Artificial intelligence is not simply layering on top of existing technology infrastructure. It is dismantling it and demanding a rebuild, according to Steven Dickens, CEO of HyperFRAME Research.
"It's making a fundamental change all the way up and down the stack, whether that's compute infrastructure, networking, operations, the way applications are deployed and managed, Kubernetes, virtual machines, literally all the way up the stack," Dickens said in an interview with The Forecast recorded in April at Nutanix’s 2026 .NEXT conference in Chicago.
"All of that stuff that I've kind of grown up and known over the years is all being thrown up in the air and we're almost relearning the entire stack."
He sat down with The Forecast to assess how enterprise AI is reshaping infrastructure, partnerships and the day-to-day demands placed on technology teams. His perspective carries particular weight: he spent years at IBM before founding HyperFRAME Research, a firm that tracks innovation across the IT industry.
The view he offered is one of cautious optimism grounded in historical pattern recognition. AI is generating enormous enthusiasm and investment, but the real transformation, Dickens argues, is still in its earliest stages.
Dickens draws from The HyperFRAME Research Lens: The State of the AI Stack, which reported on the first half of 2026 survey findings from 544 qualified enterprises. Only 14% classify their core data architecture as fully modernized for AI workloads today, while 23% are still running a legacy on-premises data warehouse.
"We're still so early," he said. "It's the first innings of a seven game World Series for me."
Cloud computing has evolved for 20 years to reach its current form. Kubernetes, the container orchestration platform now regarded as critical enterprise infrastructure, has been evolving for just over a decade. The iPhone, which redefined mobile computing, shipped in 2007 and continues to shape how enterprise applications are conceived and built. Dickens places AI on that same arc that will take years to fully resolve.
The maturation of Kubernetes is central to his argument. What began as a project embraced primarily by tinkerers and cloud providers has evolved into production-grade enterprise infrastructure. That evolution, Dickens contends, positions Kubernetes as the natural home for the wave of AI applications enterprises are now planning to deploy.
"I think AI is going to happen on Kubernetes from an app deployment perspective," he said. "If an enterprise has got a thousand apps today, they're going to have 300 new apps. They're all going to be agentic AI apps. I don't see those going onto bare metal or onto virtual machines. I see them going on to containers."
The ecosystem around AI infrastructure is undergoing its own consolidation, explained Dickens.
"The ecosystem's having to coalesce and reestablish itself around a new norm, which is this AI factory [GPU clusters designed for AI training and inference] and how we deploy," Dickens said. "We've thrown the stack up in the air. We're now watching those partnerships and all of that ecosystem start to coalesce."
His projection: the industry is roughly a year into that realignment, with another year of positioning before the focus shifts decisively to execution. For enterprise IT leaders, that window represents both an opportunity and an obligation.
Video transcript:
Jason Lopez: Artificial intelligence moving faster than any transition that came before it. In the past, a few shifts fundamentally changed information technology. From the mainframe to the personal computer to the internet, these shifts have reshaped how we communicate and how we do business. Experts say AI is rewriting the stack all over again.
Steven Dickens: It's interesting how all of us everywhere up and down the stack have all had to become AI-savvy analysts over the last two or three years. It's making a fundamental change all the way up and down the stack, whether that's compute infrastructure, networking, operations, the way applications are deployed and managed, Kubernetes, virtual machines, literally all the way up the stack, MCP, server deployments, how that's everything my entire landscape that I've known for far too many years. There's some gray in the hair and a little bit less of it. All of that stuff that I've kind of grown up and known over the years is all being thrown up in the air and we're almost relearning the entire stack. So it's a wild time to be an analyst. It's a wild time to be tracking space, but super excited.
Jason Lopez: From his years at IBM to founding Hyperframe Research, Steve Dickens has built a reputation for seeing across the entire technology stack from mainframe to cloud to now AI.
Ken Kaplan: Seems like you're going to have to have a lot of trust in people or you have to be really smart and it's a combination of all of those things. How do you feel like people are getting on with it?
Steven Dickens: I mean, so how do we keep up to speed as on this? We're really blessed. The vendors are always giving us briefings, getting on our calendars and giving us fantastic access to speak to their best and brightest. I mean, AI helps us from a perspective of being able to synthesize data and pull data together really quickly and be able to sort of ask and interpret and fact check and be able to sort of build a three-dimensional view of what we're being told. How does this relate to that, to this, to that? That would have been a multi-hour sort of Google search type journey previously. It's an AI journey. Taking it outside of me as an analyst and thinking about what other people are keeping up, I think we're still so early. Still so early. It's the first innings of a seven game World Series for me.
