Ethan Mollick's D&D Encounter Test: watching AI models fight mind flayers
eldonredwards · x · 2026-09-08
Ethan Mollick proposes the "Encounter Test," a nerdy AI benchmark: have a model simulate a D&D fight between two creatures and see how long it takes to break the rules. GPT-4o did best; Gemini was cute with similar outcomes.
Two years later he re-ran it with a stronger model ("GPT-6"): its mind flayer even got a "101-bone skeletal rig." He built Underdark Duel, a reproducible web page logging every roll, with adjustable encounters and downloadable records. Under 5E 2014 rules he found no major errors, though the mind flayer's tactics could be more creative.
Related event: Ethan Mollick Proposes 'Encounter Test' Benchmark Using D&D Battles(2 posts)→
More from Fun
- Beeple amplifies viral take: artists don't need studios in the AI era — soleio · 2026-09-08
- Microsoft shipped Friday; by Tuesday a stranger had reverse-engineered it with official tokens — unixterminal · 2026-09-08
- Planning an office layout with GPT 6 Astra and Blender — Rasmic · 2026-09-08
- Clicks Communicator praised as the best-looking keyboard phone, with a stock-price joke — cocktailpeanut · 2026-09-08
- First viewer finishes hkproj's ~19-hour AI course video — hkproj · 2026-09-08
- Dev texts random SF track ideas to Astra agent, which ships each one — blakesamic · 2026-09-08