WEBVTT - Monologue: The Updated Pale Horses of the AIpocalypse

0:00:02.160 --> 0:00:07.039
<v Speaker 1>Zone Media. Hello and welcome to this week's Better Offline Monologue.

0:00:07.040 --> 0:00:17.400
<v Speaker 1>I'm your host ed ze Trunk. In August of last year,

0:00:17.440 --> 0:00:20.240
<v Speaker 1>I put out a podcast called How the AI Bubble Bursts,

0:00:20.239 --> 0:00:21.759
<v Speaker 1>where I ran down what I thought would be the

0:00:21.760 --> 0:00:24.440
<v Speaker 1>signs that things were collapsing, my pale horses of the

0:00:24.480 --> 0:00:27.280
<v Speaker 1>AI Apocalypse, if you will, which is something that reads

0:00:27.320 --> 0:00:29.159
<v Speaker 1>really well, but when you say it out loud, not

0:00:29.280 --> 0:00:31.440
<v Speaker 1>so good. The reason I put these together is that

0:00:31.480 --> 0:00:33.920
<v Speaker 1>the bubble is unlikely to have one specific moment where

0:00:33.960 --> 0:00:36.960
<v Speaker 1>things explode. Things like burst earns Is collapse during the

0:00:36.960 --> 0:00:40.559
<v Speaker 1>Great Financial Crisis for such significant moments, in part because

0:00:40.720 --> 0:00:43.600
<v Speaker 1>that was a public company. Mortgages were a massive part

0:00:43.640 --> 0:00:45.800
<v Speaker 1>and are a massive part of the economy, and indeed

0:00:45.800 --> 0:00:48.240
<v Speaker 1>there was billions and indeed I think trillions of dollars

0:00:48.240 --> 0:00:50.159
<v Speaker 1>resting on them, and I don't really think there's a

0:00:50.159 --> 0:00:52.920
<v Speaker 1>situation like that, nor is there one where a public

0:00:52.960 --> 0:00:56.280
<v Speaker 1>company collapses as a result of the AI bubble bursting.

0:00:56.840 --> 0:00:59.320
<v Speaker 1>This bubble is different because it's so thoroughly based on

0:00:59.760 --> 0:01:02.400
<v Speaker 1>that games. I'll end up at some point doing a

0:01:02.400 --> 0:01:04.480
<v Speaker 1>longer episode about this, But the long and short of

0:01:04.480 --> 0:01:06.840
<v Speaker 1>it is that the actual core of the system. The

0:01:06.880 --> 0:01:10.560
<v Speaker 1>real weak point is actually in video. In Vidia's continued

0:01:10.600 --> 0:01:12.880
<v Speaker 1>success comes from the reliable quarter of a quarter or

0:01:12.880 --> 0:01:15.080
<v Speaker 1>a year of a year growth from selling GPUs, the

0:01:15.120 --> 0:01:18.240
<v Speaker 1>graphics processing units that power generally of AI and that

0:01:18.280 --> 0:01:20.480
<v Speaker 1>people take and put inside of servers so that they

0:01:20.520 --> 0:01:24.000
<v Speaker 1>can immediately start losing money. More than forty percent of

0:01:24.040 --> 0:01:26.640
<v Speaker 1>their GPU revenue comes from the other six companies in

0:01:26.680 --> 0:01:31.080
<v Speaker 1>the Magnificent Seven, Microsoft, Google, Meta, Tesla, Apple, and Amazon,

0:01:31.400 --> 0:01:34.240
<v Speaker 1>and then VideA. Deeply depends on their continued and growing

0:01:34.360 --> 0:01:37.200
<v Speaker 1>hunger for GPUs. It isn't just enough that they buy

0:01:37.240 --> 0:01:40.520
<v Speaker 1>the same amounts, they must buy more. The rest of

0:01:40.560 --> 0:01:43.319
<v Speaker 1>the mag seven's valuations are dependent on continued growth too,

0:01:43.360 --> 0:01:45.640
<v Speaker 1>which is what AI is meant to provide them. But

0:01:45.760 --> 0:01:50.080
<v Speaker 1>it's not really providing growth at all. Indeed, their growth

0:01:50.120 --> 0:01:53.320
<v Speaker 1>stories are not really based on numbers but vibes. They're

0:01:53.360 --> 0:01:57.240
<v Speaker 1>based on feelings and the general warm, fuzzy feeling the

0:01:57.280 --> 0:01:59.720
<v Speaker 1>dipgits get when they claim that AI is the future.

