Exploration and evolution: Cloudera’s Brian Martin on AI’s path forward

Should IT executives be thinking about AI in philosophical terms? Or as the next step in challenging the technological frontier?

Why not both? That’s how Cloudera CTO of Applied AI and VP of Industry Solutions Brian Martin sees it. Drawing on his own background, Martin believes that, as we manage the early stages of an AI revolution, we need to think about the “how” and “why” of AI’s role in the tech stack.

We sat down with Martin at Cloudera’s EVOLVE26 conference on October 27 and talked about his personal history with AI, how we’re in a modern sci-fi world, and how to move down the right path.

How long have you been interested in AI?

My undergraduate education in the ’90s was in cognitive science, linguistics, computer science, philosophy—AI without being called AI. The things that we talked about as science fiction back then over beverages are now becoming reality!

The step beyond that, this cognitive enterprise that I talk about, the company that acts as an organism and that responds organically—that is still to come, and we are only starting to approach the point where you can think of the transition.

What do you mean by transition—what are you expecting?

Instead of the people telling the computers what to do, it’s going to be the computers and the systems asking the people what they need to do.

Whether you’re in pharma, and a computer system tells a scientist, “You need to go in the wet lab and synthesize this drug to test it,” or whether it’s at a law firm where the system is going to say, “You need to discuss with the client the potential exposure because of this scenario that is evidenced through case precedent law,” in every industry we see the opportunity where the company [is] being the organism.

There’s a sense that AI is the Wild West, and there’s an open lane for exploration. Is that how you see it?

The Wild West approach enables a lot of exploration, but to deliver the maximum path to the end, you have to move from exploration to exploitation, and not in a pejorative way.

When the wagon trains went to cross the Mississippi River and move west in the United States, how did they settle on the Oregon Trail? The specific paths that took them there, exploration died—I think it’s, effectively, evolution.

An evolution of AI pathways?

The strongest or the easiest path creates that path forward. And I think that that’s what we’ll see [with AI]. The benefit of this exploration allows us to clearly identify here is the safest, the optimal, most efficient path forward, and that’s what I hope everything will eventually consolidate down to.

That’s where we start to say, “Okay, now we need regulation. If you’re going to operate outside this path, here are the controls for risks you need to have in place.”

But if you’re on this path, we’ve already clarified what you need, you already know what you need. You can move as fast as you like.