I can't think of great interventions for ensuring third-party model access.

·LessWrong··

SummaryI'm increasingly convinced that model access parity is a big deal and we are not on track to achieve it. By model access parity, I mean a small gap between (i) the model access for lab employees and (ii) the model access for external safety researchers, third-party auditors, and other actors trying to make the future go well). See here for an introduction.The basic case is this: (1) Regardless of the strategic landscape, outsiders are well-suited to many crucial activities. (2) Outsiders ...

Read full article →

Related Articles

Since Linux 6.9, LUKS suspend stopped wiping disk-encryption keys from memory
IngoBlechschmid · Hacker News · 5h ago
Kimi K2.7 Code is generally available in GitHub Copilot
unliftedq · Hacker News · 15h ago
Department of Commerce has lifted export controls on Claude Fable 5 and Mythos 5
Pragmata · Hacker News · 1d ago
Claude Code is steganographically marking requests
kirushik · Hacker News · 2d ago
The fall of the theorem economy
varjag · Hacker News · 12h ago