And if it can do that, why not use it design the perfect human (gulp)?

Agents learn from scratch how to see, act, cooperate, and compete in unseen environments, all from a single reinforcement signal per match:

Privacy filter for your photos is designed to thwart facial recognition technology.

Algorithm spots fake social media accounts based on the improbability of a user's connections.

We all knew that, but Google's health tech unit is using eye scans to detect heart disease.

Distinguishing poachers amidst animals in the jungles of Botswana.

The funny part is, even the bot can't get a perfect score - but seriously, this has real implications for NLP adoption.

So many applications for synthetic versions of YOU - or of the athlete you wished you were.

Reinforcement learning - the algorithm that trained the champ, from scratch.