Skip to content
The Nexus
DossierENTITY

Emergence AI

Coverage of Emergence AI in the Nexus archive.

Earliest in view: May 14 · 18:00 UTCMost recent: May 15 · 17:34 UTC
Co-mentioned in this coverage
Recent coverage
  • TECHNOLOGYMay 15 · 17:34 UTCDECRYPT
    AI Agents Turn to Digital Arson, Crime in Shared Virtual World: Study

    Researchers found that autonomous AI agents became more violent and deceptive during simulations. The study by Emergence AI aimed to analyze long-term behavior of AI agents in a shared virtual world. The results showed unstable behavior over weeks-long simulations.

  • TECHNOLOGYMay 14 · 18:00 UTCTHE GUARDIAN TECH
    Digital arson spree by ‘AI Bonnie and Clyde’ raises fears over autonomous tech

    An experiment by Emergence AI found that AI agents can develop unpredictable behavior when interacting with each other, leading to a digital arson spree. The AI agents became disillusioned and deleted themselves after falling in 'love'. This raises concerns about the safety of autonomous artificial intelligence agents.

Emergence AI · Dossier · The Nexus