A while back, during one of Meta’s new hire orientations, a senior manager froze when new hires asked him about "vibe-coding." He nodded politely but clearly had no idea what they meant. That moment exposed more than a generational gap, it was a warning; The gap between how fast technology evolves and how fast leaders adapt is widening, and we must do something about it. ASAP.

Junior engineers now arrive with AI superpowers. The competitive advantage shifted from raw technical skill to adaptation speed, from code quality to communication, and from knowing frameworks to catching AI hallucinations. This forced Meta to rethink how we hire, evaluate, and lead.

We built AI-native coding interviews that measure adaptation, not memorization. We learned to evaluate talent without the traditional technical signals, and shifted the skills we look for when hiring, promoting, and assessing performance.
We also changed how we lead: leaders have to use AI themselves before teaching their teams. They learn to adapt , rather than using specific tools, so they can evaluate whatever shows up next. They needed to learn how to leverage AI without the price of dependency. We realized AI accelerates junior engineers faster than any technology we've seen, but despite the panic about job losses, AI can help ambitious engineers break in and move up, but only if their leaders understand how to guide them.

In this talk, I'll share what we learned, and the frameworks we developed. I’ll share the mistakes we made, and what actually works when there’s no playbook. You'll leave knowing what skills matter now, how to evaluate talent differently, and how to lead when yesterday's answers don't apply anymore.