Monologue: Concentration Risk

0:03 Today in sentences that are not in the Bible, YouTuber Mark Plyer bought a major stake in camera maker GoPro and then GoPro turned into an AI data center neocloud.

0:14 I'm sick and tired of the goddamn AI bubble, I swear to Christ. Bird up, this is Better Offline and I'm your host Ed Zitron. So today we're going to talk through a term you may or may not have heard before, concentration risk. It's a term that refers to having all your eggs in one or a few baskets, becoming overly reliant on a few investments, customers, or particular business lines to the point that without them, your business or portfolio would suffer massive harms or just explode. In banking specifically, to quote the National Credit Union administration, It refers to any single exposure or group of exposures with the potential to produce losses large enough relative to capital, total assets, or overall risk level to threaten a financial

1:03 institution's health or ability to maintain its core operations. I bring this all up because you're going to hear this term or variations of this term a lot in the next few months and years as the AI bubble unravels, because just about every part of the industry involves its own flavor of concentration risk. Let's start at the top.

1:22 Per data from fintech firm Ramp, 80% of OpenAI and Anthropix Enterprise revenues come from 1% of their customers, a number that hasn't improved over the last three years. Ramp's lead economist, Ara Karazian, notes that the top 1% skews heavily towards the tech sector and AI products and services, and that this was a level of concentration risk unseen in any other software category they tracked. The dataset, which includes big companies like Visa and Cursor, as well as a great deal of startups and regular-sized companies, is very indicative of the overall spend of the AI industry, with the caveat that it doesn't include massive players like Microsoft or major banks. To be clear, I'm guessing about Visa

2:01 and Cursor. Any customer on RAMP can opt out of research. I have no idea, but I'm going to assume that there are big companies in there. I also want to be specific that when RAMP says AI products and services, that includes AI startups that sell subscriptions with subsidized token spend, meaning that users can burn far more than their subscription price in tokens. So on a $ 20 a month subscription, you can burn $ 30, $ 40, $ 100. This means that the money made by Anthropic or OpenAI from an AI startup in that 1% spend is contingent on their continued ability to raise venture capital dollars. To simmer all this down, It means that the vast majority of enterprises, which is where the real money is in software and the real

2:40 growth is, just don't spend that much money on AI. Those that do spend the most on it are heavily concentrated in either AI companies that either use a lot of tokens internally because they're bankrolled by venture capital, AI companies that allow their users to blow unsustainable amounts of money on tokens, bankrolled by venture capital, tech companies that are currently under heavy pressure to spend money on AI tokens, and I assume a few whale customers of some sort.

3:04 This means that 80% of OpenAI and Anthropix Enterprise revenues, which make up the vast majority of their total revenues, are dependent on what are likely hundreds of customers spending outsized amounts of money on AI tokens, with an indeterminately large chunk of them being AI startups that can only do so as long as venture capital allows them to.

3:22 I also, and this is a gut feeling, wouldn't be surprised if the AI startups spend way more on tokens for writing LLM code internally, considering how every time I see somebody going nuts on AI Twitter, It's usually a VC-backed startup. It also means, as I've hinted, that outside of the tech and AI world, very few companies are willing to pay very much for AI, which is catastrophic on just about every level, with no clear sign as to how you reverse that trend. AI has been in every media outlet and discussed in every boardroom and company for the last three years. Every single company has on some level dabbled in using AI. Most businesses have been given the green light to spend a bunch of money

4:00 on AI. And in the end, it seems that the only people the tech industry can get to spend money on AI is the tech industry itself. This is OpenAI and Anthropic's underlying exposure because these customers are also prime targets to move to either cheaper models that they train themselves because they're open source or eventually on device models.

4:19 Even if these customers choose to stay with Anthropic and OpenAI, a chunk of this spend is contingent on venture capital funding, like I've said, and the rest is contingent on whether tech firms continue to be willing to spend money at scale. 80% of the revenue concentration depends on spending and capital that varies from unreliable to actively unstable. Meanwhile, these two AI labs represent a massive concentration risk for Microsoft, Google, Amazon, Oracle, CoreWeave, and anyone else that sells compute to them, with Anthropic and OpenAI standing over 1.1% trillion worth of compute commitments based on demand that's mostly coming from a very small subset of customers. These are, from what I can tell,

4:56 take-or-pay agreements where they agree to buy that compute capacity regardless of how much capacity they actually end up using and how much revenue they actually bring in. As a reminder, both Anthropic and OpenAI are woefully unprofitable and lose tens of billions of dollars a year. To give you an idea of the concentration risk, OpenAI's compute spend and revenue share represent about 70% of Microsoft's AI revenues in fiscal year 26, which just ended in June, or a little over 70% of Microsoft's entire fiscal year revenue that year.

