⚙️ The First Reinforcement Learners

Before Q-learning, there was the Soviet “self-learning automaton.

When people talk about reinforcement learning, they usually start the story with Richard Sutton and Andrew Barto in the 1980s — or maybe with Q-learning in the 1990s.

But if you roll the timeline back 30 years and shift the map east, you’ll find something remarkable happening in Moscow, at the Institute of Automation and Remote Control (IARC).

In 1959, a …