Claude Mythos Hits 50% Success on 16-Hour Tasks in METR Time Horizons

刚刚，Claude Mythos打爆AI评测天花板！超指数狂飙，2027奇点加速

Claude Mythos Preview reached a 50% success rate on METR Time Horizons tasks that take humans 16 hours, while only 5 of 228 tasks exceeded the 16-hour range, so the article says METR lacks enough samples to quantify longer-horizon performance.

Why it matters: HKR-H/K/R all pass: the 16-hour task result is a strong hook, and the METR sample caveat adds substance. Capped at 82 because only 5 tasks exceed 16 hours, so the 2027 extrapolation is not same-day P1 material.