Most conversations about AI at work treat it like a software rollout: Plug in frontier models, and expect the company will automatically move faster. Leading AI transformation across hundreds of thousands of people at Google has taught me otherwise. The technology is scaling rapidly—in fact, we’re seeing breakthrough capabilities emerging daily—but real value only moves as fast as people adopt and adapt how they work with these ever-evolving tools. When you bolt agents onto traditional silos and multi-stage approval, you don't transform your business—you just reach the bottleneck faster.
For decades, companies ran like assembly lines, passing tasks down a chain of narrow specialists. Agents break that logic. In our internal labs, small, cross-functional teams are using agent coding swarms to rebuild complex software and systems in a fraction of the usual time, and we’re piloting multi-disciplinary roles that unify development, product, and design workflows. This more fluid style of working empowers our teams to own and drive projects end-to-end.
The real breakthrough happens when cutting-edge technology meets a ready culture. At Google, we lean on our foundational growth mindset—the belief that curiosity and continuous learning will always outperform static expertise. It takes a massive commitment to upskilling, helping every employee shift from executing within the boundaries of a narrow focus area to orchestrating systems with judgment, taste, and curiosity. And it asks something new of leaders: creating the psychological safety to let go of old routines, and reinvesting the hours saved into bigger, bolder bets.
Frontier technology sets the floor, but we’re learning that structure, fluency, and leadership will set the ceiling.
