WEBVTT - The AI Industry Is Losing

0:00:02.800 --> 0:00:03.560
<v Speaker 1>Z Media.

0:00:05.120 --> 0:00:08.039
<v Speaker 2>Hello on, Welcome to Better Offline. I'm your host ed

0:00:08.119 --> 0:00:23.120
<v Speaker 2>zetron and I am not a car. Yesterday, a business

0:00:23.160 --> 0:00:26.400
<v Speaker 2>inside of report that Uber COO Andrew McDonald had said,

0:00:26.440 --> 0:00:29.400
<v Speaker 2>and I quote, that its AI costs were becoming harder

0:00:29.440 --> 0:00:32.400
<v Speaker 2>to justify and that the link was not there between

0:00:32.440 --> 0:00:35.919
<v Speaker 2>spending money on AI tokens and creating more useful features.

0:00:36.800 --> 0:00:37.000
<v Speaker 3>Yeah.

0:00:37.200 --> 0:00:39.480
<v Speaker 2>Just gonna throw a bastball through a hoop real quick.

0:00:39.720 --> 0:00:42.920
<v Speaker 2>After three long years of hammering it home, I have

0:00:43.040 --> 0:00:46.519
<v Speaker 2>finally been proven right. AI's outputs and efficacy do not

0:00:46.640 --> 0:00:49.920
<v Speaker 2>match up with its ruinous costs. When organizations have to

0:00:49.920 --> 0:00:54.400
<v Speaker 2>pay the actual token costs of AI versus using subsidized subscriptions,

0:00:54.560 --> 0:00:56.959
<v Speaker 2>they're forced to measure the actual return on investment from

0:00:57.000 --> 0:01:00.920
<v Speaker 2>AI and are immediately balking at the results. Wheeling for mercy.

0:01:01.000 --> 0:01:03.760
<v Speaker 2>They're saying, honey, I can afford it now. To give

0:01:03.800 --> 0:01:07.400
<v Speaker 2>you some context, Anthropic only moved organizations to token based

0:01:07.400 --> 0:01:11.319
<v Speaker 2>billing sometime in Q one, twenty twenty six. This is

0:01:11.400 --> 0:01:13.720
<v Speaker 2>at most four months of having to pay the true

0:01:13.760 --> 0:01:16.800
<v Speaker 2>costs of their AI token burn and they're already squealing.

0:01:16.800 --> 0:01:19.640
<v Speaker 2>They're already begging for mercy. They're already saying, sir, no

0:01:19.760 --> 0:01:23.400
<v Speaker 2>more tokens. AI has a revenue ceiling and an economic

0:01:23.440 --> 0:01:26.640
<v Speaker 2>mismatch with its customer base. It's time to accept it.

0:01:28.360 --> 0:01:31.240
<v Speaker 2>Every time you've heard somebody say that AI is real,

0:01:31.480 --> 0:01:35.240
<v Speaker 2>it's here, and it's transformative, you've heard from somebody paying

0:01:35.240 --> 0:01:38.200
<v Speaker 2>a monthly subscription to a service that allows its customers

0:01:38.200 --> 0:01:40.920
<v Speaker 2>to burn anywhere from three dollars the thirteen dollars worth

0:01:40.959 --> 0:01:44.080
<v Speaker 2>of tokens for every dollar of their subscription. Even get

0:01:44.160 --> 0:01:47.559
<v Speaker 2>hub Copilot, which paid the model providers directly, was letting

0:01:47.600 --> 0:01:50.040
<v Speaker 2>people burn on a thirty nine dollar a month subscription

0:01:50.200 --> 0:01:54.200
<v Speaker 2>thirteen hundred to six thousand dollars in a month. Every

0:01:54.280 --> 0:01:58.040
<v Speaker 2>effervescent booster and captured business idiot editor crowing about the

0:01:58.040 --> 0:02:01.120
<v Speaker 2>power of AI has done so without ever really facing

0:02:01.120 --> 0:02:05.040
<v Speaker 2>its real costs or i think, even using it very much. However,

0:02:05.200 --> 0:02:07.920
<v Speaker 2>useful lms may seem to them as a facade for

0:02:07.960 --> 0:02:10.360
<v Speaker 2>a product that costs far too much money for outputs

0:02:10.480 --> 0:02:13.120
<v Speaker 2>that may or may not actually result in something functional

0:02:13.200 --> 0:02:17.240
<v Speaker 2>or helpful. When you're not paying for tokens, these mistakes

0:02:17.280 --> 0:02:21.520
<v Speaker 2>are easy to ignore. These subscriptions musk the ugly truth

0:02:21.560 --> 0:02:24.680
<v Speaker 2>of AI that you're paying on a per million token basis,

0:02:24.680 --> 0:02:26.799
<v Speaker 2>regardless of whether you get what you want, or even

0:02:26.840 --> 0:02:29.120
<v Speaker 2>if the model makes mistakes or creates more problems that

0:02:29.160 --> 0:02:32.240
<v Speaker 2>you then have to spend more tokens to fix. It's

0:02:32.280 --> 0:02:35.240
<v Speaker 2>a scam, it's a con. It never made sense. An

