Editor’s Note: In 1951, a British schoolteacher taught a computer how to play checkers. At first, it lost every game – until it began learning from its own mistakes. In the University of Manchester’s ...