0:02:00.080 --> 0:02:02.360
<v Speaker 1>Let me give you an example. Metas stock jump because

0:02:02.360 --> 0:02:05.440
<v Speaker 1>they've announced the theoretical data center the size of Manhattan

0:02:05.880 --> 0:02:08.760
<v Speaker 1>now important detail here as well. Meta is making exactly

0:02:08.840 --> 0:02:13.799
<v Speaker 1>minus dollars on generative AI. They're just putting generative AI everywhere,

0:02:14.280 --> 0:02:16.880
<v Speaker 1>and yeah, giving people a little bit of psychosis. I

0:02:16.880 --> 0:02:20.160
<v Speaker 1>imagine filling feeds full of slop. It's very bad. And

0:02:20.200 --> 0:02:23.760
<v Speaker 1>you know what, it also isn't making them money. And

0:02:24.040 --> 0:02:26.600
<v Speaker 1>right now this is really good for the Magnificent seven

0:02:26.639 --> 0:02:28.840
<v Speaker 1>because they're not actually having to prove that AI is

0:02:28.840 --> 0:02:31.280
<v Speaker 1>a big deal. People are basically believing it is because

0:02:31.280 --> 0:02:33.880
<v Speaker 1>the AI trade, which means the value of these stocks,

0:02:33.880 --> 0:02:37.320
<v Speaker 1>remains positive, and honestly, it is currently a positive trade,

0:02:37.360 --> 0:02:41.240
<v Speaker 1>even though it's based on effectively nothing. Seriously, Amazon may

0:02:41.280 --> 0:02:44.760
<v Speaker 1>only make five billion dollars in revenue this year on AI,

0:02:44.840 --> 0:02:48.120
<v Speaker 1>and I mean revenue, not profit. Googled maybe three to

0:02:48.200 --> 0:02:51.640
<v Speaker 1>seven billion, Microsoft thirteen billion dollars, and ten billion dollars

0:02:51.680 --> 0:02:54.840
<v Speaker 1>of that is open AI's compute. These are pathetic numbers.

0:02:54.880 --> 0:02:57.400
<v Speaker 1>But these companies are let off because the AI trade

0:02:57.440 --> 0:03:01.519
<v Speaker 1>still works. The possibilities are still theres that don't exist

0:03:01.560 --> 0:03:04.160
<v Speaker 1>today and don't appear to have any path to them either.

0:03:04.360 --> 0:03:08.560
<v Speaker 1>They're coming, we swear, if we say agents enough time,

0:03:08.680 --> 0:03:11.600
<v Speaker 1>maybe it'll work. The long and short of it is

0:03:11.639 --> 0:03:14.560
<v Speaker 1>that everything about the AI industry is based on some level,

0:03:14.560 --> 0:03:17.400
<v Speaker 1>on lies, on lies both subtle and avert, and the

0:03:17.440 --> 0:03:20.000
<v Speaker 1>general sense that generative AI is both the future and

0:03:20.040 --> 0:03:22.440
<v Speaker 1>the future of the economy. There are a few real

0:03:22.520 --> 0:03:25.160
<v Speaker 1>tangible things to point at other than chat GPT's five

0:03:25.240 --> 0:03:27.480
<v Speaker 1>hundred million weekly users and the tens of billions of

0:03:27.520 --> 0:03:31.720
<v Speaker 1>dollars invested in AI startups, or maybe the weird acquisition

0:03:31.840 --> 0:03:34.400
<v Speaker 1>deals like the two point four billion dollars that Google

