Chatbots say things, agents do things: a clinician's case for medical AI agents
jonc101x · x · 2026-10-28
Kicking off a Stanford HAI-related thread, clinician jonc101x offers a crisp dichotomy: "Chatbots say things. Agents do things."
He argues AI chatbots are already remarkable at answering medical test questions, but doctors don't spend their days answering quizzes. Even purely cognitive computer-based work — reviewing charts, pulling echocardiograms from two years ago, reconciling morning labs, deciding what to do, placing orders, and documenting encounters — is the real workload that end-to-end agents would need to handle.
The accompanying graphic drives the point home: models are impressive on medical exams, but whether they can actually "do the work" remains an open question. The thread tees up Stanford HAI's healthcare AI programming.
More from AGI Musings
- OpenAI researcher: frontier AI ideas can't be patched for safety after the fact — willdepue · 2026-10-28
- Big Short's Eisman: the whole market is now 'one trade' on AI — and it 'better succeed' — SumitGup · 2026-10-28
- Most LLM Users Ask Bad Questions, and School Is to Blame — TheMoonMidas · 2026-10-28
- Blogger pushes back on 'tool AIs want to be agent AIs': don't let agents swallow craft — akbirthko · 2026-10-28
- Blogger predicts a physics-accurate world model would unlock mind-boggling synthetic data — imjustnewatai · 2026-10-28
- Biomedical AI veteran: solving biology is limited by data, not compute — _smileyball · 2026-10-28