In 1914, Spanish inventor Leonardo Torres y Quevedo unveiled a contraption he'd spent years perfecting. Nicknamed El Ajedrecista ("the chess player"), the machine had one enormous key difference from earlier chess 'machines' like the 18th-century phony automaton known as the Mechanical Turk: it required absolutely no human input to beat its opponent. 

Quevedo and his automaton

Designed to simulate the end of a chess match, it could handle only three pieces: a king and rook manipulated by the computer, and a king controlled by a human opponent. Using an algorithm, the computer would then calculate how to checkmate its opponent, no matter where the three pieces had been placed on the board. In the first version of the game, mechanical arms then moved the pieces for the computer; a 1920 update designed by Quevedo's son used magnets to make it appear that the pieces moved on their own.

With limited gameplay, Quevedo's invention wouldn't necessarily have been very fun for a human player. Still, it was hugely powerful for what it was: an elegant statement of the potential of computing technology.