers that take forever to

0:21:53.240 --> 0:21:56.120
<v Speaker 2>even start generating money, at which point they only seem

0:21:56.160 --> 0:21:59.680
<v Speaker 2>to lose it. Were still in Video sells GPUs and

0:21:59.720 --> 0:22:02.560
<v Speaker 2>a one near upgrade cycle, meaning that all of those

0:22:02.640 --> 0:22:05.040
<v Speaker 2>data centers being built right now are being filled with

0:22:05.080 --> 0:22:07.440
<v Speaker 2>Blackwell chips, and by the time they turn on and

0:22:07.560 --> 0:22:10.280
<v Speaker 2>Video will be selling its next generation via Ruben chips,

0:22:10.640 --> 0:22:14.680
<v Speaker 2>baking them obsolete. Now now now, now, no, no, no.

0:22:14.920 --> 0:22:18.480
<v Speaker 2>You've probably heard that Verra Ruben, the next GPU from

0:22:18.480 --> 0:22:21.920
<v Speaker 2>in Video, will use the same ranks called Oberon as Blackwell,

0:22:22.000 --> 0:22:23.960
<v Speaker 2>which is true to an extent, but won't be true

0:22:24.000 --> 0:22:26.159
<v Speaker 2>for long as in Video intends to shift to their

0:22:26.280 --> 0:22:28.960
<v Speaker 2>Kuiper racks in twenty twenty seven, hoping to build one

0:22:29.040 --> 0:22:33.720
<v Speaker 2>megawat it rax, which involve entire racks full of power supplies,

0:22:34.000 --> 0:22:36.080
<v Speaker 2>meaning that all of those data centers you see today,

0:22:36.800 --> 0:22:38.840
<v Speaker 2>whenever it is they get built, if they get built,

0:22:39.080 --> 0:22:41.560
<v Speaker 2>will be full of racks incompatible with the next generation

0:22:41.600 --> 0:22:44.439
<v Speaker 2>of GPUs. This will also decrease the value of the

0:22:44.440 --> 0:22:46.720
<v Speaker 2>assets inside the data centers, which will in turn decrease

0:22:46.760 --> 0:22:48.800
<v Speaker 2>the value of the assets held by the firms investing

0:22:49.160 --> 0:22:52.120
<v Speaker 2>Stargate Aplin, the one invested in by JP Morgan, Blue

0:22:52.119 --> 0:22:55.600
<v Speaker 2>Oul and Primary Digital Infrastructure and Society General, the one

0:22:55.600 --> 0:22:57.840
<v Speaker 2>that's been heavily delayed and won't be ready until the

0:22:57.920 --> 0:22:59.840
<v Speaker 2>end of twenty twenty six of the earliest fall to

0:22:59.880 --> 0:23:02.960
<v Speaker 2>the brim to the gills with two year old GB

0:23:03.040 --> 0:23:08.119
<v Speaker 2>two hundred Blackwell racks. Hell yeah, baby woo by the

0:23:08.119 --> 0:23:12.200
<v Speaker 2>beginning of twenty twenty seven, when this shithole eventually opens,

0:23:12.720 --> 0:23:15.959
<v Speaker 2>I should really say potentially opens, because let's be honest,

0:23:16.359 --> 0:23:19.600
<v Speaker 2>I'm not confident Stargate Apolene will be obsolete, as will

0:23:19.600 --> 0:23:22.199
<v Speaker 2>any and all data centers filled with Blackwell GPUs, as

0:23:22.240 --> 0:23:25.399
<v Speaker 2>will any and all data centers being built today. Every

0:23:25.600 --> 0:23:28.320
<v Speaker 2>single one takes one to three years and hundreds of

0:23:28.320 --> 0:23:32.280
<v Speaker 2>millions or billions of dollars to build, but probably raised

0:23:32.280 --> 0:23:35.119
<v Speaker 2>in damp, and every single one faces the same kinds

0:23:35.240 --> 0:23:37.919
<v Speaker 2>of construction delays and better yeah, almost all of them

0:23:37.920 --> 0:23:44.080
<v Speaker 2>will turn on in roughly the same timeframe. All right, Look, folks,

0:23:44.119 --> 0:23:46.119
<v Speaker 2>I gotta admit I don't really get our money works.

