WEBVTT - Monologue: OpenAI Could Kill SoftBank and Oracle

0:00:03.200 --> 0:00:07.200
<v Speaker 1>Media. What's up? It's better offline and I'm your host

0:00:07.400 --> 0:00:22.320
<v Speaker 1>ed Zetron It's your better offline monologue for the week, folks.

0:00:22.360 --> 0:00:24.200
<v Speaker 1>Has been a long few months, and I've taken the

0:00:24.280 --> 0:00:27.560
<v Speaker 1>vast majority of this week after recovery, after writing about

0:00:27.840 --> 0:00:30.480
<v Speaker 1>h two hundred thousand words in the last few months,

0:00:30.520 --> 0:00:32.800
<v Speaker 1>not even including my upcoming book, The Hater's Guide to

0:00:32.800 --> 0:00:35.479
<v Speaker 1>Silicon Valley, coming out in twenty twenty seven. If you

0:00:35.520 --> 0:00:37.400
<v Speaker 1>want to support my work, I'm currently doing a week

0:00:37.440 --> 0:00:39.800
<v Speaker 1>long sale for the premium newsletter ten bucks off the

0:00:39.840 --> 0:00:43.199
<v Speaker 1>annual rate for life. That'll be in the notes. Please do.

0:00:43.440 --> 0:00:45.920
<v Speaker 1>If you feel like supporting me, do that. New's learned.

0:00:45.960 --> 0:00:49.200
<v Speaker 1>The podcast are now my by principal source of income.

0:00:49.280 --> 0:00:51.720
<v Speaker 1>It's happened. The era of smiles has begun. Join me

0:00:51.760 --> 0:00:54.560
<v Speaker 1>and receive the bounties of words every week except this one,

0:00:54.560 --> 0:00:57.240
<v Speaker 1>and I take it the week off. So let's begin,

0:00:57.320 --> 0:00:59.240
<v Speaker 1>And I wanted to start by saying how you the

0:00:59.280 --> 0:01:02.280
<v Speaker 1>listener might help me. First of all, my signal is

0:01:02.320 --> 0:01:05.360
<v Speaker 1>Ezitron dot seventy six. The things I'm looking for right

0:01:05.400 --> 0:01:08.160
<v Speaker 1>now that would help me are anthropics S one or

0:01:08.200 --> 0:01:11.160
<v Speaker 1>any and all documentation of its cloud spend and financial condition.

0:01:11.360 --> 0:01:13.920
<v Speaker 1>Especially if you can get me something from Amazon Web

0:01:14.000 --> 0:01:16.480
<v Speaker 1>Services or Google Cloud that shows me how much this

0:01:16.520 --> 0:01:19.200
<v Speaker 1>company is spending. That'd be very very helpful to me,

0:01:19.800 --> 0:01:22.560
<v Speaker 1>as would any and all information about the AI revenues

0:01:22.600 --> 0:01:25.800
<v Speaker 1>of any major tech company that aren't public knowledge, and

0:01:25.880 --> 0:01:28.319
<v Speaker 1>any and all information about the cloud spend of open Ai,

0:01:28.440 --> 0:01:31.840
<v Speaker 1>Anthropic Perplexity, or any other major AI startup, and of

0:01:31.880 --> 0:01:34.360
<v Speaker 1>course any of the financials of any AI startup. You

0:01:34.400 --> 0:01:36.880
<v Speaker 1>can get me the same ghosts with token spend at

0:01:36.920 --> 0:01:41.680
<v Speaker 1>any corporation. Please use signal, Please take photos of any documents.

0:01:41.720 --> 0:01:44.080
<v Speaker 1>Do not send me PDFs. Do not send me documents

0:01:44.080 --> 0:01:46.400
<v Speaker 1>for your safety, m for mine, and please please please

0:01:46.440 --> 0:01:48.520
<v Speaker 1>do not send me stuff that's already public. This isn't

0:01:48.560 --> 0:01:51.240
<v Speaker 1>because I'm chiding you. I'm just saying people get excited

0:01:51.280 --> 0:01:52.800
<v Speaker 1>that they want to help and they send me stuff

0:01:52.800 --> 0:01:56.160
<v Speaker 1>that's already out there. I've probably already seen it. I'm

0:01:56.200 --> 0:01:59.080
<v Speaker 1>a freak. I'm looking at this stuff all day. All right, folks,

0:01:59.120 --> 0:02:02.400
<v Speaker 1>let's roll to the actar monologue. So earlier in the week,

0:02:02.440 --> 0:02:04.480
<v Speaker 1>I put out a fifteen thousand word piece called the

0:02:04.480 --> 0:02:07.360
<v Speaker 1>Open Ai Bubble, a massive guide to how open ai

0:02:07.480 --> 0:02:11.200
<v Speaker 1>is the cause, enabler, and justification for the entire AI bubble.

