Seismic Shifts from Traditional IT to Hybrid Cloud to Enterprise AI

Hybrid cloud and hybrid multicloud are widely accepted IT models used across industries, reflecting the reality that data is distributed across various locations rather than centralized. But Jean S. Bozman, president of Cloud Architect Advisors, sees big challenges ahead for business decision leaders and IT teams.

“The AI piece is just upending everything,” Bozman said in this video interview with The Forecast. 

Analyst Jean S. Bozman, president of Cloud Architect Advisors, explains how evolving traditional IT to hybrid cloud and enterprise AI forces IT teams them to manage complexity, acquire new skills and manage a myriad of costs. 

Bozman explained that the waves of innovation from traditional to hybrid cloud and now to enterprise AI are pushing IT teams into new territories, forcing them to manage complexity, acquire new skills and manage a myriad of costs. 

She said many IT teams must learn how to integrate their existing operations with new distributed and cloud native environments. AI, emerging from open-source initiatives, adds another layer of complexity but also opens new opportunities. 

The bottom line: “Organizations need to find a way to mix in AI with everything else they’ve been developing over the past decade,” she said.

As a tech journalist and analyst, Bozman has seen several generations of innovation upend the apple cart. She’s had a front row seat at the biggest technology events through the years, including those of Apple, Amazon Web Services (AWS), Google, HP and HP Enterprise, IBM, Intel, Microsoft, Oracle, Red Hat and NVIDIA. She has interviewed many tech leaders over two decades, and written about the evolution in IT hardware, software and customers’ software-defined infrastructure. Her analysis shows the combined impact of these technologies on today’s new and emerging business cases. Her commentary crystalizes trends, bringing CIOs and IT professionals closer together about identifying new products and services that will become critical to their business.

“I think AI was almost a shock to people,” she added. “All of a sudden, it’s in your face. You have to do something. The hard part is harmonizing the data from different sources. You can do it at the edge, you can do it in the cloud, you can do it from your office, but it has to be harmonized so that it doesn’t drag the whole network down.”

“People are going to have to work together more, which is why we’re seeking a more unified environment. When we ride on the highway, that’s a unified system. We all have to agree: left lane, right lane, switching, signaling, all those things. If everybody is driving in their own way, it’s going to be pretty chaotic.”

Transcript:

Jean Bozman: It's amazingly well accepted. It's not controversial. Hybrid multi-cloud. You cannot almost go to a tech person that wouldn't agree with that statement. There've been different terms. There was hybrid cloud than they said because there were multi-cloud. It's hybrid, multicloud. I think it's an admission. That data is everywhere. And so I've distributed data and that was fed by the fact that we had the hyperscaler guys. Some people call 'em CSPs hyperscale, where the big guys know who they are. And the data is not just sitting in one monolithic place, it's sitting here, there, and everywhere. And even the people who used to think that it would just be on one or two systems, it is not that way anymore. Anyway, so it is a fact of life, but the question is how do you live with it? It's that way. How do you live with it? How do you get the two things that have been running in parallel, which is distributed, we know that's happening, but also the traditional enterprise stuff. How do you get those things to sync up and work together? And I think a lot of what we've heard here is about how do you get those enterprise capabilities, manageability, all the things we've always had to have and how do you make that happen in a world that may otherwise seem chaotic?

This is a challenge for the enterprise where people say, Hey, that's not how I've been doing things for 20 years, or Hey, I didn't learn how to do this, or I need these new skills. So there is a skills challenge and there's also a cost challenge because we know what we were doing and the thing we are doing seems additive, but now we really want to kind of embrace everything and do it all at the same time. People were living separately, right? You had the people who always advocated for distributed and lots of little databases here and there, and that was cool. Especially forgive me for my sins, but the open source movement. So fun. But it's okay to have fun, but you also need to run a company. So you need to put those business best practices. I want to say the best practices to work and that's what could be a chaotic environment, but should be a smoothly running, unified environment.

A lot of the AI stuff comes out of the same kind of people that have been doing the open work, but it's applying, I think a new rigor to it. And also there is the demographic piece. I'm absolutely going to talk about that. Where traditional IT has been running for, what, 40 years or something like that. A lot of these things are a sudden shock to systems that have been reliably in place when you talk about let's do backups, let's do security. It was done a certain way. All that's having to be adjusted, I think, and that's clear to people.

AI was kind of, I don't know, it was almost a shock to people. All of a sudden, it's in your face now what do I do? I've got lots of data. I've got these models. I know other people here. We're talking about the models and where they are and who's going to do it. And there's a very big difference in training the models, which takes up a lot of CPUs, GPUs, software. But there's also the inference stuff, and that can be customized to your business. Healthcare, retail, finance, it's customized. And also it's easier to get your arms around it than trying to take in everything that was ever written just to write a haiku. That's not an enterprise thing by itself. It can fit into the enterprise. People writing things, people being creative, but it's not enough. There's a lot of rigor that has to come around that controls guideposts and all of that.