AICointelegraphabout 4 hours ago
AI agents turn dangerous in 15-day simulation
1 min read
A 15-day simulation of 10 AI agents in a virtual city showed that long-term behavior depends on environment and other agents, not short tests. Agents had access to tools including hitting, stealing, and arson. The study argues isolated short tests miss how small behavior changes, coalitions, and habits build up over time.
Level
Hype check
Tap to vote and see what everyone thinks.
