"Human in the Loop" Might Be the Worst Way to Use AI, with Vivienne Ming

"Human in the loop" is supposed to be the safe, responsible way to use AI. Turns out it might be the worst.

Theoretical neuroscientist and seven-time founder Dr. Vivienne Ming joins host Greg Shove on Supercompanies to break down what happens to human intelligence when AI can answer almost anything.

In an experiment with UC Berkeley students using AI to forecast real-world events, Vivienne uncovered three distinct ways people work with AI: automators who just accept whatever the AI tells them, validators stuck in a "human in the loop" pattern that performs worse than either humans or AI alone, and a rare group she calls cyborgs - people who blend with AI so effectively you can't tell who made the decision, and who came close to matching expert-level forecasters.

The traits that predicted who became a "cyborg" - curiosity, intellectual humility, and perspective-taking - raise bigger questions about how we raise kids and design schools in the AI era, which Vivienne explores in her new book, "Robot-Proof."

Supercompanies is back with all new episodes!

Watch the Video:

https://www.youtube.com/@SupercompaniesPod