0:02:11.200 --> 0:02:14.240
<v Speaker 1>Writ large and how everything kind of collapses when it dies.

0:02:15.160 --> 0:02:16.919
<v Speaker 1>I wrote it because I've heard a lot of people

0:02:16.960 --> 0:02:19.680
<v Speaker 1>say it's an open ai bubble, not an AI bubble

0:02:19.680 --> 0:02:21.639
<v Speaker 1>over the last year, as a way of waving off

0:02:21.680 --> 0:02:25.160
<v Speaker 1>potential detractors suggesting that open ai could collapse and everything

0:02:25.160 --> 0:02:28.000
<v Speaker 1>else would be left behind him. Fine, you're living at

0:02:28.000 --> 0:02:30.000
<v Speaker 1>a bubble yourself if you think this way, that's a

0:02:30.080 --> 0:02:32.720
<v Speaker 1>crazy way to live one's life. And please, please please

0:02:32.760 --> 0:02:34.919
<v Speaker 1>stop emailing me about the bailout. I already I've gone

0:02:34.919 --> 0:02:38.079
<v Speaker 1>over the summits anyway. In truth, open ai is the

0:02:38.160 --> 0:02:41.200
<v Speaker 1>largest consumer of AI compute accounts for over eight hundred

0:02:41.200 --> 0:02:45.080
<v Speaker 1>billion dollars of remaining performance obligations across hyperscalars and neo clouds.

0:02:45.200 --> 0:02:48.000
<v Speaker 1>The people that rent how ai GPUs and basically just

0:02:48.040 --> 0:02:51.320
<v Speaker 1>exist a funnel Nvidia money, and likely over one hundred

0:02:51.360 --> 0:02:55.120
<v Speaker 1>billion dollars the future revenues of Broadcom and Nvidia themselves.

0:02:55.720 --> 0:02:57.520
<v Speaker 1>I go into detail about it in the piece, but

0:02:57.560 --> 0:02:59.720
<v Speaker 1>I kind of wanted to go over it a little

0:02:59.720 --> 0:03:03.280
<v Speaker 1>today in today's monologue two. Open ai is also a

0:03:03.320 --> 0:03:06.440
<v Speaker 1>material risk to both soft Bank and Oracle. If open

0:03:06.480 --> 0:03:08.760
<v Speaker 1>Ai can't go public, soft Bank is going to face

0:03:08.760 --> 0:03:12.400
<v Speaker 1>a massive liquidity crisis. It has forty five point nine

0:03:12.480 --> 0:03:15.280
<v Speaker 1>billion dollars in debt and matures within the next twelve months,

0:03:15.360 --> 0:03:17.720
<v Speaker 1>and they'll either have to refinance it or pay it.

0:03:18.440 --> 0:03:20.560
<v Speaker 1>Except that's going to be much more difficult because they've

0:03:20.600 --> 0:03:24.079
<v Speaker 1>jettisoned most of its valuable stocks in t mobile and

0:03:24.120 --> 0:03:26.400
<v Speaker 1>in video, all while pushing up against the limits of

0:03:26.400 --> 0:03:28.360
<v Speaker 1>what it can borrow against its holdings in the chip

0:03:28.360 --> 0:03:31.760
<v Speaker 1>maker arm which it bought I think think twenty twenty

0:03:31.800 --> 0:03:34.000
<v Speaker 1>three for thirty two billion dollars. It went public at

0:03:34.040 --> 0:03:36.280
<v Speaker 1>fifty billion dollars, but the holdings are worth one hundreds

0:03:36.320 --> 0:03:40.800
<v Speaker 1>of billions of dollars. Except that's kind of a problem,

