01
Observation & state
Perceiving a changing reality while representing uncertainty clearly.
What is true now?An independent research program
Medicine, redesigned
for machines that can act.
How should medicine change when intelligent machines can perceive patient states, select interventions, act in the world, and learn from their consequences?
Clinical action loop Working model
01State
02Observe
03Decide
04Act
05Respond
06Outcome
learning closes the loop
The premise
Most medical AI predicts.
The next generation will participate.
Medicine was built around humans as the sole agents of observation, judgment, and intervention. This research explores the data, interfaces, safeguards, and institutions required when intelligent systems move from describing care to taking part in it.
Read the position essayResearch framework
Not a technology map.
A complete action loop.
The central question is not only what a system knows, but what it can observe, how it acts, and how consequences remain visible, governable, and learnable.
02
Action & response
Closing the loop between a choice, its real-world effect, and the next decision.
What happens next?03
Outcome & accountability
Connecting meaningful outcomes, human oversight, safe interruption, and learning.
What remains accountable?Living bibliography
Reading evidence through one frame.
Papers, datasets, and systems organized around the same sequence: state → action → response → outcome.
Open the research libraryThinking in public
Publishing ideas before the field settles.
Paper analyses, synthesis essays, and working hypotheses in Korean and English.