Will GenAI Take Jobs? No, Says Docker CEO

Developers are no longer a backroom persona that gets thrown whatever and then have to manage themselves, said Scott Johnston, the CEO of Docker, in a wide-ranging conversation this Fall at DockerCon in Los Angeles.

Enterprise decision-makers realize developers make the architectural decisions for the application and the resulting tools needed to build them, which leads to billions of dollars of spending, Johnston said.

“And it took a while, but organizations have now connected those dots,” Johnston said, “And as a result, we’re seeing, at least in our business, a continued shift in spend that prioritizes not just a great developer experience, but a high velocity in how tools get developed.”

But what about this velocity, now with the advent of generative AI embedding into the developer experience?

“Our conviction is that continuing to deliver more and more products into that space will result in growth in the business,” Johnston said. “We’ll be able to grow the customer base, grow the customer spend with Docker, and hire more engineers, and just keep that loop going.”

He pointed to the use cases in play that were impossible at scale in the years before LLMs emerged.

“There are entire categories that are being formed as you and I sit here,” Johnston said. “And those categories are either right where we are, or they’re right adjacent to us, that we can provide product and great experiences.”

But will GenAI take jobs? Johnston says no.

“I also heard very authentically that GenAI doesn’t take jobs,” Johnston said about his experiences at DockerCon. “It’s going to take the repetitive toil and the repetitive code off developers’ plates so they can focus on ‘like, okay, how can I solve this problem more cleverly? How can I solve this problem more elegantly? And, wow, that’s a bigger problem than I had time to think about before. But now I have the time back because of GenAI to think about that problem.’ “So it’s speed and time, Alex, and I think they can be different sides of the same coin in some sense, right?”

GenAI, velocity — it will result in more developers coming into the market and more demand for apps, Johnston said. They’ll suck up that supply. It’s like building more bike lanes and making the experience better. Johnston used the highway analogy, but the notion is the same. Look at Amsterdam. They made it easy for bicyclists to ride throughout the city. The bike lanes are everywhere, separated from the street traffic. What’s the result when bike lanes expand and multiply and the experience is awesome? Bike traffic increases just as traffic does for cars when highways get built.

And now, with apps, there’s just more supply that comes with the ability to develop faster.

“So it used to be all developers were assembly coders,” Johnston said. “So when, when higher level languages like C came along, like, did all the assembly code go away? No, the market for C developers just blew up. And then Python came along and interpretive languages, it blew up further…every time you simplify with abstraction, it just expands the market and then on the demand side for applications.”

IDC reported in January that there’s a demand for more than 750 million apps in the next two years, more than written in the past 40 years.

“And we’re not surprised by that because every time you give powerful tools to a creative class, and developers are a creative class, they figure out fantastic, innovative ways to apply that we didn’t see coming. So no, I think it’d be a multiplayer, expander.”

More Episodes from DockerCon 2023

Debugging Containers in Kubernetes — It’s Complicated

Docker CTO Explains How Docker Can Support AI Efforts