When AI Makes Innovation Easier, Will We Lose the Ability to Innovate

Today at Stanford's Digital Economy Lab, I listened to Professor Daniel Rock of the University of Pennsylvania share how AI is changing scientific research and innovation, hosted by economist Erik.

Research shows that scientists using AI save about seven hours a week on average — yet scientific breakthroughs have not grown at the same pace. Why? Daniel proposed an intriguing "bottleneck law": innovation passes through four stages — defining the problem, finding solutions, validating results, and building consensus. AI has dramatically lowered the cost of finding solutions, but experimental validation, human judgment, and knowledge integration have not accelerated in step. The bottleneck has simply moved.

This reminded me of the Jevons effect in economics — why greater efficiency can lead to more demand and usage — which explains a seeming paradox of the AI era: as AI lowers the cost of innovation, we may generate more ideas, experiments, and products, yet truly valuable innovations may not grow proportionally.

What stayed with me was the discussion of the "Bananas Problem": when everyone asks AI what to research or what business to start, will the answers grow more and more alike, until innovation itself becomes mediocre? Daniel told his students: getting the right answer earns a B; getting it right and offering an insight different from AI's earns an A.

There's an even more practical question: young people used to grow into experts through large amounts of foundational work, mistakes, and reflection. Now that AI has taken over that work, how will the next generation develop independent judgment?

As an executive coach, I increasingly believe that when companies bring in AI, they shouldn't only ask how many jobs it replaces or how much cost it saves. They must redesign the path by which talent grows — so young people learn to ask questions, validate, and challenge, building judgment through real problems. AI can find answers for us, but which problems are worth solving, and why, remains a human responsibility. Innovation is not just technical capability; it is curiosity, independent thinking, and the courage to take a different path.