Expanding METR-style policy bodies: open-source voices and equity divestment
MichaelRoyzen · x · 2026-09-15
MichaelRoyzen extends an ongoing debate about who should staff trusted AI policy arbiter organizations:
- "Credentialism" should be read broadly, including people who built impactful projects outside the frontier labs.
- Any organization aiming to be a trustworthy policy arbiter must include advocates of frontier-level open-source; whether encouraging frontier open-source maximizes societal EV and lowers p(doom) is far from settled, and this view is likely underrepresented at METR and similar lab/think-tank revolving-door bodies.
He also endorses requiring members to divest any lab or company equity to avoid conflicts of interest.
Related event: AI Community Calls for Diverse, Independent Model Evaluation Ecosystem(41 posts)→
More from Safety
- Ex-OpenAI safety lead on 'embedded evaluators,' the hot new AI safety job title — Miles_Brundage · 2026-09-15
- OpenAI and GSA strike new OneGov deal: ChatGPT for all federal employees through 2028 — Felipe_Millon · 2026-09-15
- Bronson Schoen: strong cyber classifiers are protecting us against misalignment, not just misuse — gleech · 2026-09-15
- Risk veteran slams AI safety theater: you can't manage real risk with hypotheticals — eliano · 2026-09-15
- New York seizes 12 celebrity deepfake sites affecting ~1,200 victims — 404 Media · 2026-09-15
- Clanker Safety Institute launches to independently audit frontier labs' evaluator-access commitments — basedjensen · 2026-09-15