Now you can finally conquer all those vintage games you loved to play with a little help from artificial intelligence!
November 19th, 2018
Surfing the web in search of cool hacks and new games, we ran into a cool post published on technologyreview.com. It talks about how a new Python library provides a way to train a reinforcement-learning algorithms to play just about any old video game.
The library works as a wrapper around the popular game emulator MAME and the “readme” section shows how to write a quick program to master classic games, using Street Fighter 3 as an example.
But, what is reinforcement learning?
Reinforcement learning is inspired by the way animals seem to learn in response to positive feedback. DeepMind, the subsidiary of Google that aims to develop “artificial general intelligence,” famously used reinforcement learning to train programs to play Atari games. It was also the basis of AlphaGo, a program that proved capable of playing the ancient board game Go with superhuman skill (groundbreaking because the game is so complex and challenging to master).
The intersection between games and AI is an interesting one. While DeepMind popularized the idea of using games to benchmark progress in AI, it stretches back a long way. One of the earliest “AI” programs (although it was incredibly simple) was developed by the AI pioneer Arthur Samuel for playing checkers.
Reinforcement learning requires enormous amounts of data, and it’s often difficult to get it to work. Hence there aren’t many practical applications for the technology as yet. Still, it’s fun to see these games becoming accessible to reinforcement learning!
For the original article visit technologyreview.com