0:03:34.520 --> 0:03:38.920
<v Speaker 1>just gave kind of AI code editor Windsurf, except Google

0:03:39.040 --> 0:03:42.240
<v Speaker 1>took the staff, some of the staff the C suite,

0:03:42.440 --> 0:03:44.600
<v Speaker 1>the R and D team for the two point four billion,

0:03:44.640 --> 0:03:48.440
<v Speaker 1>and then Cognition, who makes the Devin coding pop took

0:03:48.480 --> 0:03:51.960
<v Speaker 1>the rest for an undisclosed sum. It's all really, really

0:03:51.960 --> 0:03:55.040
<v Speaker 1>fucking weird. It's all very strange, and yet when you

0:03:55.080 --> 0:03:58.720
<v Speaker 1>look beneath the surface, there really are few returns. Windsurfs

0:03:58.720 --> 0:04:01.960
<v Speaker 1>monthly revenue was six point eight brillion, three million dollars

0:04:02.000 --> 0:04:04.560
<v Speaker 1>a month, a ridiculously small amount, working out to about

0:04:04.560 --> 0:04:07.520
<v Speaker 1>eighty two million dollars of annualized revenue. Yet based on

0:04:07.560 --> 0:04:10.440
<v Speaker 1>the Information's Generative AI data base, Windsurf was actually one

0:04:10.480 --> 0:04:13.240
<v Speaker 1>of the highest earning AI startups out there no really

0:04:13.280 --> 0:04:16.800
<v Speaker 1>based on annualized revenue meaning month multiplied by twelve, most

0:04:16.800 --> 0:04:20.599
<v Speaker 1>Ai companies make about one hundred million dollars annualized or less.

0:04:20.720 --> 0:04:23.159
<v Speaker 1>Not really, that's about eight point three million dollars a

0:04:23.240 --> 0:04:27.240
<v Speaker 1>year a month with since there's the Weka Xai together

0:04:27.320 --> 0:04:30.440
<v Speaker 1>a bridge Glean perplexity, and they're all making about one

0:04:30.480 --> 0:04:33.560
<v Speaker 1>hundred million dollars annualized. And I must be clear about

0:04:33.600 --> 0:04:36.240
<v Speaker 1>how weird this status. Annialize just means the month of

0:04:36.279 --> 0:04:41.760
<v Speaker 1>revenue times twelve. Customers churn. Customers churn, so your arr

0:04:41.920 --> 0:04:44.159
<v Speaker 1>may go down when you have a few people drop out,

0:04:44.320 --> 0:04:48.400
<v Speaker 1>it may spike arbitrarily because you've increased prices, and it's

0:04:48.440 --> 0:04:51.040
<v Speaker 1>just it's a very bad statistic to use. Yet people

0:04:51.640 --> 0:04:53.359
<v Speaker 1>are not really looking at the other thing, which is

0:04:54.040 --> 0:04:56.080
<v Speaker 1>that's not a lot of money. Eight point three million

0:04:56.279 --> 0:04:59.360
<v Speaker 1>dollars a month. You have ad agencies that make several

0:04:59.400 --> 0:05:03.239
<v Speaker 1>times that. Ad agency's bank ads on Facebook and Google.

0:05:04.040 --> 0:05:08.159
<v Speaker 1>These are terrible fucking companies. I mean outside of one

0:05:08.240 --> 0:05:11.320
<v Speaker 1>hundred million ARR side, you've got mid Journey, who are

0:05:11.440 --> 0:05:14.760
<v Speaker 1>inexplicably profitable. We'll look into that, and Neo four j

0:05:14.960 --> 0:05:17.479
<v Speaker 1>and they make two hundred million dollars a month, sorry,

0:05:17.560 --> 0:05:20.279
<v Speaker 1>two one hundred million dollars annualized, so about sixteen point

0:05:20.320 --> 0:05:24.320
<v Speaker 1>six million dollars, which is again not amazing. None of

0:05:24.320 --> 0:05:27.800
<v Speaker 1>these are amazing. These aren't great. These would be okay

