Having a “human in the loop” is the typical recommendation for organizations using artificial intelligence for jobs in which there’s no margin for error.
But what happens when it’s just not possible to keep a human in the loop?As the use of AI ramps up within organizations and as businesses delegate more tasks to agentic AI, it’s a challenge that many corporate leaders are starting to grapple with.
“We’re seeing a material increase in the speed with which we can create things,” Zach Maybury, chief technology officer at online sports betting platform DraftKings said during a panel discussion at Fortune’s flagship technology conference, Brainstorm Tech, this month.
Maybury said his company already deals with trillions of transactions and highly distributed workloads. Once you introduce agentic AI into the mix—with AI agents talking directly to other AI agents—the volume and complexity of the operation is far too vast for traditional approaches.
“I can’t insert humans into those loops,” Maybury said. “We will never have enough humans to insert in all those loops.”
