Google DeepMind Trains 'Artificial Brainstorming' in Chess AI | Quanta Magazine
Briefly