0:03:40.920 --> 0:03:43.000
<v Speaker 1>which I'll get to in a minute. But soft Bank

0:03:43.040 --> 0:03:46.120
<v Speaker 1>will never truly go bankrupt. That doesn't mean it's not

0:03:46.200 --> 0:03:48.720
<v Speaker 1>a problem if open ai fails, because it will be

0:03:48.760 --> 0:03:51.400
<v Speaker 1>put in a historic crisis, one worse than the dot

0:03:51.440 --> 0:03:55.680
<v Speaker 1>com bubble, if where Masayoshi's done to bet everything like GeoCities,

0:03:55.960 --> 0:03:57.880
<v Speaker 1>and even the horrors of twenty twenty two and twenty

0:03:57.960 --> 0:04:00.400
<v Speaker 1>twenty three when the Vision Funds had lost is of

0:04:00.400 --> 0:04:03.760
<v Speaker 1>what thirty two billion dollars. Back then, soft Bank had

0:04:03.760 --> 0:04:06.160
<v Speaker 1>plenty of other valuable stock and ways to raise debt.

0:04:06.720 --> 0:04:09.080
<v Speaker 1>They had the Ala Barber stock, They had their holdings

0:04:09.120 --> 0:04:12.200
<v Speaker 1>in Betfair and other things like that, except most of

0:04:12.200 --> 0:04:14.520
<v Speaker 1>that is gone now. And now S and P Global

0:04:14.600 --> 0:04:18.000
<v Speaker 1>is downgraded its outlook to negative, which means they think

0:04:18.040 --> 0:04:20.719
<v Speaker 1>they might actually downgrade its credit in the future. And

0:04:20.800 --> 0:04:23.719
<v Speaker 1>just to be clear, they specify that the reason that

0:04:23.839 --> 0:04:27.120
<v Speaker 1>S and P did that was open ai. They said

0:04:27.120 --> 0:04:31.160
<v Speaker 1>the words, they literally said it was the case. Except now,

0:04:31.279 --> 0:04:33.000
<v Speaker 1>what the hell is soft Bank meant to do? It

0:04:33.000 --> 0:04:35.240
<v Speaker 1>doesn't have anything else to flog really, other than it's

0:04:35.360 --> 0:04:38.080
<v Speaker 1>arm stock. And the problem there is that soft Bank

0:04:38.160 --> 0:04:41.360
<v Speaker 1>is the largest holder of armstock in the world, which

0:04:41.400 --> 0:04:44.120
<v Speaker 1>makes it difficult for it to liquidate too much without

0:04:44.120 --> 0:04:46.240
<v Speaker 1>spooking the market. Think of it like this. If the

0:04:46.320 --> 0:04:48.200
<v Speaker 1>largest holder of a stock says, oh, I'm just going

0:04:48.279 --> 0:04:51.440
<v Speaker 1>to start selling it off for liquidity, it officially stops

0:04:51.440 --> 0:04:54.040
<v Speaker 1>being about the stock in question, but the company's selling it.

0:04:54.279 --> 0:04:56.320
<v Speaker 1>The value of arm will turn into a bet on

0:04:56.320 --> 0:04:58.760
<v Speaker 1>whether soft Bank can afford to pay its own bills.

0:04:59.520 --> 0:05:01.719
<v Speaker 1>It's already taking out a bunch of margin loans on

0:05:01.760 --> 0:05:04.120
<v Speaker 1>the stock and further lands will be difficult to raise

0:05:04.160 --> 0:05:07.359
<v Speaker 1>for the very same reason. While it would be likely

0:05:07.440 --> 0:05:10.800
<v Speaker 1>rescued by the Bank of Japan and other local financial institutions,

0:05:10.839 --> 0:05:14.039
<v Speaker 1>there has never been a more dangerous time for Masayoshi's son.

