OpenAI and DeepMind Might Be Aligning on AI Safety
Sam Altman called Demis Hassabis's idea thoughtful, which stands out since these two rarely speak publicly.
@sama4 days ago
TL;DR:
- OpenAI and DeepMind could be coordinating on safety or release standards for advanced models.
- This matters because what these two do often sets the tone for the rest of the industry.
- No proposal details are out yet, so it's hard to judge how far this might go.
Headline
Sam Altman backed a proposal from Demis Hassabis.
Summary
The OpenAI boss called the idea from the DeepMind CEO thoughtful. We don't know what the proposal said, but it's unusual to see these two leaders talk openly.
Analysis
This suggests they might agree on safety or testing standards. These two engaging publicly usually means something's up that affects both. Could mean shared rules for releasing advanced models. No details though, so hard to tell how far it goes.
Impact Assessment
Significance: Medium Categories: AI Policy, AI Safety, Industry Trend