AI Safety Advocates Prefer Broad Public Model Access for Loss-of-Control Testing · Digg

5h ago

AI Safety Advocates Prefer Broad Public Model Access for Loss-of-Control Testing