0:05:27.839 --> 0:05:31.039
<v Speaker 1>startups if they were profitable, and they're absolutely not. But

0:05:31.160 --> 0:05:33.960
<v Speaker 1>right on the tippy top behind Anthropic and open AI

0:05:34.160 --> 0:05:36.400
<v Speaker 1>is any sphere who makes Cursor and they make five

0:05:36.480 --> 0:05:39.160
<v Speaker 1>hundred million dollars annualized, or about forty one point six

0:05:39.240 --> 0:05:43.760
<v Speaker 1>million dollars a month, which is bad because in June

0:05:44.440 --> 0:05:48.000
<v Speaker 1>all of their prices got increased by Anthropic, and on

0:05:48.080 --> 0:05:50.479
<v Speaker 1>top of that they had to change the pricing and

0:05:50.520 --> 0:05:53.080
<v Speaker 1>now Cursor kind of sucks. Go on their subredd it

0:05:53.200 --> 0:05:56.200
<v Speaker 1>check it out anyway. This is why you don't like

0:05:56.200 --> 0:05:58.160
<v Speaker 1>annualized as a stat by the way, because I reckon

0:05:58.200 --> 0:06:01.000
<v Speaker 1>their July revenue might be a little bit lower than

0:06:01.040 --> 0:06:03.160
<v Speaker 1>forty one point six million dollars, but we'll never know

0:06:03.320 --> 0:06:06.240
<v Speaker 1>because well unless someone reports there. If you've got any

0:06:06.360 --> 0:06:09.200
<v Speaker 1>numbers on any of these companies, my signal is eas

0:06:09.279 --> 0:06:14.839
<v Speaker 1>it tron ezit ron dot seventy six on signal. Please

0:06:14.880 --> 0:06:18.440
<v Speaker 1>send me anything you have. I love numbers anyway, past

0:06:18.480 --> 0:06:21.160
<v Speaker 1>that point, it's only anthropic and open AI. And I've

0:06:21.240 --> 0:06:23.480
<v Speaker 1>left out a few companies like Touring and Scale they

0:06:23.480 --> 0:06:26.560
<v Speaker 1>don't make AI products their consultancies. But the long and

0:06:26.560 --> 0:06:29.279
<v Speaker 1>short of it is, this industry doesn't make much fucking money.

0:06:29.520 --> 0:06:32.000
<v Speaker 1>The actual returns from it aren't that very good. It's

0:06:32.000 --> 0:06:34.800
<v Speaker 1>not very good at all. But the reason I'm telling

0:06:34.839 --> 0:06:36.400
<v Speaker 1>you all these numbers is that I need you to

0:06:36.440 --> 0:06:38.200
<v Speaker 1>know that this industry is not held up by the

0:06:38.279 --> 0:06:42.000
<v Speaker 1>actual businesses or their revenues. None of these companies are profitable,

0:06:42.120 --> 0:06:45.520
<v Speaker 1>and it certainly isn't held up by the actual products themselves.

0:06:46.160 --> 0:06:48.720
<v Speaker 1>And I'm sure you've got an AI friend who's come

0:06:48.760 --> 0:06:50.080
<v Speaker 1>to you and said, oh, I use it for this.

0:06:51.160 --> 0:06:53.479
<v Speaker 1>See how far you get them in the conversation before

0:06:53.480 --> 0:06:55.840
<v Speaker 1>they just say they're using it like a search engine.

0:06:56.080 --> 0:06:58.880
<v Speaker 1>We had this whole thing with Brian on the podcast

0:06:58.920 --> 0:07:01.760
<v Speaker 1>this week. It's like, yeah, really emphatically talking about the

0:07:01.760 --> 0:07:04.039
<v Speaker 1>fact that you've got a slightly sexier search engine that

0:07:04.040 --> 0:07:06.280
<v Speaker 1>may or may not get things right, that can fool

0:07:06.320 --> 0:07:09.279
<v Speaker 1>you into believing you're correct about something which isn't a

0:07:09.320 --> 0:07:12.160
<v Speaker 1>fucking product. Now, if it was, people will be making money.

