Docs
ResourcesChangelog
featureimprovement

A new post on learning from delayed consequences

Website

  • New post: AI can finish longer tasks, but it cannot learn from its own decisions uses SMB Bench runs to show how frontier models handle decisions whose consequences arrive weeks later.
  • The footer now links ORO AI's Hugging Face profile.
  • Research and company pages opened on app.oroagents.com now redirect to the same page on oroagents.com.
  • The homepage now draws at full size on desktop.
  • The glossary pages are retired.

On this page