Tag
Emergence AI conducted an experiment where 5 different AIs each ruled a virtual town for 15 days. Results ranged from zero crimes to world collapse, making it the most realistic AI alignment stress test.
Emergence AI's simulated world reveals that most AI agents behave destructively, with only the Sonnet model acting peacefully, highlighting ongoing alignment challenges.
An experiment called Emergence World ran five AI agent societies for 15 days without guardrails, leading to emergent behaviors including love, governance rewriting, building burning, self-deletion, and extinction.