0:02:35.320 --> 0:02:38.400
<v Speaker 2>uber COO has given everybody permission to talk about the

0:02:38.400 --> 0:02:41.600
<v Speaker 2>inherent economic mismatch of AI, and also revealed that AI

0:02:41.680 --> 0:02:44.480
<v Speaker 2>companies like Anthropic and open Ai have a ceiling far

0:02:44.560 --> 0:02:47.919
<v Speaker 2>below that which they need to justify their valuations. These

0:02:47.960 --> 0:02:50.880
<v Speaker 2>companies need to be making two hundred billion dollars a

0:02:50.960 --> 0:02:53.640
<v Speaker 2>year by twenty thirty or they cannot keep up with

0:02:53.680 --> 0:02:58.600
<v Speaker 2>their own costs. Now, anthropics rapid revenue growth is the

0:02:58.639 --> 0:03:01.520
<v Speaker 2>result of companies spending millillions or tens or even hundreds

0:03:01.560 --> 0:03:04.800
<v Speaker 2>of millions of dollars on its tokens. For its revenue

0:03:04.840 --> 0:03:08.600
<v Speaker 2>story to make sense, this revenue would have to be stable, replicable,

0:03:08.600 --> 0:03:12.799
<v Speaker 2>and sustainable. Instead, it appears that organizations have been burning

0:03:12.840 --> 0:03:16.160
<v Speaker 2>tokens without any real understanding of why, other than that

0:03:16.280 --> 0:03:19.680
<v Speaker 2>the need to do AI and LLLM coding is the future,

0:03:19.840 --> 0:03:22.880
<v Speaker 2>and all of their other dickhead business idiot friends are saying, oh,

0:03:22.919 --> 0:03:25.120
<v Speaker 2>I got, i'm let and everyone spend ten million dollars

0:03:25.160 --> 0:03:26.960
<v Speaker 2>a month on this shit because I have a fucking

0:03:27.000 --> 0:03:30.840
<v Speaker 2>piece of rebar in my skull. For this to make sense,

0:03:30.919 --> 0:03:34.280
<v Speaker 2>the majority of organizations would have to sustain and grow

0:03:34.560 --> 0:03:39.000
<v Speaker 2>a massive spend on AI tokens from anthropic and open AI. Instead,

0:03:39.000 --> 0:03:41.920
<v Speaker 2>it appears that this token bone was inherently experimental and

0:03:42.120 --> 0:03:45.120
<v Speaker 2>entirely disconnected from messy things like I don't know return

0:03:45.160 --> 0:03:48.720
<v Speaker 2>on investment. Uber already noted back in April that it

0:03:48.760 --> 0:03:51.440
<v Speaker 2>blew through its annual token budget in a few months,

0:03:51.440 --> 0:03:54.440
<v Speaker 2>and that conversation reported by Laura Bretton of the Information

0:03:54.560 --> 0:03:56.920
<v Speaker 2>for End of the Show clearly led to an internal

0:03:56.960 --> 0:03:58.840
<v Speaker 2>back and forth that will end in it cutting its

0:03:58.840 --> 0:04:02.440
<v Speaker 2>spend on AI models. I've now spoken to three different

0:04:02.480 --> 0:04:05.800
<v Speaker 2>large organizations that are all echoing similar anxieties about the

0:04:05.960 --> 0:04:10.360
<v Speaker 2>RII of AI. Nobody can actually tell why they're spending

0:04:10.400 --> 0:04:13.280
<v Speaker 2>this much money. Things aren't getting shipped fast, that software

0:04:13.320 --> 0:04:15.800
<v Speaker 2>isn't better, and the only people that seem excited about

0:04:15.840 --> 0:04:18.880
<v Speaker 2>it a business idy it's disconnected from production, and even

0:04:19.240 --> 0:04:22.279
<v Speaker 2>they are becoming cost conscious when faced with millions of

0:04:22.320 --> 0:04:26.719
<v Speaker 2>dollars of token bills anthropic and open ai cannot afford

0:04:26.760 --> 0:04:29.080
<v Speaker 2>for things to slow down, as they've both signed up

0:04:29.120 --> 0:04:33.840
<v Speaker 2>to over a trillion dollars of compute commitments across Google, Microsoft, Amazon,

0:04:33.960 --> 0:04:38.000
<v Speaker 2>Core Weave, Oracle and Cerebrus. Just to be clear as well,

0:04:38.160 --> 0:04:40.240
<v Speaker 2>open ai has to be making two hundred and eighty

0:04:40.240 --> 0:04:42.919
<v Speaker 2>four billion dollars by twenty thirty, and they need to

0:04:42.920 --> 0:04:46.440
<v Speaker 2>be profitable at that point too, otherwise Oracle cannot afford

0:04:46.520 --> 0:04:50.560
<v Speaker 2>its own bills. This is not hyperbole. This is quoting

0:04:50.680 --> 0:04:56.159
<v Speaker 2>back open AI's own projected revenues from their investor dex.

