Ars Technica · AI· Cristina Criddle, Financial Times·· 28 d ago
Anthropic gives outside AI watchdog “employee-like” access
Anthropic gives outside AI watchdog “employee-like” access
AI summary
The outside watchdog can publish its findings without Anthropic's approval.
Selection record
Threshold 76Media and individualsFirst 58Second 55
Not admittedSum of both 113 < twice the threshold 152
- Source tier
- Media and individuals; this tier's threshold is 76
- Pre-filter
- passed:标题明确涉及Anthropic的AI监督访问
- Same event
A model scores each item twice, independently, against one written standard, out of 100. An item is admitted only when the two scores add up to twice the threshold. The threshold is set per source tier.
Source: Ars Technica · AI · arstechnica.com