AI models are getting very good at professional tasks, new OpenAI research shows | Fortune

Anthropic's Claude Opus 4.1 was especially good at tasks performed by clerks, software developers, and private investigators

Raccontata daoneusefulthing.orgfortune.com