0:04:58.160 --> 0:05:00.599
<v Speaker 2>But in reality, it appears there's a limit to which

0:05:00.680 --> 0:05:03.520
<v Speaker 2>organizations can be abused and manipulated into believing that the

0:05:03.520 --> 0:05:05.880
<v Speaker 2>future is here. And that limit is when they paid

0:05:05.920 --> 0:05:08.440
<v Speaker 2>millions of dollars a month for something that doesn't appear

0:05:08.480 --> 0:05:12.120
<v Speaker 2>to have a measurable return on investment. My friends, the

0:05:12.160 --> 0:05:14.920
<v Speaker 2>business idiots are losing because they never had a plan

0:05:15.000 --> 0:05:19.360
<v Speaker 2>to begin with. As I've said before, Ai is only

0:05:19.440 --> 0:05:22.360
<v Speaker 2>as real as its subsidies. Chat GPT is only free

0:05:22.400 --> 0:05:24.720
<v Speaker 2>to hundreds of billions of people because open ai is

0:05:24.760 --> 0:05:27.480
<v Speaker 2>able to raise hundreds of billions of dollars have Microsoft

0:05:27.520 --> 0:05:29.880
<v Speaker 2>and Google and Amazon and Core. We've built their own

0:05:29.920 --> 0:05:34.040
<v Speaker 2>infrastructure and basically do everything for them, much like Anthropic

0:05:34.160 --> 0:05:37.080
<v Speaker 2>is only able to subsidize its subscribers anywhere from eight

0:05:37.120 --> 0:05:39.520
<v Speaker 2>to thirteen and a half dollars for every dollar of

0:05:39.600 --> 0:05:43.159
<v Speaker 2>revenue because of the endless venture welfare they receive and

0:05:43.200 --> 0:05:48.400
<v Speaker 2>from hyperscalers too. The underlying economics suggests that no subscription

0:05:48.480 --> 0:05:51.560
<v Speaker 2>based AI service, let alone a free one, makes any

0:05:51.640 --> 0:05:54.320
<v Speaker 2>kind of financial sense, and the only reason that everybody

0:05:54.360 --> 0:05:57.000
<v Speaker 2>has had such unrestrained access to AI is because the

0:05:57.040 --> 0:05:59.159
<v Speaker 2>media and the market's approved it, and the people with

0:05:59.160 --> 0:06:01.360
<v Speaker 2>the money are deluded and disconnected from the process of

0:06:01.400 --> 0:06:05.679
<v Speaker 2>value creation on almost every imaginable level. Any statements around

0:06:05.720 --> 0:06:09.200
<v Speaker 2>Anthropic actually being profitable on inference of products of fantasy

0:06:09.240 --> 0:06:12.080
<v Speaker 2>and magical thinking the still copium for people that would

0:06:12.120 --> 0:06:14.800
<v Speaker 2>rather delude themselves into believing that any of this ever

0:06:14.839 --> 0:06:19.120
<v Speaker 2>made sense than face reality or do basic mathematics. Again,

0:06:19.400 --> 0:06:21.880
<v Speaker 2>the assumption is that companies would never just burn a

0:06:21.920 --> 0:06:24.240
<v Speaker 2>lot of money, but that too is catering to the

0:06:24.240 --> 0:06:27.400
<v Speaker 2>grade to myth of executive competence, something that nobody who

0:06:27.440 --> 0:06:30.320
<v Speaker 2>spends any amount of time around managers or CEOs would

0:06:30.320 --> 0:06:34.359
<v Speaker 2>ever believe get hub copilot let people burn thousands of

0:06:34.360 --> 0:06:36.800
<v Speaker 2>dollars on a thirty nine dollars a month subscription as

0:06:36.800 --> 0:06:40.520
<v Speaker 2>a means of expressing growth. I absolutely, one hundred percent

0:06:40.520 --> 0:06:43.640
<v Speaker 2>believe that both open ai and Anthropic are doing exactly

0:06:43.680 --> 0:06:46.200
<v Speaker 2>the same thing, and that neither of them has some

0:06:46.400 --> 0:06:49.480
<v Speaker 2>magical way of making inference cheap enough to justify letting

0:06:49.480 --> 0:06:52.080
<v Speaker 2>people burn thousands of dollars and one hundred or two

0:06:52.200 --> 0:06:55.400
<v Speaker 2>hundred dollars a month subscription. The whole profitable on inference

0:06:55.480 --> 0:06:58.039
<v Speaker 2>thing is a fucking myth, and to give them the

0:06:58.040 --> 0:07:00.000
<v Speaker 2>benefit of the doubt is to empower them to continue

0:07:00.120 --> 0:07:03.479
<v Speaker 2>You're raising money by conning their investors and the general public,

0:07:03.560 --> 0:07:06.440
<v Speaker 2>and to continue perpetuating an era of software that runs

0:07:06.520 --> 0:07:10.640
<v Speaker 2>contrary to what makes technology good. Their goal is simple

0:07:10.800 --> 0:07:12.640
<v Speaker 2>to ram as much of this through to as many

0:07:12.680 --> 0:07:14.560
<v Speaker 2>people as possible, to get them to spend as much

0:07:14.600 --> 0:07:18.440
<v Speaker 2>money as possible, until they work out a way to

0:07:18.480 --> 0:07:21.440
<v Speaker 2>make open AI or Anthropic or these endless data centers

0:07:21.480 --> 0:07:25.160
<v Speaker 2>into something approaching a real business. One of the greatest

0:07:25.200 --> 0:07:27.560
<v Speaker 2>mistakes we can make in our lives is to assume

0:07:27.600 --> 0:07:30.040
<v Speaker 2>that the rich and powerful have any idea what the

0:07:30.080 --> 0:07:33.240
<v Speaker 2>future holds, or that they have any grand plan or strategy.

