AI’s long-term build-out is outpacing short-term demand, for now

Yes, it's counter intuitive but it's no longer AI is no longer anymore about demand. Now it's about financing. And we're seeing issues through our work for XOVR, which owns private equity in that the ETF. Um, we see uh we have exposure to a secondary market and we're seeing cracks in the valuations of companies like Open AI and Anthropic. And the reason is that there's a big mismatch of these companies having long-term commitments when it comes to infrastructure investment and compute, but only short-term commitments from their customers. And for example, Anthropic has, um, has 500 billion, that's half a trillion of commitments when it comes to long-term commitments in infrastructure and compute over the next few years. But only a handful of their large customers that account for most of the revenue, uh, have, um, signed for long-term contracts. So there's a big mismatch here, which concerns us when when it comes to these companies, the big LLMs, Open AI, Anthropic and other companies like this.

And you argue Eva, there are in fact historical parallels you say here to the telecom build out 25 years ago. Walk me through that.

Yes. Um so we see a big parallel. It's not a bubble, so it's not we do not suggest that this is a bubble, but um back in 99, uh internet was real, it was very real, but the the investors and PE money and capital was chasing, there was actually an irrational exuberance, that's the right term, chasing the telecom industry. But their expectations were never to be met because if their expect expectations were met, that would be a disproportionate amount of the global GDP. The same thing is happening today. AI is real, is here to stay, but all this PE money, all all this capital has so high expectations that if we put all these numbers together, the growth expectations, the earnings expectations, uh the capital that's being invested in infrastructure and compute, if you put this all together, it's an a disproportionate amount of the global GDP, which means that it it will never be achieved.