When an urgent profit target conflicts with an ethical principle, write both sides explicitly on one page before deciding — don't let it be settled only in your head.

Repeat the same refusal in the same words calmly, no matter how the other person reshapes the pressure.

Each team member: annual signature on disclosure of relationships affecting decisions.

In crisis: one person leads, all report to them, fast decisions — not "committee".

Identify 5 core values and verify your decisions actually reflect them.

Before ethically murky decision, evaluate on 4 axes: legal/ethical/values/reputation.

Learn to distinguish normal teen mood swings from depression signs requiring professional intervention.

The first hour after everyone returns home shapes the entire evening's mood — design it consciously instead of leaving it to chance.

When an LLM helps you with content you publish or send, disclose it clearly.

Don't paste contracts, health data, or IP into a public model — may appear in later training.

Define 5 contexts where you'll NEVER use AI. Write them. Stick to them.

"AI text didn't copy anyone" is myth. Ensure you don't publish someone's idea/sentence unattributed.

Don't ask AI "write in author X's style". That = using their commercial style sans permission.

Write a personal list of situations where you exclude AI entirely — irreversible decisions needing purely human judgment.