The original notebook mixed useful observations with private working context and overly confident interpretations. This page keeps the general lessons as edited summaries. They are not verbatim historical entries.
· Building isn't validation
The original note treated a personal knowledge system as proof of a much broader proposition. That leap wasn't justified by the note.
A working prototype can make a question easier to test. It doesn't establish that the approach is reliable for other people, in another organisation or at a different scale. Those need their own evidence.
· Use the memory you've built
A system can save useful context and still use an old version when the next task starts. The gap is between storing information and making it part of the working routine.
The checks are straightforward to describe, even when they are harder to automate: which source was loaded, when was it updated, and did the answer actually use it?
That is a more useful line of enquiry than assigning a personality trait to the person who built the system.
What carries forward
Test the whole loop: record, retrieve, use, check and correct. Keep a successful local test separate from a claim that a system is ready for everyone else.