Why Enterprise AI Economics Are Changing

**Brian Gracely** (0:06)
Good morning, good evening, wherever you are. Welcome back to The Enterprise AI Show. I'm your host, Brian Gracely. And today, what I want to try and do, instead of maybe doing some of the things we've done in the past in terms of weekend perspectives or interviews with founders and leaders and so forth, is really kind of dive into some, what am I gonna do with this information? I think we spent a lot of time in this show, both over the last year or so, but in the past, kind of trying to educate people and identify new things that were out there, trying to identify and educate them about trends that were happening. We'll still do that on things like AI News the month and so forth, but I feel like in this point in time, especially with something new, and it's been a while since we've had something truly new like this with AI. Oftentimes, the thing that you're looking for is not so much education on what's new, but really, what do I do with that? How can I go deal with that? And so, I wanna dig into, have this list that I've been working on. I call it my top 10 enterprise AI questions or top 10 enterprise AI kind of challenges that people have right now. And I wanna kind of dig into that in a series of shows. We're gonna start with this one. We're really gonna kind of dig into the idea of token economics. And more importantly, is the sort of price that you're paying for AI today going to be something that's gonna be sustainable? Or does the overall macro of what's going on in the industry look like it's gonna change what the economics look like? And if it does, how do you go about kind of getting a handle on that? What's it gonna look like in terms of what you do on a day-to-day basis versus what your business does and so forth? So we're gonna dig into that right after the break.
Today's show is sponsored by Sharegate. You're out of time. Copilot needs to be deployed ASAP, but your tenant really isn't ready for it. Years of data, permissions and users lurking at shadows, ready to be exposed by AI. Sharegate Protect sees it all so you can find the exposure risks, fix them fast, and deploy AI with confidence. Microsoft 365 Governance, they've got this. Learn more at sharegate.com/protect.
Today's show is sponsored by Nasuni. There's a growing gap in AI right now between what's possible in theory and what successfully works at scale inside an enterprise. The difference comes down to unstructured file data. Many AI initiatives struggle because the file data they depend on is scattered, unstructured and disconnected from where and how work actually happens. Nasuni changes that. It brings your unstructured file data into a single secure foundation, so AI, both generative and agentic, can access it with the context, governance and performance it needs in production. Bring AI to where your unstructured data lives. See what it takes to activate your data for AI and request a demo at nasuni.com/ai.
We're back. And I want to dig into this thesis that I sort of mentioned and hinted on at the top of the show, which is this idea that up until now, and maybe over the last two and a half, three years, the price that we've been paying for AI has been somewhat subsidized, maybe highly subsidized. And it may not necessarily be a great prognosticator of what AI is going to cost in the future. Now, I feel like it's an interesting question, especially if we look at it from an enterprise perspective. Because on one hand, we have always seen, over time, a certain amount of things happen as new things come along. Typically, the technology, to a certain extent, becomes cheaper. Just as the classic volume play, as we see more and more people adopting the internet, or more and more people adopting iPhones, or adopting cloud computing, whatever it might be.
Because of that volume, people that are providing that are able to, due to economies of scale, able to drive the price down. And while over time, our volume of usage goes up, hence maybe our total spend goes up, over time in the past, we've already seen the price of things come down. Now, the challenge of that is, it feels like, and I want to dive into this, the economics around AI are somewhat different than we've seen in the past. Right now, we have a couple of things. Number one, we are building fundamentally AI on the most expensive technology we've ever seen before. Not only the ability to collect the data, to train the data is incredibly expensive with incredibly high cost resources to go by doing that, i.e. data scientists and so forth. We're building that upon the most expensive factories to go by doing this. They are extremely expensive. They are very difficult to build. They are running into both economic and political challenges of going by doing this. They have to be cooled in a different way, they need energy in a different way, and a volume that is different than we've ever seen before. Then they run on top of the most expensive computing unit that we've ever seen before.