"The real sin is if we miss something."
Q. How do you measure the return on R&D investment?
I don't know, in any particular year, which parts of our R&D spending have been the most fruitful.
We have a very unusual model in that if I say that something is going to be important . . . like I say that TV in the future will be delivered over an IP network — you know I said that 12 years ago. Well, about eight years went by and nothing happened. Microsoft was spending more than $100 million a year on that R&D, and we were ahead of our time. Well, the market will tell us if we are ahead of our time on that or not, but that's okay. The real sin for us is if we miss something.
Q. So you can afford to be patient?
Yes. The payoff when you get something right is enough that being a few years early is not a bad thing.
Like pen computing. In the 1990s the venture capitalists all said, "We're not putting another dime into that." The time horizon was too long. But the payoff, once you get it right — [someday] every student will be taking notes on a tablet.
Q. Microsoft is often called a "fast follower."
[Look at] RealOne: We shipped a media player before that company was founded, period.
Google didn't [do search] first: AltaVista, guys who worked for us, came first. We did many things where we were first. Then we have other areas where we do catch-up, like the browser.
Q. Is one strategy better than the other?
Oh, it's always better to be first.
Actually, it's very expensive. Internally, people give me no end of grief about things like TV. We'd rather have more competitors investing in long-term research. It would be healthier for the industry if more people would do that. And that might cause us not to be first on as many things. [But] the things that we've been really successful in, we were first.
Q. Green field or incremental innovation: Which is more important?
I'm afraid that taxonomy is pretty tough.
What about speech recognition? In the 1960s, people did 20% accuracy speech recognition. And the things we've done over a period of years to get to 30%, 40%, 50% — is that incremental? Our speech server absolutely is a collection of about 5k incremental improvements in speech recognition that lead to a paradigm shift of how you interact with the computer. Most breakthroughs are a compilation, where a ton of incremental improvements come together.
Q. As a society, have we gotten better at innovation?
There have been lots of improvements.
Innovation builds on itself in the sense that the Internet and the PC are themselves tools of innovation that allow far more rapid innovation than ever before. [But] most people in the world live in very poor conditions, and our ability to focus innovation on the things that would save lives is extremely poor. For people living in the United States, we've got erectile- dysfunction medicines and cell phones with nice color screens. There is a very inequitable spread in terms of how creativity is being applied. We're doing better than we've ever done, and people are healthier and better off. But are we really making the advances to improve human conditions where most people live? Not even close.