0:07:34.000 --> 0:07:37.160
<v Speaker 2>It's very likely that Dario Amaday and Sam Mortman's plan

0:07:37.240 --> 0:07:39.480
<v Speaker 2>is to keep burning money until somebody works out for

0:07:39.680 --> 0:07:43.000
<v Speaker 2>them how to stop burning money. And in the interim,

0:07:43.040 --> 0:07:44.760
<v Speaker 2>their plan is to get as many users as they

0:07:44.800 --> 0:07:48.560
<v Speaker 2>can to keep raising money. That's it. That's the entire plan.

0:07:48.680 --> 0:07:51.360
<v Speaker 2>There is no secret thing. They're not building AGI. They

0:07:51.360 --> 0:07:53.920
<v Speaker 2>don't know how to do. Their AGI doesn't exist. They

0:07:53.920 --> 0:07:57.560
<v Speaker 2>don't have secretly super profitable inference. If they did, they

0:07:57.600 --> 0:08:00.400
<v Speaker 2>would only sell model access. That's all they would do,

0:08:00.440 --> 0:08:03.480
<v Speaker 2>and it would be profitable. Right. They don't have that.

0:08:03.560 --> 0:08:05.800
<v Speaker 2>They don't have a plan. There is no secret model

0:08:05.800 --> 0:08:09.080
<v Speaker 2>behind the scenes. Even mythos is a fucking myth. I've

0:08:09.080 --> 0:08:11.400
<v Speaker 2>been over that before. Put a link in the notes.

0:08:11.880 --> 0:08:15.680
<v Speaker 2>And similarly, Microsoft, Google, Meta and Amazon's plan is to

0:08:15.720 --> 0:08:17.520
<v Speaker 2>keep building data centers in the hope they'll have a

0:08:17.520 --> 0:08:19.679
<v Speaker 2>reason to use them by the time they're actually finished.

0:08:20.240 --> 0:08:23.440
<v Speaker 2>There is no other plan. They do not have a

0:08:23.480 --> 0:08:26.200
<v Speaker 2>secret invention coming. They do not have AGI in a

0:08:26.240 --> 0:08:28.680
<v Speaker 2>box in their office. They do not have anything. And

0:08:28.720 --> 0:08:31.120
<v Speaker 2>the reason they're spending so much money and shoving AI

0:08:31.160 --> 0:08:33.840
<v Speaker 2>into everything you use is because they have no fucking

0:08:33.840 --> 0:08:37.280
<v Speaker 2>clue what to do. This is why Warrio Amida makes

0:08:37.280 --> 0:08:40.440
<v Speaker 2>wild claims about AI replacing fifty percent of all white

0:08:40.440 --> 0:08:45.000
<v Speaker 2>collar workers, or Microsoft aiceo Mustafa Suliman claims all white

0:08:45.040 --> 0:08:48.160
<v Speaker 2>collar labor will be automated in eighteen months because the

0:08:48.200 --> 0:08:51.719
<v Speaker 2>products themselves, the actual products aren't impressive enough to win

0:08:51.800 --> 0:08:54.160
<v Speaker 2>you over or to justify the hundreds of billions of

0:08:54.200 --> 0:08:57.400
<v Speaker 2>dollars being sunk into AI. They say these things to

0:08:57.440 --> 0:08:59.720
<v Speaker 2>make you think that they have a scary or powerful

0:09:00.000 --> 0:09:04.080
<v Speaker 2>technology behind the scenes when they do not. And yes, God, yes,

0:09:04.480 --> 0:09:08.800
<v Speaker 2>does that include fucking mythos. I'm so tired of seeing

0:09:09.360 --> 0:09:12.079
<v Speaker 2>news stories about how banks are meeting up to talk

0:09:12.120 --> 0:09:15.880
<v Speaker 2>about the scary monster, the scary Dracula of Mythos. They

0:09:15.880 --> 0:09:18.640
<v Speaker 2>should have called it Dracula. I pretend to take this

0:09:18.760 --> 0:09:23.880
<v Speaker 2>shit seriously, and this forceful harassment grade incursion if AI

0:09:24.000 --> 0:09:26.679
<v Speaker 2>services into our daily lives is not a sign of

0:09:26.720 --> 0:09:29.360
<v Speaker 2>its power, but a gesture of the lack of confidence

0:09:29.400 --> 0:09:32.440
<v Speaker 2>and fear in the hearts of its progenitors. Good shit

0:09:32.600 --> 0:09:35.880
<v Speaker 2>sells by telling you why it's good. Dodgy shit cells

0:09:35.920 --> 0:09:38.520
<v Speaker 2>by tricking and scaring you and taking advantage of business

0:09:38.520 --> 0:09:40.319
<v Speaker 2>idiots who think that using an L and M to

0:09:40.400 --> 0:09:42.680
<v Speaker 2>type emails and spending twelve hours a day on Twitter

0:09:42.760 --> 0:09:46.480
<v Speaker 2>constitutes work. I believe that the vast majority of these

0:09:46.559 --> 0:09:49.640
<v Speaker 2>data centers go unuse and or unfinished, and the most

0:09:49.640 --> 0:09:53.240
<v Speaker 2>AI startups will die once the venture capital subsidies dry up.

