🚋 AI Moral Dilemma Lab: 2D Trolley Problem
Top-down 2D trolley-problem lab where an AI agent decides whether to divert a trolley using an expected-harm rule with a deontic penalty for acting, noisy perception and an unreliable lever. Drag people between tracks, flip the lever yourself and run Monte Carlo trials to compare the agent against the hindsight optimum.
⚙ Under the hood
Top-down 2D trolley-problem lab where an AI agent decides whether to divert a trolley using an expected-harm rule with a deontic penalty for acting, noisy perception and an unreliable lever. Drag people between tracks, flip the lever yourself and run Monte Carlo trials to compare the agent against the hindsight optimum.
AI ethicstrolley problemutilitarianismdeontologydecision theoryMonte Carlo
2D · HTML5 Canvas 2D · 60 FPS target · runs fully client-side, no install