If AI removes traditional constraints on execution, what becomes the limiting factor? AI can now generate, summarize, and analyze, but it still can’t judge. And as many business leaders are learning, that can cause a new wrinkle in workflows.

On this episode of Where AI Works, host Matthew Bidwell is joined by Ritcha Ranjan, Senior Vice President of Product at Expedia Group, to explore what happens as AI takes on more and more tasks traditionally handled by humans, forcing employees to decide when it’s actually doing things well. After all, when AI can produce endless ideas and recommendations, the real skill becomes knowing which ones to trust. Ritcha believes skepticism isn’t a bug, it’s a requirement. She explains why building effective AI systems means designing for human judgment: knowing when to keep people in the loop, how to validate outputs, and how to create feedback systems that continuously improve performance. She also shares how her teams at Expedia are experimenting with always-on AI systems, reworking product development, and preparing employees for a world where deep expertise, curiosity, and judgment matter more than ever.

Episode Highlights:

6:25 - Ritcha shares the top three misconceptions some business leaders still have about AI.

9:11 - Ritcha outlines the three-stage evolution most organizations go through when adopting AI — from “help me” to “create for me” to “run this for me.”

12:18 - Ritcha breaks down the risks/rewards of AI implementation, and how the equation changes the further down the road your company gets.

20:06 - Ritcha discusses the skills needed to thrive in an AI-forward environment, and explains how productivity doesn’t come from more output but from better judgment.