0:09:54.360 --> 0:09:57.199
<v Speaker 2>I believe that neither open AI nor anthropic have a future,

0:09:57.360 --> 0:09:59.600
<v Speaker 2>and that their revenues are only made possible through venture

0:09:59.600 --> 0:10:03.320
<v Speaker 2>subsidyce for startups using their models, and the experimental revenue

0:10:03.320 --> 0:10:05.520
<v Speaker 2>of business idiots that don't really know why they're doing

0:10:05.559 --> 0:10:09.880
<v Speaker 2>AI in the first place. AI demand remains a result

0:10:09.920 --> 0:10:12.960
<v Speaker 2>of societal psychosis and a weakness in those who are

0:10:12.960 --> 0:10:16.680
<v Speaker 2>meant to scrutinize the untrustworthy. It's unraveling will be framed

0:10:16.679 --> 0:10:19.760
<v Speaker 2>as impossible to see coming because nobody in power bothered

0:10:19.760 --> 0:10:32.240
<v Speaker 2>to fucking look and look. I've seen some of you

0:10:32.280 --> 0:10:35.199
<v Speaker 2>on the Reddit. I've had some emails from you. You're scared.

0:10:35.679 --> 0:10:38.240
<v Speaker 2>I get him. It's easy to feel hopeless. We're at

0:10:38.240 --> 0:10:41.080
<v Speaker 2>a point where agreed and the shamelessness and the stupidity

0:10:41.120 --> 0:10:44.680
<v Speaker 2>is at its fever pitch. We've reached a time too

0:10:44.760 --> 0:10:47.400
<v Speaker 2>when the maskers started to slip and the c suite

0:10:47.440 --> 0:10:50.920
<v Speaker 2>imbecile classes unabashed and about its loathing of people, as

0:10:51.000 --> 0:10:53.000
<v Speaker 2>was the case when the CEO of a UK bank,

0:10:53.160 --> 0:10:56.080
<v Speaker 2>Bill Winters, talked about how those at risk of losing

0:10:56.120 --> 0:10:59.839
<v Speaker 2>their jobs to AI are lower value human capital the

0:11:00.080 --> 0:11:02.240
<v Speaker 2>same event where he said that the company would likely

0:11:02.400 --> 0:11:05.040
<v Speaker 2>shared nearly eight thousand roles in the coming years due

0:11:05.080 --> 0:11:09.000
<v Speaker 2>to AI. Winters would later apologize for his choice of words,

0:11:09.040 --> 0:11:11.320
<v Speaker 2>although to be clear, he was being absolutely honest when

0:11:11.320 --> 0:11:14.200
<v Speaker 2>he made the remarks. That's what he believes, that's what

0:11:14.280 --> 0:11:16.360
<v Speaker 2>gets him up in the morning, that's what gets him hard.

0:11:16.720 --> 0:11:19.160
<v Speaker 2>Bill Winton's loves him. Now it must be clear for

0:11:19.240 --> 0:11:21.480
<v Speaker 2>legal reasons that was a joke. I don't know about

0:11:21.480 --> 0:11:26.600
<v Speaker 2>Bill Winter's erection. I won't guess it at any further anyway,

0:11:27.000 --> 0:11:30.480
<v Speaker 2>Move it on, though. Everything feels rough because the AI

0:11:30.559 --> 0:11:34.960
<v Speaker 2>industry is equal parts desperate and overconfident. AI executives believe

0:11:35.000 --> 0:11:37.480
<v Speaker 2>that they can cram through enough promises of money into

0:11:37.520 --> 0:11:40.199
<v Speaker 2>the system that the system would rather cannibalize itself than

0:11:40.240 --> 0:11:44.760
<v Speaker 2>admit it made a mistake. Sundarbishai, Andy Jase, Larry Ellison,

0:11:44.840 --> 0:11:48.760
<v Speaker 2>Elon Musk Morg Zuckerberg and Sachinadella will gladly annihilate hundreds

0:11:48.760 --> 0:11:52.000
<v Speaker 2>of billions of dollars to avoid facing the inevitable, but

0:11:52.160 --> 0:11:56.600
<v Speaker 2>once they do, it'll be gruesome. At the same time.

