Dopamine: Improved Reinforcement Learning and… Pleasure for Machines?

Great news: on the 27th of August 2018, Google announced via their AI blog that they were releasing a new framework, named Dopamine, for ‘Flexible and Reproducible Reinforcement Learning Research’.

Not jumping from joy yet? No worries, we’ll go through what this means, what this entails, and hopefully we’ll all have that reaction.

Reinforcement Learning

Here’s a quick recap to get everyone on the same level of understanding before we dive any…