Modern societies are obsessed with innovation. In June 2015, Google searches returned 389 million hits for “innovation,” easily beating “terrorism” (92 million), “economic growth” (91 million), and “global warming” (58 million). We are to believe that innovation will open every conceivable door: to life expectancies far beyond 100 years, to the merging of human and machine consciousness, to essentially free solar energy.

This uncritical genuflection before the altar of innovation is wrong on two counts: It ignores those big, fundamental quests that have failed after spending huge sums on research. And it has little to say about why we so often stick to an inferior practice even when we know there’s a superior course of action. The fast breeder reactor, so called because it produces more nuclear fuel than it consumes, is one of the most remarkable examples of a prolonged and costly innovation failure. In 1974 General Electric predicted that by 2000 about 90 percent of the United States’ electricity would come from fast breeders. GE merely reflected a widespread expectation: During the 1970s, the governments of France, Japan, the Soviet Union, the United Kingdom, and the United States were all investing heavily in the development of breeders. But high costs, technical problems, and environmental concerns led to shutdowns of British, French, Japanese, U.S. (and also smaller German and Italian) programs, while China, India, Japan, and Russia are still operating experimental reactors. After the world as a whole has spent well above US $100 billion in today’s money over some six decades of effort, there is no real commercial payoff.

Other promised fundamental innovations that still are not commercial concerns include supersonic passenger flight, magnetic levitation trains, and thermonuclear energy. The last one is perhaps the most notorious example of an everreceding innovative achievement.