Fuzzy Logic Evaluation
summary
Argues that fuzzy logic replaces brittle if/behavior-tree rules by mapping environment signals (health, crowding, comfort, temperature) to 0–1 via a response curve, combining them with chosen conjunction/disjunction operators (product or min, sum or maximum selection) and either picking the max or sampling by weighted probability for unpredictability. Claims this scales to real‑time crowds: an open‑source MMO prototype uses the system for all NPC decisions and runs efficiently enough to produce natural, non‑hardcoded group behavior.
AI-written summary · about
related talks
54mExperiments in Reasoning -
14mHigh fidelity modelling of traffic light control with xml logic representation
15mCurrent Practices for Building LLM-Powered Reasoning Tools Are Ad Hoc—and We Can Do Better
40mSystems Thinking for Developers
47mBounded rationality, complex Systems and Agile principles
34mLogic in Computer Science, Engineering and Industry
23mSpecifying and Verifying Future Conditions
43mThe Evolution of Design Patterns in the Age of AI