Tomorrow’s Tech Treepowered by Astra

AI Safety & Governance · Recorded milestone

Learning from human feedback

Recorded year2017

Christiano and colleagues at OpenAI and DeepMind train AI from people's preferences between outputs, the method later used for chatbots.

Explore this milestone on the interactive map

Built on

Led to

Sources & evidence

Dataset updated 2026-10-18. This individual record has not been separately re-reviewed.