5:26 And UBS estimates that OpenAI and Anthropic's compute spend will account for 48% of Google Cloud's entire revenues next year, or somewhere between $ 84 billion and $ 100 billion in 2027. That's on top of, per Barclay's, OpenAI and Anthropix estimated $ 40 billion spent on Amazon Web Services and at least $ 50 billion that both of them will spend on Microsoft Azure in calendar year 2027, which I note because of Microsoft's old fiscal year system. On the low end, that means that Anthropic and OpenAI account for over $ 174 billion worth of expected revenues from Microsoft, Google, and Amazon in 2027, which is contingent on their ability to raise venture capital or debt, which is contingent on the continued growth of

6:10 their businesses, which is contingent on growing AI spend from a small subset of customers, many of whom are funded by venture capital. The reason this hasn't been a problem yet is that when you sign these contracts, you tend to pay a little upfront fee and the capacity in question is yet to come online. That's going to start happening next year and get dramatically worse month after month as capacity starts powering up and they start actually having to pay for it. A really shittily written piece from an outlet called Groundbreaker that people keep emailing me did make a good point about this, comparing it to when the rates on millions of mortgages exploded as they hit a reset wall in 2027. where the low teaser interest

6:47 rates ended, so when you signed a mortgage, you would get like 1%, 2%, 3%, very low, exploding the monthly mortgage payments to unsustainable highs, with customers assuming when they signed it, incorrectly, that their houses would keep appreciating, they'd be able to refinance, or they could simply sell the bloody thing, which they obviously could not do when everyone was trying to do the same thing. In other words, OpenAI and Anthropic's massive compute commitments are the subprime mortgages of the AI bubble. They signed big, beautiful deals that helped hyperscalers and neoclouds post massive revenue backlogs under the belief that nothing bad would ever happen. That growth would happen unabated, and of course the money would always be

7:24 available for everyone involved. Finally, at the top of the pile sits NVIDIA, whose concentration risk lies with the hyperscalers and neoclouds themselves, who justify buying further GPUs based on demand, and I put that in air quotes here. from OpenAI and Anthropic, with said demand for services contingent on whether they can continue to raise money. Even those buying GPUs to build AI data centers for other customers are doing so because they believe there's some sort of crazy demand for AI compute, with their reference point being the massive revenue backlogs for Core, Weave, Iron, Nebius, and other neoclouds, who primarily sell compute to either OpenAI, Anthropic, or one of the hyperscalers backing them. Oh, and NVIDIA's customers are

8:04 no longer able to buy its GPUs through cash flow alone, so all of those purchases are contingent on their constantly availability of debt. None of this is very good at all. I should also add that Broadcom added on their latest earnings that Anthropic and OpenAI are going to be their top two customers. It's all very good. It's all very normal, very good. Everything's fine here, okay? Nobody freak out. Even when you put it all in a line, it all sounds really fucking bad. I'm sorry, I'm not trying to be alarmist, but even at the end of my own monologue, I'm kind of like.

8:38 Anyone else fucking think about this? Anyone else worried? No.

8:42 The answer's no. If you ask most sell-side analysts or financial journalists, they'll tell you that all of this is totally fine and it's nothing to worry about. They will assure you that these are the smartest people in the world, the most powerful companies, that they wouldn't spend all this money for no reason. that the demand for both AI compute and AI itself is real, and that the AI skeptics are cherry-picking data. Well, we're going to fucking find out, aren't we? And when we find out, I think it's going to be the thing I've been warning about. And when that happens, I've been keeping really detailed notes about all the people that tried to hand-wave this away. Because I think this is a catastrophic misallocation of capital, but

9:20 also just the largest miss in journalism history. Just unbelievable to me that when this eventually falls apart, and I am literally looking at my fucking Bloomberg terminal, and what just popped up says, Crusoe signs roughly $ 13 billion Jane Street deal for cloud computing. Now, you may think, wow, that's a different customer, Jane Street, a hedge fund. How could they possibly be involved in this? Well, you never guess what. Jane Street's a major customer of CoreWeave and an investor in CoreWeave. I bet they fucking invest in Anthropic at some point. Jesus fucking, did they invest in anthropic All right, no, I got to end this. I got to end this goddamn monologue. Look, I'll be back next week. I still have yet to come up with

10:04 what I'm going to do, but it's going to be great. My cat just knocked over an empty Diet Coke can and that's very annoying. But nevertheless, I will be back. I love you all. I appreciate you listening. I'm Ed Zetron and this has been Better Offline.