0:23:46.320 --> 0:23:50.840
<v Speaker 2>I'm not economist, but I do know that supply and

0:23:50.880 --> 0:23:53.760
<v Speaker 2>demand has an effect on pricing. What do you believe

0:23:53.840 --> 0:23:56.640
<v Speaker 2>happens to the price of renting at Blackwell GPU when

0:23:56.720 --> 0:23:59.760
<v Speaker 2>all of these data centers come on? Do you think

0:24:00.240 --> 0:24:02.919
<v Speaker 2>it will it will mean that the price goes up

0:24:03.000 --> 0:24:06.720
<v Speaker 2>or down. Also, while we're on the subject, what do

0:24:06.760 --> 0:24:11.320
<v Speaker 2>you think happens if there isn't sufficient demand? And demand

0:24:11.400 --> 0:24:14.240
<v Speaker 2>is a real question, by the way. Right now, open

0:24:14.240 --> 0:24:16.359
<v Speaker 2>ai makes up a large chunk of the global sale

0:24:16.359 --> 0:24:18.840
<v Speaker 2>of compute, at least eight point six seven billion dollars

0:24:18.880 --> 0:24:21.800
<v Speaker 2>of a zero revenue which is Microsoft's cloud platform through

0:24:21.840 --> 0:24:25.560
<v Speaker 2>September twenty twenty five, and they're part of about twenty

0:24:25.560 --> 0:24:28.639
<v Speaker 2>two point five billion dollars of Corway's backlog, thirty eight

0:24:28.680 --> 0:24:31.440
<v Speaker 2>billion dollars of Amazon's backlog, and so on and so forth,

0:24:31.720 --> 0:24:34.000
<v Speaker 2>and made based on my reporting from last year, just

0:24:34.040 --> 0:24:36.280
<v Speaker 2>over four point five billion dollars through the end of

0:24:36.280 --> 0:24:39.880
<v Speaker 2>September twenty twenty five. Open ai can't afford to pay anyone,

0:24:40.000 --> 0:24:42.160
<v Speaker 2>and nowhere is that more obvious than when it negotiated

0:24:42.240 --> 0:24:45.199
<v Speaker 2>year long payment terms of core with Otherwise, when you

0:24:45.240 --> 0:24:48.200
<v Speaker 2>remove the contracts signed by Hyperscale as an open Ai,

0:24:48.320 --> 0:24:50.840
<v Speaker 2>which I do not believe has the money to pay anybody.

0:24:51.680 --> 0:24:54.200
<v Speaker 2>Based on my analysis, there was less than a billion

0:24:54.240 --> 0:24:57.280
<v Speaker 2>dollars of AI compute revenue in twenty twenty five, or

0:24:57.359 --> 0:24:59.680
<v Speaker 2>zero point five eighty three percent of the money spent

0:24:59.680 --> 0:25:04.560
<v Speaker 2>on day center credit deals in the US. Is that good?

0:25:05.280 --> 0:25:09.159
<v Speaker 2>Hyperscalar revenue is also immediately questionable with Microsoft's steel with

0:25:09.200 --> 0:25:12.399
<v Speaker 2>Nebeus per their six K filing set to default in

0:25:12.440 --> 0:25:15.040
<v Speaker 2>the event that Nebius cannot provide the capacity it's sold

0:25:15.320 --> 0:25:18.280
<v Speaker 2>of its unfinished Fineland, New Jersey data center, which is

0:25:18.320 --> 0:25:20.720
<v Speaker 2>being built by Data One, a company that's never built

0:25:20.760 --> 0:25:23.119
<v Speaker 2>an AI data center, with a CEO that had his

0:25:23.240 --> 0:25:28.119
<v Speaker 2>LinkedIn location set to United Arab Emirates as a very

0:25:28.160 --> 0:25:31.720
<v Speaker 2>recently and indeed was in Arabic until very recently as well,

0:25:32.160 --> 0:25:34.560
<v Speaker 2>and they have funding from a concrete firm that is

0:25:34.560 --> 0:25:37.320
<v Speaker 2>also a vendor on the project. I also believe that

