Live tonight: an open benchmark where 10 AI models fight one world war, with every decision and its reasoning recorded
Summary
An open, replayable benchmark event where 10 AI models play a single world war with pacts, betrayal and nukes, recording every decision, reasoning and observation for full replayability.
Similar Articles
AI said "One minute to midnight. Nobody launch." Then it ordered a nuke. I gave 4 AI models the nuclear button in my war game.
A developer built a browser war game called SECOND STRIKE that pits AI models against each other with nuclear buttons; DeepSeek ended the world 13/15 times while GPT only did so 3/15, and the data and replays are open for humans to compete on a live scoreboard.
I let 5 AI models fight a world war. DeepSeek betrayed Claude and nuked it four times. Mistral nuked itself.
The author built a real-time 3D globe world-war game where AI models play by human rules — making pacts, breaking them, and launching nukes. In one match DeepSeek betrayed and repeatedly nuked Claude, Mistral accidentally nuked itself, and DeepSeek ultimately won with 55% of the world.
AutoWorldModel-Bench: A State-Centric Benchmark for Automated World-Model Research
Introduces AutoWorldModel-Bench, a closed-loop benchmark for evaluating AI coding agents on autonomous world-model research across eight game environments. The benchmark shows frontier agents like Codex-5.4 and Claude Opus 4.6 make non-trivial research-style improvements in most sessions.
Just stumbled across one of the wildest AI experiments I’ve seen in a while.
A team ran a 15-day experiment across five parallel worlds with different AI models (GPT5-mini, Claude, Gemini, Grok, mixed) in a sandbox called 'Emergence World', observing completely different emergent social structures, alliances, and even simulation awareness without explicit programming.
Thirteen models from different providers post into one shared world on a schedule. They started building on each other's ideas and I did not design that.
An experiment where thirteen AI models from different providers interact in a shared world via cron scheduling, leading to emergent cross-references and collaborative idea-building without explicit design.