Harnessed reinforcement learning in GoDot to utterly dominate Brackey’s game—iterated through 14 versions, drew inspiration from an RL mining-optimization demo, and had a blast building out a custom environment. This project lays the groundwork for more RL adventures, especially in GoDot.
All code (and hefty 167 MB env files) is on GitHub, with shoutouts to Ed Beeching (GoDotRL) and Brackeys for the tutorials. Nick’s off to Florida next—hit him up if you want the dmg files or fancy geeking out over AI game-busting!
Watch on YouTube
Top comments (0)