DeepMind Technologies is an artificial intelligence research firm established in 2010 and is currently a subsidiary of Alphabet Inc., the parent company of Google.
One of DeepMind's most famous successes is AlphaGo, the RL Go-playing machine that beat the current world champion and became the subject of a documentary. Another DeepMind RL agent, AlphaZero, taught itself to play and win not only Go, but also chess and shogi.
All of these machines, like the algorithms we've designed, used RL algorithms with deep learning architectures and taught themselves to play by starting with trial and error, then learning to model the games they were playing as state-action functions.
In December 2013, the DeepMind team developed and introduced deep Q-learning as an algorithm for solving optimization problems with better-than-human...