0:05:14.520 --> 0:05:16.800
<v Speaker 1>Check out my hater's guide to SoftBank. I know it's

0:05:16.839 --> 0:05:18.840
<v Speaker 1>a premium thing, but there's a deal going on. It's

0:05:18.839 --> 0:05:21.600
<v Speaker 1>a beautiful deal. I stopped doing the Trump impression now

0:05:21.720 --> 0:05:25.279
<v Speaker 1>and get onto one of his friends, Oracle Now. In

0:05:25.320 --> 0:05:28.880
<v Speaker 1>Oracles case, its entire future rides and open AI's ability

0:05:28.880 --> 0:05:30.880
<v Speaker 1>to pay it about seventy billion dollars a year in

0:05:30.960 --> 0:05:34.159
<v Speaker 1>annual revenue, and that rides on Oracle's ability to build

0:05:34.160 --> 0:05:36.880
<v Speaker 1>about seven point one gigawats or so of data sene

0:05:36.920 --> 0:05:41.960
<v Speaker 1>capacity in the next two or three years. I'm not kidding,

0:05:42.000 --> 0:05:45.719
<v Speaker 1>it's genuinely that's that's what they need to do. Oracle

0:05:45.720 --> 0:05:48.200
<v Speaker 1>and open Ai have a three hundred billion dollar contract

0:05:48.240 --> 0:05:50.920
<v Speaker 1>that's meant to have started on June first, twenty twenty six,

0:05:51.120 --> 0:05:53.640
<v Speaker 1>which said revenue coming from the Stargate data sentence, the

0:05:53.640 --> 0:05:57.600
<v Speaker 1>Oracle is building for Clammy Sammy. That's right, mister Ortmand Clamuel.

0:05:57.960 --> 0:06:01.479
<v Speaker 1>We don't like him. The problem, of course, is the

0:06:01.520 --> 0:06:04.520
<v Speaker 1>seven point one gigawats of capacity needs to be built,

0:06:04.520 --> 0:06:07.560
<v Speaker 1>and barely four hundred megawats of it actually exist. And

0:06:07.920 --> 0:06:10.240
<v Speaker 1>that's in stockg Abilene, a project that broke ground in

0:06:10.240 --> 0:06:12.159
<v Speaker 1>the middle of twenty twenty four and was meant to

0:06:12.160 --> 0:06:14.040
<v Speaker 1>be done either the end of last year or a

0:06:14.160 --> 0:06:17.839
<v Speaker 1>couple of weeks ago. And everything I hear from the

0:06:17.880 --> 0:06:20.640
<v Speaker 1>people inside Oracle is that they keep firing all the

0:06:20.680 --> 0:06:23.920
<v Speaker 1>people who know other fucking business works. It's a terrible

0:06:23.960 --> 0:06:29.080
<v Speaker 1>thing for them. Oh and another thing is that Oracle

0:06:29.120 --> 0:06:31.359
<v Speaker 1>literally warned in its anal report that it may or

0:06:31.480 --> 0:06:33.719
<v Speaker 1>may not get paid by open AI, and that said

0:06:33.800 --> 0:06:36.279
<v Speaker 1>nonpayment would potentially also mean that it couldn't leasse the

0:06:36.320 --> 0:06:39.279
<v Speaker 1>GPU capacity to anyone else, as in no one else

0:06:39.320 --> 0:06:41.360
<v Speaker 1>really exists. They could buy it, and if they did

0:06:41.400 --> 0:06:43.760
<v Speaker 1>manage to sell it to them be a massive loss.

0:06:44.600 --> 0:06:47.599
<v Speaker 1>Is that good? Also? Tell me is it good that

0:06:47.720 --> 0:06:52.440
<v Speaker 1>S and P Global also downgraded Oracle just one step

0:06:52.880 --> 0:06:56.279
<v Speaker 1>above junk grade? They're barely investment grade credit wise for

0:06:56.360 --> 0:07:00.240
<v Speaker 1>now for all, an angel status aren't incoming. Maybe on

0:07:00.240 --> 0:07:02.760
<v Speaker 1>that means another monologue you don't really care. Let's get

0:07:02.800 --> 0:07:07.320
<v Speaker 1>to the funny stuff though. Larry Ellison, CEO of sorry

0:07:07.440 --> 0:07:10.200
<v Speaker 1>Chairman of Oracle CEO is currently two different mister Bean

0:07:10.200 --> 0:07:13.240
<v Speaker 1>type characters. He's also used three hundred and forty six

0:07:13.280 --> 0:07:15.920
<v Speaker 1>million of his one point one six billion Oracle shares

0:07:15.960 --> 0:07:18.760
<v Speaker 1>for an estimated twenty one billion dollars of margin loans