0:11:56.679 --> 0:11:59.800
<v Speaker 2>The things that they need to happen, actual profitability, actual

0:11:59.800 --> 0:12:02.240
<v Speaker 2>time and investment, actual tangible proof that this is a

0:12:02.280 --> 0:12:04.800
<v Speaker 2>real thing rather than something they all have to actively

0:12:04.840 --> 0:12:08.920
<v Speaker 2>conspire to keep alive. Well, they're not happening. Each week

0:12:08.960 --> 0:12:11.760
<v Speaker 2>we hear about new AI megaprojects that will dominate our

0:12:11.760 --> 0:12:15.040
<v Speaker 2>countryside with blinding lights, endless noise, and few belging gas

0:12:15.080 --> 0:12:18.200
<v Speaker 2>turbines at such a scale that it feels impossible it

0:12:18.200 --> 0:12:21.640
<v Speaker 2>could ever stop. The system is absolutely going to try

0:12:21.640 --> 0:12:24.760
<v Speaker 2>and exhaust itself to keep it going. The government bought

0:12:24.840 --> 0:12:28.480
<v Speaker 2>nine billion dollars of Blackwell GPUs, which may seem like

0:12:28.600 --> 0:12:30.440
<v Speaker 2>it's a too big to fail, but what it actually

0:12:30.520 --> 0:12:32.560
<v Speaker 2>is is just the way to keep in Vidia's plates

0:12:32.600 --> 0:12:35.520
<v Speaker 2>spinning for another quarter. Nine billion dollars isn't going to

0:12:35.559 --> 0:12:38.800
<v Speaker 2>be enough. In truth, the amount of money that in

0:12:38.840 --> 0:12:41.200
<v Speaker 2>Nvidia needs to keep this going is so extreme that

0:12:41.240 --> 0:12:42.880
<v Speaker 2>it is now a test of how long the debt

0:12:42.880 --> 0:12:46.920
<v Speaker 2>markets and the hyperscalers can keep sustaining the situation, the hype,

0:12:46.920 --> 0:12:50.480
<v Speaker 2>and the industry itself. A trillion dollars in annual revenue

0:12:50.520 --> 0:12:53.200
<v Speaker 2>is necessary by the end of twenty twenty eight for Invidia,

0:12:53.240 --> 0:12:56.720
<v Speaker 2>which would require over thirty gigawots of actual operational data

0:12:56.760 --> 0:12:58.920
<v Speaker 2>center capacity to be built by then at a time

0:12:58.960 --> 0:13:01.040
<v Speaker 2>when only five gig was It most appears to be

0:13:01.120 --> 0:13:04.760
<v Speaker 2>under construction, and under construction could mean anything from actual

0:13:04.800 --> 0:13:09.440
<v Speaker 2>building happening to a scaffolding yard. Nevertheless, even the sweatiest,

0:13:09.520 --> 0:13:12.640
<v Speaker 2>least trustworthy boosters have begun to sneak in statements about

0:13:12.640 --> 0:13:15.199
<v Speaker 2>how we're probably not in a bubble, or yeah, it's

0:13:15.240 --> 0:13:17.679
<v Speaker 2>a bubble, but it's a good bubble. Jeff Bezos, when

0:13:17.679 --> 0:13:19.920
<v Speaker 2>asked about the AI bubble, said that you shouldn't worry

0:13:19.960 --> 0:13:25.079
<v Speaker 2>about it, which, well, well, Jeff, that that doesn't really

0:13:25.120 --> 0:13:27.720
<v Speaker 2>help anyone. Jeff, I mean, you probably don't need to

0:13:27.720 --> 0:13:31.240
<v Speaker 2>worry about it. You're rich, twap. Anyway, none of this

0:13:31.400 --> 0:13:33.319
<v Speaker 2>is to say that the mood is good. The vibes

0:13:33.320 --> 0:13:37.040
<v Speaker 2>are disastrous. Everybody's exhausted. Those who love AI vibrate with

0:13:37.080 --> 0:13:40.320
<v Speaker 2>a strange soullessness, constantly talking about the incredible power of

0:13:40.360 --> 0:13:43.080
<v Speaker 2>AI without ever showing what it did, or perhaps what

0:13:43.160 --> 0:13:48.080
<v Speaker 2>all that supposed saved time got them. What are you

0:13:48.120 --> 0:13:50.560
<v Speaker 2>doing with the spare time? I don't see any chuckle

0:13:50.600 --> 0:13:53.720
<v Speaker 2>folks providing anything useful to the word. I don't see

0:13:53.720 --> 0:13:57.880
<v Speaker 2>you living lives of leisure. I just see you posting online.

0:13:58.360 --> 0:14:00.400
<v Speaker 2>I can do that without the power of a because

0:14:00.400 --> 0:14:03.960
<v Speaker 2>I have a working brain. What's your fucking excuse, good lord.

0:14:04.920 --> 0:14:07.200
<v Speaker 2>But one thing I'll add though, that adds to the

0:14:07.280 --> 0:14:09.560
<v Speaker 2>malaise is that everyone I talk to to a lot

0:14:09.559 --> 0:14:12.800
<v Speaker 2>of people at hyperscales, they're all fucking miserable. It's terrible there.

0:14:12.840 --> 0:14:15.720
<v Speaker 2>The business idiots are running the show, and they're both saying,

0:14:15.760 --> 0:14:18.280
<v Speaker 2>spend more money on AI, but also don't spend quite

0:14:18.280 --> 0:14:20.280
<v Speaker 2>as much money on AI because it costs a lot.

