Samuel J. Palmisano

IBM isn't investing billions of dollars every year into research and development - and winning more patents than our top 10 competitors combined for more than a decade - as an academic exercise. But research is now being driven much more by what people need rather than just by what is possible.

Early in the 1990s, I flew alone in a dandelion-yellow, single-engine, 180-horsepower Piper Cherokee from Westchester County Airport in New York westward to the Rocky Mountains, landing and refuelling a good many times in middle-sized cities and towns along the way.