Nick spent his Sunday night training an AI agent using Reinforcement Learning in GoDot to conquer Brackey’s notoriously hard game—he found the inspiration in a mining-optimization RL demo and now sees a ton of potential for custom environments. The code (14 versions!) is up on GitHub, and he can share the hefty environment files on request.
Huge shout-out to Ed Beeching (GoDotRL) and Brackeys for the inspo! Nick’s flying to Florida tomorrow to present—hit him up if you’re around, or connect on LinkedIn, Facebook, GitHub, Patreon, or join the Discord for more coding fun.
Watch on YouTube
Top comments (0)