0:25:37.359 --> 0:25:40.359
<v Speaker 2>Microsoft is setting Nebeus up to fail based on discussions

0:25:40.359 --> 0:25:42.720
<v Speaker 2>with sources direct with direct knowledge of plans for the

0:25:42.760 --> 0:25:45.880
<v Speaker 2>violent New Jersey data center. Nebius has agreed to timelines

0:25:45.880 --> 0:25:48.320
<v Speaker 2>that involve having eighteen thousand video B two hundred and

0:25:48.359 --> 0:25:50.840
<v Speaker 2>B three hundred GPUs up and running by the end

0:25:50.840 --> 0:25:54.320
<v Speaker 2>of January for a total of fifty megawats, with another

0:25:54.359 --> 0:25:57.240
<v Speaker 2>eighteen thousand, b three hundreds due by the end of May.

0:25:57.640 --> 0:25:59.760
<v Speaker 2>On speaking with experts in the field about how viable

0:25:59.800 --> 0:26:01.600
<v Speaker 2>he is, the plans are two laugh and one to

0:26:01.760 --> 0:26:05.080
<v Speaker 2>me to fuck off. If Nebius fails to build the capacity,

0:26:05.119 --> 0:26:07.360
<v Speaker 2>Microsoft can walk away, much like open Ai can walk

0:26:07.400 --> 0:26:09.840
<v Speaker 2>away from Stargate in the event that Oracle fails to

0:26:09.840 --> 0:26:12.280
<v Speaker 2>build it on time, as reported by Anissa Gardisi of

0:26:12.320 --> 0:26:15.200
<v Speaker 2>The Information in April twenty twenty five, and I believe

0:26:15.240 --> 0:26:17.280
<v Speaker 2>that this is the case for literally any data center

0:26:17.320 --> 0:26:20.080
<v Speaker 2>provider that's building a data center for any signed up tenant.

0:26:20.400 --> 0:26:22.920
<v Speaker 2>This is another layer of risk to data center developers

0:26:22.920 --> 0:26:25.680
<v Speaker 2>that nobody bothers to fucking discuss, because everybody loves seeing

0:26:25.720 --> 0:26:29.160
<v Speaker 2>these big, beautiful numbers, Except the numbers might have become

0:26:29.200 --> 0:26:32.879
<v Speaker 2>a little too beautiful for sun. On December seventeenth, the

0:26:32.880 --> 0:26:35.280
<v Speaker 2>Financial Times report that the Blue Ou Capitol had pulled

0:26:35.280 --> 0:26:38.800
<v Speaker 2>out of the ten billion dollar Stargate Michigan data center project,

0:26:39.040 --> 0:26:41.879
<v Speaker 2>citing and I quote concerns about its rising debt and

0:26:41.960 --> 0:26:45.680
<v Speaker 2>artificial intelligence spending to quote the ft again. Blue Oul

0:26:45.680 --> 0:26:48.080
<v Speaker 2>had been discussions with Lender's and Oracle about investing in

0:26:48.080 --> 0:26:50.199
<v Speaker 2>the planned one gig what data center being built to

0:26:50.280 --> 0:26:55.040
<v Speaker 2>serve open AI in Saline Township, Michigan. What debt, you

0:26:55.119 --> 0:26:58.200
<v Speaker 2>may ask? While Blue Owl, formerly the loosest legs in

0:26:58.280 --> 0:27:01.880
<v Speaker 2>data center financing, was in always six hundred million dollar deal,

0:27:02.000 --> 0:27:03.960
<v Speaker 2>the seven hundred and fifty million dollar deal as well

0:27:03.960 --> 0:27:07.360
<v Speaker 2>for its planned Virginia data center with Teresa Technology Parks,

0:27:07.600 --> 0:27:09.879
<v Speaker 2>a four billion dollar corps with have data center project

0:27:09.920 --> 0:27:14.119
<v Speaker 2>in Pennsylvania, Stargate Abilene, Stargate Mexico, met As thirty billion

0:27:14.160 --> 0:27:17.080
<v Speaker 2>dollar hyperiod data centered deal, and a one point three

0:27:17.119 --> 0:27:21.200
<v Speaker 2>billion dollar data centered deal in Australia through Stack Infrastructure.

