#moravecs-paradox

Question 1: Single choice

mediumAI-checkedAI Foundations & History

Early AI programs proved theorems, solved geometry problems and played decent chess, but could not recognize a face or cross a room without bumping into things. What did this contrast reveal?

Tasks easy for people, like perception and movement, can be the hardest to program with symbolic rules1
Chess and theorem proving need far more computation than vision, so they should have been solved later2
Robots of the time lacked batteries strong enough to move around3
Perception becomes easy once enough logical rules have been written down4

Timed test