Testing leads to failure, and failure leads to understanding.

There was a wonderful little short four-year time period when marvelous things happened. It started in 1908, when the Wright brothers flew in Paris, and everybody said, 'Ooh, hey, I can do that.' There's only a few people that have flown in early 1908. In four years, 39 countries had hundreds of airplanes, thousands of pilots.

When there's ever a breakthrough, a true breakthrough, you can go back and find a time period when the consensus was 'well, that's nonsense!' so what that means is that a true creative researcher has to have confidence in nonsense.

Space travel is the only technology that is more dangerous and more expensive now than it was in its first year. Fifty years after Yuri Gagarin, the space shuttle ended up being more dangerous and more expensive to fly than those first throwaway rockets, even though large portions of it were reusable. It's absurd.

NASA's myriad failures are in many ways the natural consequence of a catastrophic combination of bureaucracy, monopoly, and a calcifying aversion to the kind of risk necessary for innovation.

The most impressive airplane ever, I believe, was designed only a dozen years after the first operational jet. Stayed in service till it was too rusty to fly, taken out of service. We retreated in '98 back to something that was developed in '56. What? The most impressive spaceship ever, I believe, was a Grumman Lunar Lander.