0:07:18.800 --> 0:07:21.360
<v Speaker 1>personal ones, just to be clear. Otherwise, he has about

0:07:21.400 --> 0:07:23.800
<v Speaker 1>ten billion dollars in cash and fifteen billion dollars in

0:07:23.840 --> 0:07:26.680
<v Speaker 1>Tesla shares. Why do you care about that? Well, this

0:07:26.720 --> 0:07:29.320
<v Speaker 1>is material because he also has to contra up about

0:07:29.320 --> 0:07:32.760
<v Speaker 1>twenty billion dollars in cash actual cash to push through

0:07:32.760 --> 0:07:35.920
<v Speaker 1>the paramount Warner Brothers deal by September thirtieth. Otherwise, he

0:07:35.960 --> 0:07:38.440
<v Speaker 1>gets find about five hundred and six hundred million dollars

0:07:38.480 --> 0:07:41.520
<v Speaker 1>a quarter, and to do so, regardless of whether it happens,

0:07:42.040 --> 0:07:44.360
<v Speaker 1>whether it's stopped by then, or whether it's meant to

0:07:44.400 --> 0:07:46.720
<v Speaker 1>go through, he's likely going to have to take out

0:07:46.800 --> 0:07:49.920
<v Speaker 1>further margin loans on his Oracle stock, And depending on

0:07:49.920 --> 0:07:52.040
<v Speaker 1>how said stock is doing, getting that loan might be

0:07:52.080 --> 0:07:55.160
<v Speaker 1>quite problematic. I think had just said problematic, but I'm

0:07:55.160 --> 0:08:12.280
<v Speaker 1>just going to keep on going. So to explain A

0:08:12.360 --> 0:08:14.680
<v Speaker 1>margin loan works by offering up a certain amount of

0:08:14.680 --> 0:08:17.080
<v Speaker 1>shares is collateral a little like a mortgage. So if

0:08:17.120 --> 0:08:19.520
<v Speaker 1>you are taken out a twenty five percent loan to

0:08:19.600 --> 0:08:22.600
<v Speaker 1>value loan for one billion dollars, you'd offer up about

0:08:22.640 --> 0:08:24.840
<v Speaker 1>two hundred and fifty million dollars in stock, maybe a

0:08:24.840 --> 0:08:27.640
<v Speaker 1>little more depending on how volatile it is. When the

0:08:27.680 --> 0:08:30.120
<v Speaker 1>value of said stock drops below a certain amount, you

0:08:30.200 --> 0:08:32.839
<v Speaker 1>get forced to do something called a margin core, while

0:08:32.840 --> 0:08:35.120
<v Speaker 1>a margin call happens to our main, where you're forced

0:08:35.120 --> 0:08:37.640
<v Speaker 1>to either proffer up more cash to cover the shortfall, or,

0:08:37.679 --> 0:08:40.880
<v Speaker 1>far more likely, in Ellison's case, give up more Oracle stock.

0:08:41.960 --> 0:08:44.840
<v Speaker 1>The problem that Ellison faces is the very same one

0:08:44.880 --> 0:08:48.000
<v Speaker 1>that SoftBank has with Arm. He owns forty percent of

0:08:48.040 --> 0:08:51.439
<v Speaker 1>Oracle shares, which means that liquidating them would be very difficult.

0:08:51.559 --> 0:08:55.120
<v Speaker 1>If he started selling them off, the market would say, oh, Christ,

0:08:55.160 --> 0:08:58.240
<v Speaker 1>is he going to sell his forty percent holdings in Oracle? Sell?

0:08:58.400 --> 0:09:01.320
<v Speaker 1>Sell it all? The Piss index is crashing. And to

0:09:01.360 --> 0:09:04.200
<v Speaker 1>make that as worse, he's collateralized over fifty billion dollars

0:09:04.240 --> 0:09:06.720
<v Speaker 1>worth of Oracle shares the ones I mentioned earlier, which

0:09:06.720 --> 0:09:09.439
<v Speaker 1>means that if Oracle stock drops below, say I don't

0:09:09.480 --> 0:09:13.160
<v Speaker 1>know sixty dollars a share, He'll be put in a

0:09:13.200 --> 0:09:15.839
<v Speaker 1>vicious cycle where he'll either have to hand over more

0:09:15.880 --> 0:09:18.720
<v Speaker 1>and more shares to cover the shortfall, or worse, still

0:09:19.360 --> 0:09:22.120
<v Speaker 1>have to sell them to keep up. Hey, hey, stop laughing,

0:09:22.400 --> 0:09:25.560
<v Speaker 1>stop it, stop laughing. Mister Ellison is in real trouble.

