Building a game and training a PPO agent with Grok
tetsuoai · x · 2026-09-01
The author used Grok to build a browser game called Roofline, then used Grok again to train a PPO (Proximal Policy Optimization) agent to play the live page. The best recorded run achieved a score of 39,359, a distance of 4,924 meters, and a combo of 117. A link to the game is provided.
Related event: Dev Uses Grok to Build a Game and Train a PPO Agent to Play It(2 posts)→
More from coding & agent
- Hands-On Workshop: Build an LLM Wiki as Long-Term Memory for Your Agents — Al_Grigor · 2026-09-01
- Burned 1 Billion Tokens Building an Agent Client, Got an Admin Panel Instead — sujingshen · 2026-09-01
- User claims Codex is far ahead, calling a hyped coding bot overrated — demian_ai · 2026-09-01
- Teknium says don't worry about PR rebasing, team will salvage code — Teknium · 2026-09-01
- CyberStrikeAI: Open-Source AI-Native Cybersecurity Command Center — Shruti_0810 · 2026-09-01
- Using Claude to Plan Outdoor Robot Plant Segmentation Research — _Stocko_ · 2026-09-01