
DeepMind's Demis Hassabis on the future of AI (from The TED Interview)
The Next Big Idea
00:00
The Dopemen Reward Mechanism in the Brain
Alfago is a computer programme that can learn to beat itself in real time. It was developed by the same team behind alfa zero, which played chess against itself and went from knowing nothing other than the rules of chess to being the greatest chess player ever lived. Alfago uses an adversarial model where you have two versions of the sophar playing against each other with different strategies. The system has also been used for games such as agnosticism or starcraft.
Transcript
Play full episode