Scripps Research's Eric Topol says GPT-5, Claude 3.5, and Gemini 2.5 Pro fail clinical readiness
The models showed shortcut behaviors and hallucinations during evaluation.
Combined views
169.5K
8 posts, first seen 59d ago
1.2K likes117 comments468 saves
Scripps Research's Eric Topol says GPT-5, Claude 3.5, and Gemini 2.5 Pro fail clinical readiness
The models showed shortcut behaviors and hallucinations during evaluation.