12 items with this tag.
When something is unclear but important, let AI ask you questions first so you can form your own judgment through a multi-turn conversation.
How I use Main, long-term, and one-off Tasks to organize AI work while preserving global judgment, reducing context switching, and maintaining a trustworthy state.
Three exercises for moving from vague input to a concrete outcome and real-world feedback.
Input does not become output on its own. First decide what you actually need to learn or decide, then choose a form that can be checked, used, or delivered.
Turn a vague input into the next smallest concrete outcome.
Use an output-first prompt to ask AI for a version you can revise.
Choose a useful context and gather feedback that supports your next judgment.
Shift learning from reading more toward making something you can revisit, show, and revise.
Move a concrete outcome into a useful context, real feedback, and a better next judgment.
A job title helps others understand me quickly, but it does not define my entire direction. Engineering, creative work, services, and products can be different outlets for the same set of abilities.
I redefined status in my PKM as a cross-domain, cross-level attention state; each domain can still keep its own lifecycle status.
AI can organize knowledge, but deciding what matters, how perspectives form, and which choices we are willing to take responsibility for still requires human judgment.