RL is more information inefficient than you thought

Since it is not explicitly stated, "RL" in this article means Reinforcement Learning.

https://en.wikipedia.org/wiki/Reinforcement_learning