If you look at the cloud, it's taken 20 years for that to evolve. Look at Kubernetes, what are we on? 13 years of that, 13, 14 years of that. You look at mobile, what the iPhone came out 2007. These are still trends that are shaping the industry today and we're at two, three years into this with AI. So it's that level of tectonic shift and that's how early I think we still are.
Ken Kaplan: Let's talk about Kubernetes. Is it becoming mainstream now? And then what's that role in the rise of AI now?
Steven Dickens: So I think it's gone from Tinkerer's project. It was kind of in the cloud providers and kind of people were playing with it. Where I've seen over the last three or four years, I talk about the hoodie to smart coat ratio of how many people are kind of involved. We hear now talk about platform engineering. These are grown up terms that you're talking about when you talk in the context of putting a workload into production for a mission critical application. I think Kubernetes is there now. Then when you look at the confluence of that with AI, I think AI is going to happen on Kubernetes from an app deployment perspective. We're going to see some bare metal. We're going to see some virtual machines. But I think if an enterprise has got a thousand apps today, they're going to have 300 new apps. They're all going to be agentic AI apps.
I don't see those going onto bare metal or onto virtual machines. I see them going on to containers. That to me, so you look at AI and you look at the maturity of Kubernetes, it's perfect, perfect timing. If we'd have had AI come along and Kubernetes was two or three years old, we wouldn't have seen the same explosion. But we've got a really robust and mature way to deploy all these applications at scale, both cloud and on - prem. So it's a perfect timing for me.
Ken Kaplan: We've seen DevOps and we've seen containers, but you have to manage it. You have to be buttoned up. Can you just describe what's happened in the past year or two around that?
Steven Dickens: I talked about platform engineering before. I think it was a tinkerer's project, a Kubernetes cluster, probably up until four years ago I would say. Maybe I'm doing that at this service, but that's how it certainly felt to me. I go to CubeCon, I speak to the Red Hat guys, the Susa guys, the Marantis guys, the canonical guys. I chat to a bunch of these guys and it felt always to me like a tinkerer's project. That's hard pivoted and changed. Platform engineering for me is looking at non-functional requirements, security, availability, performance, scalability. Things that a tinkerer doesn't think about. They're like, "How can I just get this out fast?" We've moved beyond that to now it's got to not only be delivered fast, that's remained. Is it production ready? Can we secure it? What about backup? What about availability? What happens when something goes bang in the middle of the night and we need to DR over it? How are we going to manage this? How are we going to build skills around it? Those are grown up boring things, but they're absolutely vital if you're going to have this as a platform to support next gen AI apps and mission critical apps in the enterprise. So that for me is how that's all evolved.
Ken Kaplan: There are a lot of partnerships it feels like so - and-so's investing in the other one. I've been covering it for over 20 years, the industry, and I haven't seen it so frenetic.
Steven Dickens: Yes. I mean, I think where we see, if I was looking ahead, I think we've thrown the stack up in the air. We're now watching those partnerships and all of those ecosystem start to coalesce. So we've seen it this week with Nutanix, partnerships with NetApps and Everpure and Lenovo in the storage space.That would have been unheard of four years ago when you were a HCI focused company. We're seeing that across the stack. Nvidia's Rise and everybody at GTC clamoring to get Jensen to sign one of their sort of racks or servers. The ecosystem's having to coalesce and reestablish itself around a new norm, which is this AI factory and how we deploy. So for me, what's fascinating is how all that comes together. I think we're probably a year into that ecosystem sort of the strata in that ecosystem starting to be established. I think there's still about another year of that to go and then it's going to be execution.
Related:
What Is an AI Agent Gateway? How Enterprise IT Governs Agentic AI at Scale
Dual Native Runs Containers in Virtual Machines or Bare Metal
Memflation and the AI Supply Chain: How IT Leaders Are Adapting
How Kubernetes Catalyzes Enterprise IT
Dual Native Runs Containers in Virtual Machines or Bare Metal
ECI 2026 Report Shows Strain Between AI Innovation and IT Governance
AI's Defining Decade: Nutanix Leaders on Inference, Agents, and the Road Ahead
Jason Lopez is executive producer of Tech Barometer, the podcast outlet for The Forecast. He’s the founder of Connected Social Media. Previously, he was executive producer at PodTech and a reporter at NPR.
Ken Kaplan contributed to this article. He is Editor in Chief for The Forecast by Nutanix. Find him on X @kenekaplan and LinkedIn.
© 2026 Nutanix, Inc. All rights reserved. For additional information and important legal disclaimers, please go here.
Related Articles