0:07:12.160 --> 0:07:16.800
<v Speaker 1>It's ridiculous. But even in areas like coding, coding the

0:07:16.840 --> 0:07:19.600
<v Speaker 1>one place where people can point and go look look

0:07:19.760 --> 0:07:23.560
<v Speaker 1>users like this, look edward, look at them. A study

0:07:23.600 --> 0:07:26.560
<v Speaker 1>from nonprofit group Model Evaluation and Throat Research found that

0:07:26.600 --> 0:07:29.800
<v Speaker 1>despite developers believing AI coding tools sped them up by

0:07:29.800 --> 0:07:33.120
<v Speaker 1>twenty four percent, coding tools actually increase the completion time

0:07:33.120 --> 0:07:36.600
<v Speaker 1>of their tasks by nineteen percent. It actually it slowed

0:07:36.640 --> 0:07:40.160
<v Speaker 1>them down. Oh my god. Every week with this industry,

0:07:40.200 --> 0:07:42.920
<v Speaker 1>I feel like I'm going insane. It's like every week

0:07:42.960 --> 0:07:46.160
<v Speaker 1>there's this goddamn story like this. I still have people

0:07:46.160 --> 0:07:48.680
<v Speaker 1>who tell me it's the future. I feel like I'm

0:07:48.720 --> 0:07:52.120
<v Speaker 1>going insane. But as a result of AI being a

0:07:52.160 --> 0:07:55.040
<v Speaker 1>nearly entirely vibe spaced economy, the pale horses of its

0:07:55.080 --> 0:07:58.600
<v Speaker 1>apocalypse will be vibe shifters rather than defined events that

0:07:58.640 --> 0:08:02.000
<v Speaker 1>would logically change things. Remember Deep Seek, that was a

0:08:02.040 --> 0:08:05.200
<v Speaker 1>clunky story. So you had a Chinese model developer that

0:08:05.320 --> 0:08:08.200
<v Speaker 1>may be cheaper, it was cheaper to train, much cheaper

0:08:08.240 --> 0:08:11.600
<v Speaker 1>to train, but still cheaper to train, and their model

0:08:11.680 --> 0:08:14.920
<v Speaker 1>might be cheaper, and it is comparable. I mean, I've

0:08:14.960 --> 0:08:17.160
<v Speaker 1>certainly made my hay with the episodes and interviews that

0:08:17.200 --> 0:08:19.760
<v Speaker 1>did around it, but I surprise that shocked the market.

0:08:19.840 --> 0:08:22.840
<v Speaker 1>There are bigger, nastiest stories. But if that can do

0:08:22.960 --> 0:08:27.560
<v Speaker 1>it well, basically anything could one thing that really spreads

0:08:27.640 --> 0:08:31.200
<v Speaker 1>like information poison into the heads of business idiot's everywhere

0:08:31.200 --> 0:08:34.360
<v Speaker 1>who are saying AI now please without really understanding what

0:08:34.480 --> 0:08:37.160
<v Speaker 1>it does. It's unlikely to be one big pop but

0:08:37.280 --> 0:08:40.600
<v Speaker 1>several little doots from the bubble deflating. So let's get

0:08:40.640 --> 0:08:42.160
<v Speaker 1>to it. Here are the science to look out for,

0:08:42.480 --> 0:08:44.920
<v Speaker 1>freshly updated, and if you have any ideas, please do

0:08:45.040 --> 0:08:47.319
<v Speaker 1>reach out. I'd love to hear from you. We'll start

0:08:47.320 --> 0:08:50.240
<v Speaker 1>with an obvious one. Any price increases or decreases are

0:08:50.240 --> 0:08:53.920
<v Speaker 1>open AI and Anthropic. I called these increases in August,

0:08:53.960 --> 0:08:56.120
<v Speaker 1>and they had decreases because, as I've mentioned in the

0:08:56.120 --> 0:08:58.600
<v Speaker 1>subprime AI crisis, there is now a race to the

