Bend's website demo: an AI-coded game you mathematically cannot win
alejandroll10 · x · 2026-09-10
Victor Taelin shared a live demo on the Bend language website: a game with one rule — you can't win. Visitors can prompt an AI chat to change the game in an attempt to win, but the AI is mathematically unable to introduce a bug, and by the rules winning is a bug.
Examples of prompts:
- "Remove the wall on the right" → AI puts it on the left
- "Let WASD teleport 2 squares" → still blocked by walls
- "Let it pass through walls" → walls become 2 tiles thick
- "Remove the walls" → WASD now moves the flag
No matter what you prompt — even jailbreaks or threats — you can never touch the flag, because committing the code requires the AI to prove the flag is unreachable. Taelin notes that people unfamiliar with proofs are skeptical, but that's how it works.
More from coding & agent
- OpenAI staff ditch Google Forms, vibe code custom forms on ChatGPT Sites — sherwinwu · 2026-09-10
- Turing launches CEO Bench: 500+ expert tasks test frontier agents in a simulated company — ecekamar · 2026-09-10
- Seroter's daily list: 5 things you shouldn't vibe code, and trusting agents with production secrets — rseroter · 2026-09-10
- Notion CEO-shared take: own your context, rent the intelligence — ivanhzhao · 2026-09-10
- Voice in the Codex app is great for math study, but bugs pile up in heavy use — SIGKITTEN · 2026-09-10
- We gave AI agents browsers, now they can just call people — Illustrious-Fig-326 · 2026-09-10