Will's Blog: Models like o3 and Gemini 2.5 Pro feel like "Jagged AGI": unreliable, even at some mundane tasks, but still offering superhuman capabilities in many areas (Ethan Mollick/One Useful Thing)

21 April 2025

Models like o3 and Gemini 2.5 Pro feel like "Jagged AGI": unreliable, even at some mundane tasks, but still offering superhuman capabilities in many areas (Ethan Mollick/One Useful Thing)

Ethan Mollick / One Useful Thing:
Models like o3 and Gemini 2.5 Pro feel like “Jagged AGI”: unreliable, even at some mundane tasks, but still offering superhuman capabilities in many areas — New models and new thresholds — Amid today's AI boom, it's disconcerting that we still don't know how to measure how smart …

Posted from: this blog via Microsoft Power Automate.