Ideas and working notes
Questions I am still working through.
These are not finished essays. They are the questions I keep returning to, held open on purpose so the thinking stays inspectable while it changes.
Current questions
3 questions, none of them settled.
-
Memory and identity Open question
What should an AI remember — and how should that memory earn trust?
A system that remembers everything is a liability; one that remembers nothing is useless. The open part is what earns a place in memory, and what has to be shown before that memory is trusted.
Open the question -
Authority and agents Open question
How do we delegate work without making accountability disappear?
Delegation to software moves the work but not the responsibility. The open part is where responsibility actually lands once a system acts on someone's behalf.
Open the question -
Evaluation and approval Open question
Where should automation stop and human judgment begin?
Some decisions should stay with a person even when a machine could make them faster. The open part is how to tell which ones, before rather than after.
Open the question
How to read this
The states are meant literally.
| State | What it means | In this list |
|---|---|---|
| Open question | A question I am actively holding open. There is no position yet, and saying otherwise would be dishonest. | 3 |
| Working position | A view I currently hold and act on, stated plainly, still open to being wrong. | 0 |
| Tested | A position that has survived contact with real work, with the evidence shown. | 0 |
| Revised | A position I changed, kept visible along with what changed it. | 0 |
| Retired | No longer something I think is useful. Kept for the record rather than deleted. | 0 |
Keep going