The optimism trap

Listen to the predictions from this era.

In 1958, Herbert Simon wrote that within ten years, a computer would be chess champion, prove new mathematical theorems, and compose music of significant aesthetic value. In 1965, he predicted that machines would be capable of doing any work a man can do within twenty years.

He was wrong. So was Minsky, who said in 1970 that "within a generation, the problem of creating artificial intelligence will substantially be solved."

These weren't idle guesses. These were careful, serious people who had already built things that worked. They extrapolated from what they'd done and got the timeline catastrophically wrong.

Why? Because the problems they'd solved were, it turned out, the easy ones.