01Aug 06, 2026
AI and research
When AI Polishes the Paper, What Remains Our Work?
AI can change how a paper is expressed. It cannot take responsibility for its evidence, interpretation, or conclusion.
8 min readRead note
ROBIN / PERSONAL OPERATING SYSTEM
I think by building,
and I build what I think.
I'm Robin — a biomedical engineer working across silicon, materials, AI agents, and wet-lab practice. I use this site to turn research, reflection, and action into one system that can be tested and revised.
robin://field-notesv2.0.1
NODE 01CALIBRATABLEsilicon ↔ wet-lab ↔ mind
01 observe()
02 build()
03 review()
04 reuse()
02 build()
03 review()
04 reuse()
LEGO
ROBIN
ROBIN
KNOW DO REVIEW REUSE KNOW
Research map / 01
Six fields.
One way of working.
Each theme is a module that can grow independently and still connect to the others. Choose one to return to its working notes.
Working notes / 02
Break the problem.
Let evidence speak.
Working versions of ideas about research, AI agents, personal systems, and life. The edges stay visible here; answers can wait.
02Aug 06, 2026
Scientific reasoning
The Valley Before Fluency
Research growth is not a steady climb. It is the gradual construction of feedback loops inside uncertainty.
6 min readRead note
03Aug 06, 2026
AI and research
The Old Grammar of Research Is Changing
AI agents may change not only the speed of research, but the order and location of its reasoning.
9 min readRead note
04Aug 06, 2026
AI and research
From Plugins to Closed-Loop Nodes
How an agent system can move from a plugin stack to reusable nodes that connect digital action with physical feedback.
10 min readRead note
05Jul 18, 2026
Clinical-to-chip
The chip is not the system
What changes when a silicon device is designed alongside the living context it is meant to measure.
8 min readRead note
06Jul 04, 2026
Wet-lab systems
A prototype is a question with edges
A field note on building wet-lab systems that make uncertainty visible early, before it becomes expensive.
6 min readRead note
Wang Yangming portrait · Cai Shixin · Shanghai Museum · public-domain image
THOUGHT PROTOCOL / 03
Knowing and doing
are one loop.
I read Wang Yangming’s idea of knowing and doing as a working method, not a wall quote: a model should change what I do, and an action should produce new understanding.
知
行
行
KNOWING_DOING_PROTOCOL
KNOWING / DOING
Knowledge is not the endpoint. It has to enter an experiment, a system, or a life before we know whether it holds.
01KNOW / UNDERSTAND
Observe, read, and model the shape of the problem.
↓
02DO / BUILD
Turn understanding into an experiment, a piece of code, a material, or a real decision.
↓
03REVIEW / MAKE IT YOURS
Record the result, expose the assumptions, and carry the pattern into the next action.
AVATAR / 01
PERSONAL NODE
PERSONAL NODE
designed to be rebuiltAbout Robin / personal node
Stay curious.
Stay rebuildable.
I work across silicon, materials, wet-lab practice, and AI agents. This is not a résumé. It is a rebuildable personal node: turning ideas into experiments, experiments into understanding, and understanding back into life.
Open a conversation