0:09:25.640 --> 0:09:28.600
<v Speaker 1>Larry Ellison could be in a real pickle. Stop it.

0:09:29.200 --> 0:09:33.240
<v Speaker 1>I just kid and fuck Larry Ellison anyway. The worst

0:09:33.320 --> 0:09:36.200
<v Speaker 1>part is that open ai doesn't even have to collapse

0:09:36.240 --> 0:09:39.400
<v Speaker 1>for this to happen. The AI bubble isn't based on

0:09:39.400 --> 0:09:44.520
<v Speaker 1>anyone's actual revenues, productivity or ROI. No Hyperscala actually reports

0:09:44.520 --> 0:09:48.400
<v Speaker 1>their AI revenues, Oracle kind of bundles them within OCI

0:09:48.520 --> 0:09:52.800
<v Speaker 1>their infrastructure bucket, and outside of open Ai and Anthropic

0:09:52.880 --> 0:09:56.040
<v Speaker 1>the industry is actually incredibly small revenue wise, very loud

0:09:56.080 --> 0:09:59.640
<v Speaker 1>and annoying, though the markets have conflated akapex bubble as

0:09:59.679 --> 0:10:02.480
<v Speaker 1>in GPUs and the surrounding gear for data centers with

0:10:02.559 --> 0:10:05.040
<v Speaker 1>a thriving AI industry, which is a nice way of

0:10:05.080 --> 0:10:08.600
<v Speaker 1>saying that the current slate of prices for Microsoft, Google, Amazon,

0:10:08.640 --> 0:10:12.000
<v Speaker 1>Meta and Oracle and their current AI focused bumps are

0:10:12.000 --> 0:10:16.439
<v Speaker 1>based on vibes, and mania vibes can break mania subsides.

0:10:17.160 --> 0:10:19.520
<v Speaker 1>All it takes for things to go pair shaped is

0:10:19.520 --> 0:10:23.240
<v Speaker 1>for the market to believe, to become convinced that Oracle

0:10:23.280 --> 0:10:26.079
<v Speaker 1>won't get paid by open Ai for them to start

0:10:26.160 --> 0:10:29.000
<v Speaker 1>dumping the stock. All they have to believe is that

0:10:29.120 --> 0:10:32.880
<v Speaker 1>all of that capex, all of that negative cash flow

0:10:34.040 --> 0:10:37.160
<v Speaker 1>is just for nothing, is for a company that either

0:10:37.200 --> 0:10:40.520
<v Speaker 1>won't exist or won't be able to actually afford all

0:10:40.520 --> 0:10:42.280
<v Speaker 1>of that data center capacity. Every time I say this

0:10:42.320 --> 0:10:46.320
<v Speaker 1>out loud, I feel crazy. The biggest myth though about

0:10:46.360 --> 0:10:49.640
<v Speaker 1>Oracle is that it's this healthy, profitable business and that

0:10:49.720 --> 0:10:52.880
<v Speaker 1>this AI thing is just yet more good stuff that

0:10:52.880 --> 0:10:56.480
<v Speaker 1>they've added on top, when in fact, every other business

0:10:56.520 --> 0:11:00.280
<v Speaker 1>segment other than renting out aigpus is either plat code

0:11:00.440 --> 0:11:04.440
<v Speaker 1>or in active decline, meaning that Oracle's entire future rides

0:11:04.480 --> 0:11:07.200
<v Speaker 1>and whether it can monetize its fin doom relationship with

0:11:07.280 --> 0:11:13.400
<v Speaker 1>Nvidia and Jensen Wong and really Oracle's story is fundamentally insane.

