Exclusive: Red Hat’s CTO sees thousands of AI models in the cloud’s future

Red Hat CTO Chris Wright knows a thing or two about adaptability. Growing up in a family that moved around a lot, Wright found himself confronted with wildly different environments, urban and rural, mountainous and flat, but he learned how to get along in each new location. That adaptability has helped Wright go with the flow in a rapidly-changing technology landscape. 

Silverlinings recently interviewed Wright to learn more about what it’s like to be CTO, where Red Hat is headed in 2024 and how artificial intelligence (AI) is set to change the landscape. 

According to Wright, the role comes with plenty of challenges. One is the expectation that he knows "everything in our portfolio down to the code, everything that’s not in our portfolio at an architectural level – you know, basically everything, which isn’t realistic." So, learning to say "I don't know" has been key. 

He said he's also often pulled in many different directions. To deal with this, Wright focuses on tasks where he can add the most unique value and delegates the rest.  

"I’m focused on keeping us at the nexus of technology trends, open-source communities and customer needs," he said. 

All eyes on AI

One technology that meets those criteria is artificial intelligence (AI). And according to Wright, AI is about to explode. 

He explained that while there are a limited number of models today, ultimately AI is not a one-size-fits-all solution for enterprise applications. In the future, each application will need its own AI model and in some cases, multiple AI models to serve different processes within the app. This sudden boom in the number of models could cause mayhem if not managed. 

“An enterprise today is built of thousands of different applications. I think we would be naive to expect that a future enterprise has thousands of applications and one [AI] model,” he told Silverlinings.

“There will be thousands of applications and thousands of models, possibly a multiplier on the number of models relative to the applications." 

He added these models will "be a combination of enterprise-built models as well as partner-provided commercial models." 

This, of course, has a few implications. 

The first is that enterprises will have to think about how to manage the sprawl of a new form of infrastructure in their business, he said. “We’re still really new to that space, so I expect to see a lot of work there,” Wright added. 

Open source rises … again 

The second point Wright made is that that open source will likely play a more prominent role in AI going forward. 

Wright isn’t the only one who thinks so. Cockroach Labs CEO Spencer Kimball recently tipped open-source solutions to emerge to meet demand from enterprises looking to avoid AI vendor lock-in. 

In its 2024 Enterprise Technology Outlook, market research company PitchBook predicted AI integrations will become table stakes for enterprise software-as-a-service providers and specifically forecast a wave of open-source AI innovation on the orchestration front. 

“Open-source momentum will likely take mindshare from closed-source advances,” PitchBook’s team wrote. “We expect to see game-changing new products built on top of open-source models in 2024, while closed-source developers continue iterating on their existing products.” 

Chris Wright cycling
Wright said cycling is one of his favorite hobbies in his off time. (Chris Wright)

They warned, though, that open-source projects come with the risk that they can “quickly become obsolete if the contributors shift their focus or users identify fundamental flaws.”

Where does Red Hat fit into all this? Well, just like it has done for the application landscape, Wright said it will provide the hybrid cloud platform needed for training, tuning, testing, deploying and monitoring models. As far as external use cases go, Wright said Red Hat sees plenty of opportunity in the telco, manufacturing, healthcare, retail, government and education markets.

It is also incorporating AI into its own products. Already, for instance, it offers Ansible Lightspeed with IBM watsonx Code Assistant.

"Infusing AI into other products in our portfolio is on our roadmap for this year," Wright said.

Growing pains

Wright acknowledged there will be growing pains as AI spreads. But he said people the world over have coped with massive technological change before and the industry can help smooth the path forward by keeping humans at the heart of everything.  

“We’ve had evolutions like this in the past where we did it manually and then a machine helped us,” he said. Wright pointed to the advent of calculators in math as one example. “It’s not as if we haven’t learned through these processes how to continue to be valuable as humans, but I do think it’s useful to think about how you engage people in the change process so they feel like they’re a part of something rather than being replaced by something.”

Chris Wright skiing with the Matterhorn in the background
Wright skiing with the Matterhorn in the background. (Chris Wright)

Guessing game

When he’s not busy at the office, Wright said he can usually be found doing some kind of outdoor activity – usually cycling, skiing or hiking – or spending time with his family. The latter includes two teenage daughters, both of whom are now (to his chagrin) driving.  

“My world is split into three key areas: what I do for work, my family life and then that sort of outlet, physical activity, and each of those feed a different part of who I am. Family is a really important part of staying human,” he said.

To be human, though, is to be flawed. Wright readily admitted he’s made plenty of wrong calls over the course of his career, but added he found the process of discovering “just how wrong you are” rewarding. A prime example? Social media.

“I couldn’t see anything very interesting about the early stages of social media,” he said, since humans already had other ways to communicate via calls, texts and email. “I didn’t appreciate what it was. I wrote that off…Boy, was I wrong.”

“The number of times I made an assumption about how interesting something was or wasn’t and decided ‘that’s not interesting and so I’m not going to put any time or effort into it’ and then later needing to learn more about that space and discovering it’s rich, deep, wide, fascinating, there’s a million things happening in that space that I never fully appreciated – I’ve always enjoyed that. It’s a dose of humility,” he said.

Those snap judgments haven’t always been about technology. A longtime Linux developer and engineer, Wright said that he always thought he had no interest in managing people.

Wright joined Red Hat as an engineer in 2005, making the leap from a similar role at the non-profit Open Source Development Labs (which closed in 2007). He initially focused on virtualization engineering before moving on to become Red Hat’s Technical Lead for Cloud Engineering in late 2009 and Technical Director of Cloud Engineering in early 2012. That last hop marked the transition from his role as a technology builder to more of an architect. He was later Red Hat’s Chief Technologist before being named CTO in October 2015.

He was wrong, by the way, about not being interested in managing people.

“Turns out people are fascinating. They’re dynamic, they’re complicated. Getting teams organized and motivated and building things together very effectively, it’s a non-trivial task,” he said.

The CTO concluded: “I always encourage people: Stay curious, stay open-minded...I’ve been wrong many times and that’s when you learn.”

Read more inspiring profiles of tech leaders here