You are someone who's just been allocating, interestingly, to a public company that's, of course, based in Europe, but thinking about infrastructure, about access to the cloud. Nimbus is the name of the company you were looking at, but you're also just

thinking of the turf warfare that's out there to allocate money to generative AI. This is just going to add to the party, right? Yeah, absolutely. Well, first, thanks for having me on. But we at Excel are big believers that

we're only in the early innings of laying down the infrastructure that's going to be required to actually support this massive technological wave of AI. So this committed investment doesn't surprise me at all to see it coming in because we do need that infrastructure. So I'm excited to see more

infrastructure being laid down for it. Okay. Where are you seeing the biggest opportunities? You've just made this allocation of funds into a publicly traded provider of cloud? We've got CoreWeave already up there as a big pivot from all these companies that were looking at Bitcoin

mining and now getting into the world of cloud computing. But what next in terms of areas that you think are fruitful? Yeah, absolutely. So at Excel, we broadly invest across two different facets of AI. We think about infrastructure, which is

very much an example of that. Also investing in companies like Scale AI in the infrastructure layer, helping build the data engine for AI. But then we can't forget about the applications that are going to run on that infrastructure. So we see great opportunities in both

enterprise and consumer applications. Our most recent investment on the consumer side of things was in a company called Speak, supporting language learning for AI. So that is the tussle right at the moment. There is an awful lot of money being allocated to building picks and shovels, infrastructure, and many feel the

valuations potentially have gone out of whack there. We now go into the killer applications that many are still waiting on. Do you think we're actually there, and people aren't seeing the application of generative AI right now, the real productivity gains?

Absolutely. We see it from the ground up, and we see companies that are gaining real traction and driving real value, both in enterprise and consumer. So although it is still the early innings, and a lot of the emphasis has been on that infrastructure, the

training to get us to the point where we can actually run inference on these models and get the value out of them. I think you're going to start to see, particularly across 2025 and 2026, applications spring up both in

enterprise and consumer that drive real value to the end user. But then when I think of enterprise, you've got Salesforce all in on agent force, you've got Microsoft trying to offer it. So, Co-pilot, where is the space? Or is it actually that a lot of these startups

will end up being eaten up by these bigger enterprise companies? That's the exit strategy. Yeah, that's a good, it's a really good question. So, speaking on the infrastructure layer, what we actually see is that by pairing their software and hardware

engineers against the problem of standing up infrastructure, but doing it with an AI-specific approach, we're able to offer a level of flexibility that the large incumbents offering generalized clouds just don't have in the ecosystem today. So that means that people can come to us

that need large amounts of compute, that want bare metal. And people can also come to us that just want an on-demand flexible experience with their infrastructure. So as always, you see the innovation in the early startups tackling some of the holes and open opportunities in the large

incumbents. And I expect to see the same thing. Salesforce is obviously defending its territory, but we see a ton of activity and tech approaches and apps going after their exact market. So we'll continue to see it. Is Gen AI going to be the next sort

of over-allocated area? I mean, we all got way too in the weeds when it came to chips and semiconductors for 2024. When it comes to 2025, we're all going to be playing an agenda.

I don't think it's overhyped at all. I think the reality is we're still in the early innings, and the expectations are quite high on what agents are going to do within the enterprise application arena. And so it's easy to point at the flaws

in agents today, but I tend to look at the progress of agents, and as an investor, I look at where value is going to be created over the next decade. And over the next decade, I think you're going to see really mature agents hit

the market and scale in a meaningful way.

Laying Down the AI Infrastructure

Welcome to the Bloomberg Technology YouTube channel. Subscribe for original series and mini-docs on game-changing technologies, profiles on innovators, live conversations with experts, and breaking technology news & analysis from Bloomberg TV.

Laying Down the AI Infrastructure

Laying Down the AI Infrastructure

Scripts