Hugging Face Launches Open Alignment Initiative, Backs Anthropic's Embedded Evaluators
Key Info
Hugging Face co-founder Clement Delangue announced two steps toward safer AI: launching the Open Alignment Initiative led by Thom Wolf and Hugging Face, and requesting inclusion in Anthropic's "embedded evaluators" program, which Anthropic CEO Dario Amodei just committed to.
Highlights
- The Open Alignment Initiative aims to address AI alignment openly, rather than behind the closed doors of a few frontier labs.
- Anthropic is unilaterally committing to provide third-party evaluators with permanent, employee-level access to verify safety measures, report incidents, and assess alignment during training.
- The move follows Dario Amodei's essay "We Must Pace the Frontier," which argues the AI industry should slow down and proposes a three-part plan.