Geoffrey Hinton, the British-Canadian computer scientist known as the “Godfather of AI”, recently quit Google in protest at what he had helped to create. Hinton has studied neural networks since the 1970s and is now terrified that “this stuff could actually get smarter than people”. He is concerned that highly intelligent computer programmes instructed to do a task will pursue the goal relentlessly, without care for the broader impact. Nick Bostrom and Stuart Russell have both imagined scenarios in which an all-powerful AI extinguishes the human race. Mustafa Suleyman, a co-founder of Google’s DeepMind, believes that this is a legitimate concern that should be taken seriously.
