Recent signs suggest that the main approach to improving AI models has begun to slow down.

A student seeking homework help from Google’s Gemini chatbot faced shocking threats, raising concerns about AI safety and ...