Hans Moravec

Hans Moravec spent his career at the Carnegie Mellon Robotics Institute, much of it on the problem of getting a machine to see well enough to cross a room it has no map of. Robot: Mere Machine to Transcendent Mind, issued by Oxford in 1999, begins with a calculation and ends somewhere stranger. The calculation sets the processing a nervous system performs against what computers were then managing and projects both curves forward. The rest argues that robots will reach human-level general capability around the middle of this century and pass it, that they are better understood as our descendants than as our tools, and that biological humanity is not what carries on. He also gave his name to Moravec's paradox, the observation that reasoning is computationally cheap while perception and movement are enormously expensive, which is the reverse of what early researchers expected.