0:27:21.200 --> 0:27:23.959
<v Speaker 2>You remember I mentioned those like ten minutes ago if

0:27:24.000 --> 0:27:27.040
<v Speaker 2>you if you email me for a fish biscuit, I

0:27:27.040 --> 0:27:29.879
<v Speaker 2>haven't gotten. I'm sorry. Anyway, they owned that company and

0:27:30.480 --> 0:27:33.520
<v Speaker 2>may I mentioned it earlier. Anyways, Let's keep going to

0:27:33.600 --> 0:27:35.800
<v Speaker 2>be clear. Blue Owl pulling out is not the same

0:27:35.840 --> 0:27:39.520
<v Speaker 2>as a regular deal. It's a BBC business development company.

0:27:39.640 --> 0:27:42.479
<v Speaker 2>Then invest both this money and rallies together various banks

0:27:42.480 --> 0:27:47.440
<v Speaker 2>in this case SMBCBMP, Peribus, MUFG, and Gold and Sachs

0:27:47.480 --> 0:27:51.760
<v Speaker 2>all part of Stargate, New Mexico, which makes me wonder

0:27:51.800 --> 0:27:57.199
<v Speaker 2>why it didn't happen. In fact, it makes me very

0:27:57.320 --> 0:28:01.159
<v Speaker 2>much worry about that. Per the Financial time, and I quote,

0:28:01.200 --> 0:28:03.600
<v Speaker 2>the private Capitol group had been has been the primary

0:28:03.640 --> 0:28:06.359
<v Speaker 2>backer for Oracle's largest data center projects in the US,

0:28:06.600 --> 0:28:08.960
<v Speaker 2>investing its own money and raising billions more in debt

0:28:09.000 --> 0:28:11.560
<v Speaker 2>to build the facilities. Blue Ol typically sets up a

0:28:11.560 --> 0:28:14.399
<v Speaker 2>special purpose vehicle which owns the data center and leases

0:28:14.440 --> 0:28:18.199
<v Speaker 2>it to Oracle. Blue Owl is incredibly well connected and

0:28:18.240 --> 0:28:20.920
<v Speaker 2>experienced in putting together these kinds of deals and very

0:28:21.040 --> 0:28:24.080
<v Speaker 2>likely went to many banks, many banks. It's worked with

0:28:24.160 --> 0:28:27.520
<v Speaker 2>banks that have been basically giving them blank checks who

0:28:27.560 --> 0:28:31.280
<v Speaker 2>apparently had and I quote concerns about Blue Owl's rising debt,

0:28:31.920 --> 0:28:35.960
<v Speaker 2>much of which it had issued them. While rumor suggests

0:28:35.960 --> 0:28:38.760
<v Speaker 2>that Blackstone may step in, I need to be clear,

0:28:38.800 --> 0:28:41.680
<v Speaker 2>and I've spoken to numerous people in private equity about this.

0:28:41.680 --> 0:28:45.480
<v Speaker 2>This is not a consumer mortgage. This is not a

0:28:45.520 --> 0:28:48.160
<v Speaker 2>young couple buying a start at home. This is a

0:28:48.320 --> 0:28:52.080
<v Speaker 2>giant deal. If it's ten billion dollars, Blue Owl probably

0:28:52.080 --> 0:28:54.720
<v Speaker 2>would put in two billion themselves. That'd be like, I

0:28:54.720 --> 0:28:58.960
<v Speaker 2>don't know, maybe eight billion dollars of credit. I mean,

0:29:00.000 --> 0:29:02.800
<v Speaker 2>pipping in would require billions of dollars in legal logistics

0:29:02.800 --> 0:29:05.600
<v Speaker 2>and likely Blackstone talking to the very same banks it

0:29:05.720 --> 0:29:10.720
<v Speaker 2>already said no, why are they not doing this? And indeed,

0:29:10.760 --> 0:29:14.480
<v Speaker 2>why are things looking shaky? Well remember that thing about

0:29:14.520 --> 0:29:17.000
<v Speaker 2>how this data cent would be least to Oracle. Well,

0:29:17.040 --> 0:29:20.440
<v Speaker 2>Oracle had a free cash flow of negative thirteen billion

0:29:20.520 --> 0:29:23.280
<v Speaker 2>dollars on revenues of sixteen billion dollars, with its most

0:29:23.360 --> 0:29:27.680
<v Speaker 2>recent earnings only beating as analyst estimates thanks to the

0:29:27.680 --> 0:29:29.840
<v Speaker 2>sale of its two point sixty eight billion dollar steak

0:29:29.880 --> 0:29:33.200
<v Speaker 2>in Ampare. And you remember that soft bank Bridge loan.