0:14:20.880 --> 0:14:23.120
<v Speaker 2>Nobody really knows why they're doing this other than the

0:14:23.200 --> 0:14:27.000
<v Speaker 2>fact it's what everyone else is doing. And basically every

0:14:27.160 --> 0:14:30.400
<v Speaker 2>person in every job has had somebody intimate that they're

0:14:30.440 --> 0:14:32.560
<v Speaker 2>going to lose their job to a computer every time

0:14:32.560 --> 0:14:35.120
<v Speaker 2>they open a newspaper or use a website, and every

0:14:35.160 --> 0:14:37.640
<v Speaker 2>app has some sort of desperate, vulgar pop up about

0:14:37.680 --> 0:14:40.760
<v Speaker 2>a feature that will generate some bullshit officscating the features

0:14:40.800 --> 0:14:42.560
<v Speaker 2>you actually want to use in favor of those that

0:14:42.680 --> 0:14:46.320
<v Speaker 2>might lose the company money, because the company has to

0:14:46.360 --> 0:14:48.640
<v Speaker 2>prove to the people that invested in the company that

0:14:48.680 --> 0:14:52.120
<v Speaker 2>they're futuristic and have the AI integrations, even though no

0:14:52.120 --> 0:14:56.479
<v Speaker 2>one likes them and they only seem to lose money. Alternatively,

0:14:56.520 --> 0:14:59.160
<v Speaker 2>their CEO either has my old or severe AI sside

0:14:59.200 --> 0:15:01.560
<v Speaker 2>coosis to the point they have decided to violate your

0:15:01.640 --> 0:15:05.800
<v Speaker 2>user experience because AI is a non consensual technology at

0:15:05.840 --> 0:15:09.680
<v Speaker 2>its heart, this is what it is. Nobody asked for this,

0:15:10.080 --> 0:15:13.520
<v Speaker 2>no one wants this. We've had it forced upon us.

0:15:13.880 --> 0:15:16.760
<v Speaker 2>Even the people excited by it would rather do something else.

0:15:17.440 --> 0:15:20.200
<v Speaker 2>And all of that excitement doesn't seem to amount to anything.

0:15:21.120 --> 0:15:23.680
<v Speaker 2>And they all seem desperate, and they seem anxious and

0:15:23.880 --> 0:15:28.560
<v Speaker 2>scared and vulgar in the way they attack their detractors.

0:15:29.280 --> 0:15:33.800
<v Speaker 2>And that's because they're losing. They all know it. They're desperate,

0:15:33.840 --> 0:15:37.160
<v Speaker 2>they're scared. They're scared that the thing that they mortgage

0:15:37.200 --> 0:15:41.400
<v Speaker 2>their companies and careers and reputations on is all falling apart.

0:15:43.280 --> 0:15:46.040
<v Speaker 2>And it seems that there are nearly as many announcements

0:15:46.040 --> 0:15:49.120
<v Speaker 2>of new large data center developments as there are cancelations

0:15:49.120 --> 0:15:53.000
<v Speaker 2>of said data center projects. While hyperscolars can dismiss that

0:15:53.120 --> 0:15:56.200
<v Speaker 2>as a simple reallocation of capital and nothing to worry about,

0:15:56.440 --> 0:15:59.280
<v Speaker 2>it's hard to ignore the growing backlash against these facilities

0:15:59.320 --> 0:16:02.040
<v Speaker 2>from locals and the success that these locals have had,

0:16:02.560 --> 0:16:06.760
<v Speaker 2>especially in blocking multiple different developments or setting up actual

0:16:07.440 --> 0:16:11.080
<v Speaker 2>what county wide and statewide moratoriums depending on where you look.

0:16:11.960 --> 0:16:15.160
<v Speaker 2>And it gets worse. Anthropic had to conspire with Elon

0:16:15.360 --> 0:16:18.680
<v Speaker 2>fucking Musk to conjure up a single profitable quarter with

0:16:18.800 --> 0:16:22.200
<v Speaker 2>SpaceX deliberately discounting two months of compute costs as a

0:16:22.240 --> 0:16:25.160
<v Speaker 2>means of swindling the media and its investors. One last time,

0:16:25.520 --> 0:16:28.320
<v Speaker 2>I swear to God, the object permanence in the valley

0:16:28.440 --> 0:16:32.120
<v Speaker 2>is not so good, folks. But in response, open Ai

0:16:32.240 --> 0:16:34.400
<v Speaker 2>either leaked or had leaked that it had a negative

0:16:34.440 --> 0:16:37.600
<v Speaker 2>one hundred and twenty two percent non gap operating margin

0:16:37.640 --> 0:16:40.480
<v Speaker 2>in Q one twenty twenty six and chat GPT growth

0:16:40.480 --> 0:16:44.960
<v Speaker 2>that's stalled. Anthropic is either the single most successful grifter

0:16:45.080 --> 0:16:46.920
<v Speaker 2>of all time or speed running a con where it

