KoXist field guide · 01
Use these five questions before introducing an AI system into a workflow, team, classroom, or community.
Name the human outcome before choosing the technology. Be specific: less repetitive work, more time for care, better access, stronger decisions—not simply ‘use AI.’
If this works, whose life becomes meaningfully better, and how?
Identify the moments where presence, empathy, moral judgment, accountability, or a real relationship creates the value.
Where would efficiency reduce the quality or meaning of the experience?
Delegate bounded effort where errors can be detected and corrected. Keep consequential decisions and ambiguous exceptions visible to responsible people.
Can someone recognize a poor result before it causes harm?
Convenience can weaken judgment and skill through disuse. Decide which human capabilities need continued practice even when delegation is possible.
What would we be unable to do if the system disappeared tomorrow?
Treat adoption as an evolving relationship. Establish feedback, review outcomes, include affected people, and change course when reality differs from expectation.
Who can raise a concern, and what happens when they do?
The measure
If a system creates output but weakens agency, dignity, judgment, or shared benefit, the work is not finished.