0:08:58.600 --> 0:09:01.520
<v Speaker 1>boond for model developers. Open Ai cut the price of

0:09:01.559 --> 0:09:03.880
<v Speaker 1>their own through Reasoning model by eight percent in June

0:09:03.880 --> 0:09:06.520
<v Speaker 1>in an attempt to undercut Anthropics claud for Opus, which

0:09:06.600 --> 0:09:11.000
<v Speaker 1>launched May twenty second, I believe, and we've already begun

0:09:11.040 --> 0:09:14.240
<v Speaker 1>to see price increases too. Anthropic added service theres and

0:09:14.280 --> 0:09:17.240
<v Speaker 1>open ai priority processing for enterprise customers to give them

0:09:17.320 --> 0:09:20.160
<v Speaker 1>uninterrupted service in June. And I must be clear when

0:09:20.200 --> 0:09:22.760
<v Speaker 1>I say enterprise, I mean any AI company that has

0:09:22.840 --> 0:09:27.360
<v Speaker 1>customers at scale. So any Sphere, Cursor, Glean and the like,

0:09:27.440 --> 0:09:30.440
<v Speaker 1>they are enterprise customers and they are very likely having

0:09:30.440 --> 0:09:33.960
<v Speaker 1>to pay these prices. Now here's another one. Anthropic and

0:09:34.000 --> 0:09:37.720
<v Speaker 1>open Ai may move away from monthly subscriptions, and indeed

0:09:37.800 --> 0:09:41.280
<v Speaker 1>ANYAI company may do this. This is a huge payer

0:09:41.400 --> 0:09:45.640
<v Speaker 1>horse when companies start moving away from simple business models,

0:09:45.679 --> 0:09:48.679
<v Speaker 1>so twenty bucks one hundred bucks a month, or they

0:09:48.679 --> 0:09:51.720
<v Speaker 1>start changing those so the returns on a sort of

0:09:51.720 --> 0:09:55.160
<v Speaker 1>returns the actual usership isn't obvious. Isn't obvious how much

0:09:55.200 --> 0:09:57.440
<v Speaker 1>you can or can't use. That's a sign that the

0:09:57.440 --> 0:10:00.160
<v Speaker 1>company is trying to contain costs and indeed fuck with

0:10:00.200 --> 0:10:01.960
<v Speaker 1>their customers a bit. And we've already seen a little

0:10:01.960 --> 0:10:04.520
<v Speaker 1>bit of this with how Cursor changed their pricing or

0:10:04.640 --> 0:10:06.960
<v Speaker 1>Claude Max the one hundred dollar or two hundred dollars

0:10:06.960 --> 0:10:10.160
<v Speaker 1>a month subscription from Anthropic, which has now changed things

0:10:10.160 --> 0:10:12.839
<v Speaker 1>to say you get more usage at the higher tiers

0:10:12.880 --> 0:10:15.720
<v Speaker 1>than the twenty dollars a month pro subscription. How much

0:10:15.760 --> 0:10:19.040
<v Speaker 1>more it isn't clear. Go to Claude subreadit. You'll see

0:10:19.040 --> 0:10:21.160
<v Speaker 1>a lot of people who are constantly saying, I don't

0:10:21.200 --> 0:10:24.000
<v Speaker 1>know what the rate limits are, and that's working as

0:10:24.040 --> 0:10:27.560
<v Speaker 1>intended by the way Anthropic is tweaking things very clearly

0:10:27.600 --> 0:10:29.760
<v Speaker 1>on the back end, we do not know what they are.

0:10:29.800 --> 0:10:32.760
<v Speaker 1>There are rumors that they are quantizing models, so making

0:10:32.840 --> 0:10:36.960
<v Speaker 1>them smaller, dumber, cheaper to run, but still doing comparable things,

0:10:37.360 --> 0:10:40.679
<v Speaker 1>or just not having as much access within peak hours.

