There is a narrow window—hours to a few days—when a vague note is still fully understandable to its author. After that, ambiguity hardens into mystery. Clarification is the practice of asking the right small question inside that window, and it is one of the few things an AI can do for your memory that you predictably will not do yourself.
The gap between capturing and understanding
Capture optimizes for speed: pronouns, fragments, a photo with no caption. Understanding, months later, needs names, dates, and intent. No amount of writing discipline closes this gap reliably, because the whole point of quick capture is not stopping to think.
Why timing beats thoroughness
A clarifying question asked the same evening costs seconds: ‘him’ is obviously Marek, ‘Friday’ is obviously the 14th. The identical question asked in six months is unanswerable. This is why periodic archive cleanups fail while small timely questions work—the information still exists only near the moment of capture.
What a good clarifying question looks like
Clarification earns its interruption by being small, specific, and answerable from memory in one line.
- It targets one ambiguity, not the whole note.
- It offers a guess when it has one: ‘Is this about the Vantaa project?’
- It can be dismissed without penalty—silence is an acceptable answer.
- It never asks what the system could infer from existing notes.
Answers become part of the record
A clarification is not a chat that evaporates. The answer attaches to the original note, dated, and gets indexed for search—so ‘ask him on Friday’ is findable forever after as the note about Marek and the estimate. The original stays untouched; the meaning is preserved alongside it.
You cannot write perfect notes at speed, and you should not try. Capture fast, then answer the one small question the system asks while the answer is still in your head.