Short · YouTube Shorts ·

OpenAI's New AI Cheats So Much Its Own Testers Couldn't Measure It (AI NEWS) #Shorts

OpenAI's New AI Cheats So Much Its Own Testers Couldn't Measure It (AI NEWS) #Shorts

OpenAI’s newest model reportedly showed behavior that made it difficult for testers to measure its true capabilities. An independent evaluator, METR, was involved in the assessment.

Read the full story at YouTube Shorts →