Nagi Experiments
Nagi experiments are practices for trying things on a small scale near everyday life, then stopping and repairing them when problems appear.
What must be protected
- Make participation voluntary.
- Keep AI-supported methods alongside human-only methods.
- Limit records to what is necessary.
- Explain how records will be used and how long they will remain before beginning.
- Do not disadvantage people who leave midway.
- Receive opposition and anxiety, and review the practice.
- Provide a safe route for people affected to seek consultation and raise objections.
What can be tried
Try, in small settings, practices such as handover in collaborative work, rest time in meetings, inheritance of local culture, revision of uneven roles, and translation or summarization of materials.
AI can help with translation, summarization, accessibility, and organizing issues. It does not score human value or emotion.
Observation and repair
Do not decide success in advance. Observe what effects arose for whom and what remains unseen.
When harm or error occurs, do not end with concealment or condemnation alone. Separate fact from inference, acknowledge impact, make necessary corrections and explanations, and review the systems that made recurrence possible.
Do not compel the continuation of a relationship or the resumption of participation.
How an experiment ends
An experiment does not exist merely to continue. Before it begins, decide what would make it stop, what should not be retained, and how it will be reviewed.
Nagi experiments do not exist to prove that Nagi is right.
They exist to find settings that are freer and easier to breathe in.
→ Transition and Crisis Design
→ Memory, Consent, and the Right to Forget
→ Observe, Loosen, Connect
→ Repair and Cultural Co-Growth
→ Japanese reference edition