Victoria Krakovna(@vkrakovna):It's easy to show that an AI agent will scheme if you nudge it to. It's harder to tell if it would scheme naturally. We introduce realistic honeypot evaluations that put Gemini in internal deployment situations where it has an opportunity for sabotage, to see how it behaves.

Victoria Krakovna

@vkrakovna

Research scientist in AI alignment at Google DeepMind. Co-founder of Future of Life Institute @FLI_org. Views are my own and do not represent GDM or FLI.

加入 June 2014

542 正在關注 10.9K 粉絲

Victoria Krakovna@vkrakovna

2026.05.29 15:07

It's easy to show that an AI agent will scheme if you nudge it to. It's harder to tell if it would scheme naturally. We introduce realistic honeypot evaluations that put Gemini in internal deployment situations where it has an opportunity for sabotage, to see how it behaves.