0:11:13.480 --> 0:11:17.079
<v Speaker 1>Right now, It's spending three hundred and forty billion dollars

0:11:17.160 --> 0:11:20.280
<v Speaker 1>or more on AI data center capacity for one client

0:11:20.559 --> 0:11:23.160
<v Speaker 1>Open AI in the hopes that open Ai can bring

0:11:23.200 --> 0:11:25.840
<v Speaker 1>it seventy billion dollars a year in revenue for a

0:11:25.840 --> 0:11:28.440
<v Speaker 1>contract that began a month ago, except most of the

0:11:28.480 --> 0:11:31.719
<v Speaker 1>capacity is barely under construction, and because data centers take

0:11:31.760 --> 0:11:34.440
<v Speaker 1>eighteen to thirty six months to build, it's unlikely that

0:11:34.480 --> 0:11:36.880
<v Speaker 1>more than a single gigawatt will be ready before the

0:11:36.880 --> 0:11:40.440
<v Speaker 1>beginning of twenty twenty eight, if at all, building this

0:11:40.480 --> 0:11:43.600
<v Speaker 1>capacity is costly. An Oracle's cash flow has been negative

0:11:43.640 --> 0:11:46.360
<v Speaker 1>twenty billion dollars or worse for the last two quarters,

0:11:46.400 --> 0:11:49.480
<v Speaker 1>with plans to spend another ninety billion dollars in the

0:11:49.520 --> 0:11:53.800
<v Speaker 1>next fiscal year, which just begun in June. The twenty

0:11:53.800 --> 0:11:55.920
<v Speaker 1>two thousand people that it laid off in the past

0:11:55.960 --> 0:11:59.320
<v Speaker 1>six months were based on discussions with many sources load

0:11:59.320 --> 0:12:02.160
<v Speaker 1>bearing parts of it's infrastructure, throwing parts of the business

0:12:02.160 --> 0:12:05.920
<v Speaker 1>into active chaos and genuinely causing problems with renewals for

0:12:05.960 --> 0:12:10.280
<v Speaker 1>its Oracle database products that are the cornerstone of how

0:12:10.920 --> 0:12:14.720
<v Speaker 1>Larry Ellison's heart beats. I don't know what to say.

0:12:14.960 --> 0:12:17.439
<v Speaker 1>I don't know. I am just a guy. Why am

0:12:17.520 --> 0:12:20.360
<v Speaker 1>I the one that found this? It's sitting in broad daylight.

0:12:20.440 --> 0:12:23.880
<v Speaker 1>Maybe people just don't do mathematics. I don't know, but

0:12:23.960 --> 0:12:28.200
<v Speaker 1>this is this is quite bad. And the point I'm making, though,

0:12:28.280 --> 0:12:30.720
<v Speaker 1>just to simplify, is that for me to be wrong,

0:12:31.240 --> 0:12:33.720
<v Speaker 1>open ai will have to bring in hundreds of billions

0:12:33.720 --> 0:12:37.120
<v Speaker 1>of dollars a year in revenue for themselves by twenty three,

0:12:37.600 --> 0:12:41.720
<v Speaker 1>which will require it to become either profitable massively profitable

0:12:41.720 --> 0:12:43.760
<v Speaker 1>too like not just like a little bit of a

0:12:43.800 --> 0:12:47.640
<v Speaker 1>bit the profitability I mean twenty thirty forty billion dollars

0:12:47.679 --> 0:12:51.760
<v Speaker 1>a year plus or just continually raise money. I'm talking

0:12:51.880 --> 0:12:54.840
<v Speaker 1>hundreds of billions of dollars a year, and no, no

0:12:54.960 --> 0:12:57.960
<v Speaker 1>government bailout will cover the shortfall. The government's talking about

0:12:58.000 --> 0:13:00.720
<v Speaker 1>five percent share be forty two billion dollars. That won't

0:13:00.760 --> 0:13:03.400
<v Speaker 1>even cover their fucking influence costs if they lasted twenty

0:13:03.440 --> 0:13:08.880
<v Speaker 1>twenty seven. It's a joke. Oh and by the way,

0:13:09.480 --> 0:13:12.160
<v Speaker 1>even if open ai does that, Oracle will have to

0:13:12.280 --> 0:13:15.959
<v Speaker 1>complete the single most ambitious construction project of all time,

0:13:16.280 --> 0:13:20.120
<v Speaker 1>building enough power to power multiple cities, erecting giant data

