AI Investment Is Not Slowing Down: Jefferies’ Thill

Key Takeaways

It's been a week. It's really interesting. I just got back from Washington DC, right, Brent for the AI action plan. I listened to the president. I appreciate that Intel is not actually on your coverage list, but it was interesting that Intel wasn't in the room there in DC. Uh, Alphabet wasn't either. They had their earnings print. Would you say there's kind of one core lesson right now about what's happening in infrastructure in America and what the biggest result of this week has been from the administration? Yeah, I think the key takeaway is that the AI investment boom is still happening at a rapid rate. There's no slowdown. Google uh raised their entire capex from 75 to 85 billion on their cloud business, not their ad business. So that's uh selling uh Google cloud to enterprises. And I think it tells you that we're still in the infrastructure buildout. we continue to see, you know, the dirt, the energy, the the power lines, uh the core servers, that's really been the focus and the AI trade has not come to applications. It has not gone to Salesforce or Adobe uh to many of those names. They're not monetizing this yet. We think that happens and that has to happen for AI to work in our in our view uh that we have to go up through the infrastructure layer to the application layer. And so I think that's the next chapter. But I'd say uh so far uh alive and kicking uh in doing very well in AI infrastructure. There's been no signs of slowdown. You heard Mark Zuckerberg talk about uh spending hundreds of billions of dollars in capex uh just a couple weeks ago. Uh Google just raised their capex by 10 bill. And then we've got Microsoft Amazon on on deck. And we think that ultimately in our checks the demand for hyperscaler infrastructure is accelerating and we're seeing the the terries starting to shake off. >> Bren the the AI action plan largely focuses on deregulation and expedited permitting of all of the companies that you cover where that's relevant. How will it help them? Well, I think it should help the entire industry if we have lower bars and lower regulation. Uh we're going to see hopefully a lot more consolidation. I think the one thing we're seeing is there's a lot of companies that say they're AI first. Uh we saw this movie in the internet boom in the '90s. We saw this in the cloud boom. There's going to be massive consolidation. So, you're seeing already the beginning of a of an M&A wave. So, I think it'll help M&A. It should help the buildout. it should help protect the US in terms of what we're doing as a country in in AI. Um so I think you know the president has us right and certainly um uh you look at uh our allies and in building this it's it's a collection of of different technology companies and I think what we're we thought would happen at the beginning of AI is that we consolidate power to to a few and what we're finding is it's consolidating power to many and so there's just not enough capacity. Everyone's capacity constrained. Google said they're capacity constrained. Microsoft said they were going to be constrained, but but what's happening now is that uh you're seeing demand get spread all over and that's actually good for the ecosystem and good for many vendors. Uh over time again I think you're going to see a tremendous amount of clean up and M&A and that that needs to happen because there's probably going to be too many companies chasing this uh long term but short term I think you know there's no no question we're an AI boom and uh it's really exciting and and I think it's real. I don't think this is this is fake like we've seen in other tech trends. >> It's interesting service now Bill Mcderman yesterday saying this is an existential issue for many in CRM who aren't on the AI bandwagon and thinking about to IBM and they look like they're selling well in terms of consulting into AI. The software bit was a little bit woolly and people pulled back from what had been a ramp up in the stock. Brent, just before we push forward to what's next week, what did you make of some of the weaker plays and how much they get beaten up this week? Yeah, I mean I IBM didn't have uh a phenomenal quarter. They introduced uh you know a new infrastructure um cycle and that kind of paused demand. They also integration Hashi which is a a deal they did and so the software business decelerated at the point you know it wasn't it wasn't terrible but the stock had run up and is almost sitting at the same multiple that Microsoft was at. And I think you know IBM's a single digit grower, Microsoft is a double digit grower. So I I think the stock got mismarked. We don't have a buy on IBM. uh we like what they're doing. We like Arvin's software centric uh strategy. So we're we're fundamental fans, but stock just had a big move and it kind of didn't didn't make sense maybe where it was at based on the results. >> Yeah. >> Um so I don't I wouldn't look at IBM's results and say, you know, that they're a proxy. Uh actually the CEO upgraded the overall economy from cautiously optimistic to optimistic. And I said, well, what's next? Is there a ludic ludicrous speed or is there another speed above optimistic? But, you know, Arvin, you know, has a pretty good lens in what he sees. And I think that was probably the most interesting thing, >> which is to us that that's a good tellt tell what's about to happen next week. >> Yeah. I mean, Alvin told me he's really looking at M&A at the moment. And also that he's more optimistic about consulting for 2026 than he is about 2025. Push forward then for next week and the optimism around what this means for Microsoft spend, what this means for Amazon spend, and how much their cloud units are growing. Yeah. So, Microsoft uh we they're on an acceleration path. They've said they're capacity constraint. They're accelerating the guide to an acceleration. We're now, you know, looking at mid30% Azure infrastructure build. Uh for Amazon, last quarter they put up 20% backlog and that was up from 14. So, their backlog's growing, which is a sign of what was going to happen with revenue. So, we continue to see uh right now a continued uh acceleration across the board. And again, I I would highlight we haven't even seen AI hit these numbers in in in in a big way. I mean, for all these big companies, it's less than 5% of total revenue. >> Uh, sorry, sorry to interrupt. We just had 30 seconds left, but I wanted to ask that. Is capex still the data point to watch, or you want to see topline growth as a result of capex in the prior year? >> Booking. Bookings is number one, revenue growth, and then uh capex in that