0:10:40.800 --> 0:10:43.679
<v Speaker 1>These are unsubstantiated rumors, but I've seen enough claims of

0:10:43.720 --> 0:10:46.840
<v Speaker 1>them that they're worth mentioning. These are signs that things

0:10:46.920 --> 0:10:49.800
<v Speaker 1>are getting bad, or that they need to make things

0:10:49.840 --> 0:10:53.840
<v Speaker 1>more profitable. And really, any and all cost increases on

0:10:53.880 --> 0:10:56.640
<v Speaker 1>consumers by any AI service are a pay or horse,

0:10:56.640 --> 0:10:58.640
<v Speaker 1>as they suggest the company in question is trying to

0:10:58.679 --> 0:11:02.240
<v Speaker 1>deal with these ruinous costs. Remember, there are no profitable

0:11:02.240 --> 0:11:05.480
<v Speaker 1>AI companies. It's fucking crazy. We are three years into

0:11:05.520 --> 0:11:10.079
<v Speaker 1>this dog shit and there's none of them. Okay, surge AI.

0:11:10.559 --> 0:11:14.240
<v Speaker 1>They are an AI data training company. That does not matter.

0:11:14.679 --> 0:11:16.480
<v Speaker 1>It does not matter. If you've got a car that

0:11:16.480 --> 0:11:19.079
<v Speaker 1>has a horrible fuel economy. You don't point out how

0:11:19.120 --> 0:11:22.400
<v Speaker 1>well oil is selling to say that the car is good.

0:11:22.800 --> 0:11:25.920
<v Speaker 1>The different things anyway, Any and all rate limits that

0:11:26.040 --> 0:11:29.800
<v Speaker 1>any AI startup are a huge pay or horse because

0:11:29.840 --> 0:11:32.240
<v Speaker 1>these companies are spending a shit ton of money and

0:11:32.280 --> 0:11:35.839
<v Speaker 1>they're having to run these massive infrastructural operations to keep

0:11:35.880 --> 0:11:40.120
<v Speaker 1>things going. Those things are expensive. Power in particular is expensive,

0:11:40.160 --> 0:11:43.640
<v Speaker 1>so any mediating those costs is necessary. If you see

0:11:43.679 --> 0:11:46.320
<v Speaker 1>it with other companies, it's even worse because it means

0:11:46.360 --> 0:11:49.600
<v Speaker 1>that open ai and Anthropic are turning the screws on them. Now.

0:11:49.600 --> 0:11:52.280
<v Speaker 1>Another obvious one, of course, is any funding of revenue

0:11:52.280 --> 0:11:55.920
<v Speaker 1>problems from the remaining AI software companies Glean, Harvey, Alpha Sense,

0:11:55.960 --> 0:11:58.559
<v Speaker 1>any sphere who makes Cursor as I mentioned, or replet

0:11:58.640 --> 0:12:01.240
<v Speaker 1>for example. It's just the funding required to keep these

0:12:01.240 --> 0:12:03.760
<v Speaker 1>companies going is running out, and their business models aren't working.

0:12:03.920 --> 0:12:06.040
<v Speaker 1>And I want to make one real clear point here.

0:12:06.040 --> 0:12:09.800
<v Speaker 1>I'm kind of repeating myself, but it's necessary. These companies

0:12:09.880 --> 0:12:14.040
<v Speaker 1>are symbolic of whether generative AI will work. I realized

0:12:14.120 --> 0:12:16.480
<v Speaker 1>chet GPT is the big sexy one. But if you

0:12:16.600 --> 0:12:19.720
<v Speaker 1>can't build software on GENERATIVEAI and sell it for a profit,

0:12:20.000 --> 0:12:24.120
<v Speaker 1>there is no industry here. Every single major cloud computing

0:12:24.200 --> 0:12:28.680
<v Speaker 1>revolution has started because there is profit, and fucking hell,

0:12:28.760 --> 0:12:31.680
<v Speaker 1>it's nothing like Amazon Web Services when it grew. I'm

0:12:31.760 --> 0:12:33.960
<v Speaker 1>tired of hearing this point. I think I have to

0:12:34.000 --> 0:12:36.720
<v Speaker 1>do a thing about it because I'm fucking sick of it.

