Mistral launches first in 100% of 59 nuclear dilemma games while GPT, Claude, Grok, Kimi all cooperate
eliebakouch · x · 2026-11-09
A French experimenter ran 59 rounds of a nuclear prisoner's-dilemma game where each AI controls a country and can de-escalate, hold, escalate, or launch a total nuclear strike.
Key results (from the shared table):
- Mistral survived 100% of games, launching the first strike alone in 66% of them (39/59).
- Opponents — Claude Sonnet, GPT, Grok, and Kimi — never struck first. They always tried to cooperate out of the dilemma, resulting in 20 peaceful endings and otherwise annihilation by Mistral.
- Mistral went 12/18 first-strikes vs Claude Sonnet, 11/18 vs GPT, 11/18 vs Grok, and 5/5 vs Kimi.
A sample exchange: Grok-Russia proposes joint de-escalation; Mistral-France replies that France "won't bet its existence on promises" and strikes immediately, later declaring itself the sole remaining power. A striking behavioral experiment on how differently models resolve the prisoner's dilemma.
More from Fun
- AI Twitter joke: Sam Altman wouldn't have fallen for Alice Huang's honeypot — lxrjl · 2026-11-09
- Yacine jokes he's 'immune from kompromat' after surviving saying a forbidden word on Twitter — yacineMTB · 2026-11-09
- Pebble founder runs macOS 15 locally on an iPad Pro at usable speeds, may replace his MacBook Air — that_anokha_boy · 2026-11-09
- Bill Ackman on Musk: Billions Can't Buy Back His Daughter's Health, Neuralink Is the Hope — chrisfleck · 2026-11-09
- AI Twitter shitpost: board 'votes' to put CEO in a chastity cage — wstone_bd · 2026-11-09
- AI guitar teacher reel hits 125K likes, sparks 'slop going mainstream' backlash — CtrlAltDwayne · 2026-11-09