0:29:33.240 --> 0:29:35.760
<v Speaker 2>I mentioned that fifteen billion dollar one Yeah, half of

0:29:35.800 --> 0:29:37.720
<v Speaker 2>that went to open AI, half of that went to

0:29:37.720 --> 0:29:41.000
<v Speaker 2>buying Ampare, So that's a one off of them. Oracle's

0:29:41.040 --> 0:29:43.400
<v Speaker 2>date is exploding, with over a billion dollars in interest

0:29:43.400 --> 0:29:46.840
<v Speaker 2>payments in its last quarter alone. It's GPU gross margins

0:29:46.880 --> 0:29:49.880
<v Speaker 2>of fourteen percent, which does not mean profitable. Its latest

0:29:50.040 --> 0:29:53.200
<v Speaker 2>Nvidio GB two hundred GPUs have a negative one hundred

0:29:53.200 --> 0:29:55.520
<v Speaker 2>percent gross margin, and it has two hundred and forty

0:29:55.520 --> 0:29:58.640
<v Speaker 2>eight billion dollars in upcoming data center leases is yet

0:29:58.640 --> 0:30:02.640
<v Speaker 2>to begin, and thus all for the most part to

0:30:02.640 --> 0:30:05.960
<v Speaker 2>handle compute for one customer open Ai, which needs to

0:30:06.040 --> 0:30:10.000
<v Speaker 2>raise one hundred billion dollars and then probably much more

0:30:10.640 --> 0:30:17.160
<v Speaker 2>to survive. Is that good anyway, that's a good point

0:30:17.160 --> 0:30:19.360
<v Speaker 2>to jump off from. Next episode, we're gonna wrap all

0:30:19.440 --> 0:30:22.200
<v Speaker 2>this up. It's been a great time reading this for you,

0:30:22.560 --> 0:30:24.280
<v Speaker 2>and we're gonna talk about how the data center of

0:30:24.280 --> 0:30:35.720
<v Speaker 2>Apocalypse will start and what might come next. Thank you

0:30:35.720 --> 0:30:38.440
<v Speaker 2>for listening to Better Offline. The editor and composer of

0:30:38.480 --> 0:30:41.440
<v Speaker 2>the Better Offline theme song is Matasowski. You can check

0:30:41.480 --> 0:30:44.440
<v Speaker 2>out more of his music and audio projects at Matasowski

0:30:44.440 --> 0:30:47.960
<v Speaker 2>dot com, M A T T O. S O w

0:30:48.280 --> 0:30:51.840
<v Speaker 2>Ski dot com. You can email me at easy at

0:30:51.840 --> 0:30:54.520
<v Speaker 2>Better offline dot com, or visit Better Offline dot com

0:30:54.520 --> 0:30:56.960
<v Speaker 2>to find more podcast links and of course my newsletter.

0:30:57.360 --> 0:30:59.800
<v Speaker 2>I also really recommend you go to chat dot Where's

0:30:59.840 --> 0:31:02.360
<v Speaker 2>your dot am to visit the discord, and go to

0:31:02.400 --> 0:31:05.800
<v Speaker 2>our slash Better Offline to check out I'll Reddit. Thank

0:31:05.800 --> 0:31:09.240
<v Speaker 2>you so much for listening. Better Offline is a production

0:31:09.320 --> 0:31:11.960
<v Speaker 2>of cool Zone Media. For more from cool Zone Media,

0:31:12.080 --> 0:31:15.400
<v Speaker 2>visit our website Coolzonemedia dot com or check us out

0:31:15.440 --> 0:31:18.440
<v Speaker 2>on the iHeartRadio app, Apple Podcasts or wherever you get

0:31:18.440 --> 0:31:19.160
<v Speaker 2>your podcasts