A principle in artificial intelligence and robotics stating that tasks humans find difficult, like playing chess or solving calculus, are computationally easy for machines, whereas tasks humans find effortless, like walking or recognizing a face, require staggering amounts of computational power. It highlights the evolutionary complexity of human perception and mobility compared to abstract logic.
Heard on the job
Trying to get this bipedal robot to walk across a gravel driveway is a perfect example of Moravec's paradox in action.