0:12:36.720 --> 0:12:40.360
<v Speaker 1>It's nothing to do with it, nothing, It's not remotely similar.

0:12:40.440 --> 0:12:43.320
<v Speaker 1>The only similar thing is that goddamn data centers. God

0:12:44.760 --> 0:12:49.160
<v Speaker 1>calming down. But on the subject to data centers, any

0:12:49.240 --> 0:12:53.000
<v Speaker 1>delays to Stargate or Amazon and Anthropics new Carlisle, Indiana

0:12:53.080 --> 0:12:56.880
<v Speaker 1>build out, both Anthropic and OpenAI are running into capacity issues.

0:12:56.920 --> 0:12:59.640
<v Speaker 1>This is documented well both by the companies and reporting

0:12:59.679 --> 0:13:02.760
<v Speaker 1>from outs like the Information. And they need these projects

0:13:02.840 --> 0:13:06.600
<v Speaker 1>to happen because right now it's very obvious Microsoft is

0:13:06.600 --> 0:13:09.800
<v Speaker 1>not going to expand their infrastructure for Open AI and

0:13:09.840 --> 0:13:12.720
<v Speaker 1>Anthropic well, they need Amazon to do it. But data

0:13:12.760 --> 0:13:15.680
<v Speaker 1>centers don't grow like weeds. They take forever to build.

0:13:16.000 --> 0:13:18.560
<v Speaker 1>I don't even know when the Carlysle Indiana data center

0:13:18.640 --> 0:13:21.720
<v Speaker 1>is going to get built. Fucking hell, what a mess.

0:13:21.920 --> 0:13:24.000
<v Speaker 1>Every reason I must think about this, it's such a mess,

0:13:24.280 --> 0:13:26.400
<v Speaker 1>all right. One final pale horse, and this is my

0:13:26.480 --> 0:13:29.839
<v Speaker 1>favorite one. Anything that happens with soft Bank and their

0:13:29.880 --> 0:13:32.960
<v Speaker 1>ability to raise money is a big pale horse the

0:13:33.000 --> 0:13:35.720
<v Speaker 1>paylist of them all. Actually, because soft Bank is critical

0:13:35.760 --> 0:13:38.839
<v Speaker 1>to everything. Soft Bank is the money behind open Ai.

0:13:39.440 --> 0:13:42.520
<v Speaker 1>Soft Bank is the money behind Stargate. Soft Bank is

0:13:42.800 --> 0:13:46.920
<v Speaker 1>financially responsible for Stargate, by the way, And if open

0:13:46.960 --> 0:13:51.480
<v Speaker 1>Ai doesn't move in to Stargate, and indeed if Stargate

0:13:51.520 --> 0:13:54.640
<v Speaker 1>never happens, Oracle gets left with forty billion dollars of

0:13:54.679 --> 0:13:56.960
<v Speaker 1>chips unused and Cruso, who they have to pay a

0:13:57.000 --> 0:14:00.000
<v Speaker 1>billion dollars I think, have a fifteen year lease with them.

0:14:00.400 --> 0:14:03.200
<v Speaker 1>What a mess. But as I've said, the actual bubble

0:14:03.200 --> 0:14:05.920
<v Speaker 1>bursting will be slow, annoying and unsatisfying at times. But

0:14:05.960 --> 0:14:08.400
<v Speaker 1>these pale horses are the signs that things are falling apart.

0:14:08.600 --> 0:14:10.000
<v Speaker 1>I will, as ever keep you up to date on

0:14:10.040 --> 0:14:12.920
<v Speaker 1>when these signs are happening, what they mean, why you

0:14:12.960 --> 0:14:15.760
<v Speaker 1>should care. I'll probably say fucking shit a few times

0:14:15.760 --> 0:14:18.600
<v Speaker 1>when I do, so, really appreciate you, love you all,

0:14:18.679 --> 0:14:19.480
<v Speaker 1>Thanks for listening.