0:13:20.160 --> 0:13:23.600
<v Speaker 1>centers in multiple states, make them fully operational with far

0:13:23.679 --> 0:13:26.000
<v Speaker 1>less people than the shortage of both talent and the

0:13:26.000 --> 0:13:29.559
<v Speaker 1>core materials like electrical grade steel and also do so

0:13:29.679 --> 0:13:31.520
<v Speaker 1>on a timeline that doesn't put it in breach of

0:13:31.559 --> 0:13:36.079
<v Speaker 1>contract with open AI. Would it shock you if I

0:13:36.200 --> 0:13:39.680
<v Speaker 1>told you that this is considered radical thinking? Would you

0:13:39.760 --> 0:13:42.880
<v Speaker 1>be surprised if I told you that many journalists simply

0:13:42.920 --> 0:13:45.120
<v Speaker 1>shrug when you tell them this, and they go, they'll

0:13:45.120 --> 0:13:48.080
<v Speaker 1>work it out, they'll work it out, It'll all work out,

0:13:48.360 --> 0:13:51.040
<v Speaker 1>it'll work out. Don't worry about it. Ed, you're crazy

0:13:51.080 --> 0:13:53.600
<v Speaker 1>head Ed. Why are you outside my office? Why are

0:13:53.600 --> 0:13:55.439
<v Speaker 1>you dressed as a dug? Why do you keep saying

0:13:55.720 --> 0:13:59.600
<v Speaker 1>I'm the deterioration dug. Quack quack quack, the market's going

0:13:59.640 --> 0:14:01.719
<v Speaker 1>to hell? And the answer is that's a bit I'm

0:14:01.760 --> 0:14:04.760
<v Speaker 1>working on that I'll go to later. In any case,

0:14:04.800 --> 0:14:07.160
<v Speaker 1>I am kidding, and I do need to say. The

0:14:07.200 --> 0:14:10.240
<v Speaker 1>future really is one strewn with chaos and carcasses, and

0:14:10.240 --> 0:14:12.760
<v Speaker 1>I'm not sure the world is actually ready for it,

0:14:12.800 --> 0:14:15.400
<v Speaker 1>thanks to a captured media industry that refuses to do

0:14:15.480 --> 0:14:18.959
<v Speaker 1>the messy work of reconciling with the projections that range

0:14:19.000 --> 0:14:23.600
<v Speaker 1>from unrealistic to fantastical. When the time comes, people are

0:14:23.640 --> 0:14:25.760
<v Speaker 1>going to be asking why we didn't see this coming?

0:14:25.880 --> 0:14:28.520
<v Speaker 1>And as I've said before, it's because nobody wanted to

0:14:28.640 --> 0:14:33.080
<v Speaker 1>fucking look anyway, Thanks as ever for listening. Things are

0:14:33.080 --> 0:14:35.360
<v Speaker 1>going to get real interesting over the next few months.

0:14:35.440 --> 0:14:37.960
<v Speaker 1>I just got me a Bloomberg terminal, and it's a

0:14:38.000 --> 0:14:40.640
<v Speaker 1>source of unfathomable power that I believe will take my

0:14:40.720 --> 0:14:42.920
<v Speaker 1>coverage to the next level. If you're on there, shoot

0:14:42.960 --> 0:14:45.560
<v Speaker 1>me a message, shoot me a goot on Ploop. That's

0:14:45.560 --> 0:14:49.280
<v Speaker 1>another social network I'm on. And I really do, genuinely,

0:14:49.360 --> 0:14:52.760
<v Speaker 1>deeply and meaningfully appreciate the many wonderful messages and emails

0:14:52.800 --> 0:14:55.040
<v Speaker 1>I receive every day, and I'm so grateful to have

0:14:55.080 --> 0:14:58.200
<v Speaker 1>such incredible fans and followers. You're all amazing, like I

0:14:58.240 --> 0:15:00.360
<v Speaker 1>hear from so many of you, and you're very thoughtful

0:15:00.360 --> 0:15:02.320
<v Speaker 1>and smart, and you pick this stuff up very quickly.

0:15:02.920 --> 0:15:05.640
<v Speaker 1>I love you all, and I assume your bellow we

0:15:05.680 --> 0:15:08.640
<v Speaker 1>love you chef in response, See you next week.