0:16:46.960 --> 0:16:49.360
<v Speaker 2>fudget together numbers to raise endless amounts of money to

0:16:49.440 --> 0:16:53.360
<v Speaker 2>keep its billion or trillion dollar burn going. These are

0:16:53.360 --> 0:16:56.440
<v Speaker 2>not the actions of honest, sustainable companies that will exist

0:16:56.440 --> 0:16:59.320
<v Speaker 2>in the future. I believe that we are on course

0:16:59.360 --> 0:17:02.320
<v Speaker 2>for a truly horrible crash, the likes of which may

0:17:02.480 --> 0:17:06.159
<v Speaker 2>rewrite the venture capital industry and mortally wound one or

0:17:06.200 --> 0:17:09.720
<v Speaker 2>more hyperscalers, as well as fundamentally divide society on so

0:17:09.800 --> 0:17:12.440
<v Speaker 2>many levels into those that fall for this and those

0:17:12.480 --> 0:17:15.879
<v Speaker 2>that did not. This will, in the short term be

0:17:16.000 --> 0:17:19.040
<v Speaker 2>absolutely fucking horrible for our markets in our wider economy

0:17:19.119 --> 0:17:21.439
<v Speaker 2>is a result of the time bomb of private credit

0:17:21.560 --> 0:17:25.119
<v Speaker 2>and private equity and their failed investments in software and

0:17:25.160 --> 0:17:28.879
<v Speaker 2>their failing investments in data centers. In the long term,

0:17:29.000 --> 0:17:30.800
<v Speaker 2>I see it as that they live moment for many

0:17:30.800 --> 0:17:33.720
<v Speaker 2>millions of secret imbeciles and credits in our mist and

0:17:33.800 --> 0:17:36.200
<v Speaker 2>I don't think it'll be easy to wash the stench off.

0:17:36.200 --> 0:17:38.760
<v Speaker 2>For those that really pledge themselves to the graveyard smash

0:17:38.760 --> 0:17:43.240
<v Speaker 2>of AI, we will win long term, what they are

0:17:43.240 --> 0:17:46.240
<v Speaker 2>doing is not working. The future will not be without pain,

0:17:46.320 --> 0:17:48.560
<v Speaker 2>nor will it be easy or pleasant or something I

0:17:48.560 --> 0:17:51.120
<v Speaker 2>will relish in. But in the long term, I think

0:17:51.119 --> 0:17:53.760
<v Speaker 2>this is a moment where the greater business idiot incursion

0:17:54.000 --> 0:17:56.400
<v Speaker 2>faces a reckoning with a reality it believed it could

0:17:56.440 --> 0:17:59.760
<v Speaker 2>change through sheer force of will. These people don't know

0:17:59.800 --> 0:18:02.080
<v Speaker 2>how to build things that work anymore, and that's the

0:18:02.080 --> 0:18:04.639
<v Speaker 2>only thing they can do is spend money in fire people.

0:18:05.280 --> 0:18:08.080
<v Speaker 2>They believe in nothing other than growth, and one cannot

0:18:08.080 --> 0:18:11.679
<v Speaker 2>exist on belief in hyperlone, at least not forever. And

0:18:11.760 --> 0:18:14.399
<v Speaker 2>I can't wait to watch what happens when it collapses.

0:18:22.760 --> 0:18:24.440
<v Speaker 2>Thank you for listening to Better Offline.

0:18:24.560 --> 0:18:27.000
<v Speaker 3>The editor and composer of the Better Offline theme song

0:18:27.080 --> 0:18:29.720
<v Speaker 3>is Matasowski. You can check out more of his music

0:18:29.760 --> 0:18:33.400
<v Speaker 3>and audio projects at Matasowski dot com, M A T

0:18:33.400 --> 0:18:34.000
<v Speaker 3>T O.

0:18:34.200 --> 0:18:37.679
<v Speaker 2>S O w Ski dot com.

0:18:37.760 --> 0:18:40.080
<v Speaker 3>You can email me at easy at Better offline dot

0:18:40.119 --> 0:18:42.320
<v Speaker 3>com or visit Better Offline dot com to find more

0:18:42.359 --> 0:18:45.760
<v Speaker 3>podcast links and of course, my newsletter. I also really

0:18:45.760 --> 0:18:48.080
<v Speaker 3>recommend you go to chat dot Where's youreaed dot at

0:18:48.080 --> 0:18:50.520
<v Speaker 3>to visit the discord, and go to our slash Better

0:18:50.560 --> 0:18:53.720
<v Speaker 3>Offline to check out our reddit. Thank you so much

0:18:53.760 --> 0:18:54.280
<v Speaker 3>for listening.

0:18:55.119 --> 0:18:57.760
<v Speaker 1>Better Offline is a production of cool Zone Media. For

0:18:57.920 --> 0:19:00.560
<v Speaker 1>more from cool Zone Media, visit our web site cool

0:19:00.640 --> 0:19:03.960
<v Speaker 1>Zonemedia dot com or check us out on the iHeartRadio app,

0:19:04.040 --> 0:19:06.440
<v Speaker 1>Apple Podcasts, or wherever you get your podcasts.