Integrating ASP-based incremental reasoning in the videogame development workflo...
summary
Integration of Incremental-DLV2 into ThinkEngine for Unity wires a live, multi-shot ASP solver to reactive and planning NPC modules and three game benchmarks (Frogger, Space Invaders variant, Pacman). Experiments show average per-shot speedup versus single-shot DLV2, with the Pacman emergency planner running >2× faster (decision loop under ~0.3s vs >0.3s), but incremental grounding causes uncontrolled memory growth for very fine-grained, pixel-level encodings, so an abstraction layer and sensing optimizations are needed.
AI-written summary · about
chapters (9)
- 0:00ASP in video game AI
- 3:55Declarative tools for game AI
- 5:22Think Engine and DLV2
- 9:07Incremental ASP reasoning
- 15:14DLV2 in Think Engine
- 17:43Benchmarking classic games
- 20:12Results and performance
- 22:23Lessons and future work
- 23:28Q&A
related talks
52mThe Journey of AI in Games: A Technical and Philosophical overview
20mStudio Speaks: The Reality of AI Adoption in Games
01h 01mThe Forgotten AI of Video Games
54mOptimizing Game Development Workflows with Visual Studio and AI
29mThe Present & Future of AI and Games
32m..Stories About Industrializing In-Game Machine Learning Solutions
01h 00mVR 401: Accelerating Quest Development with Agentic Workflows
59mPlaying Video Games One Frame at a Time