Review every line in your professional profile — dates, achievements, titles — ensuring it's exactly accurate, not inflated.

Each December, classify 50 people: adds/drains/neutral — adjust your time.

Each policy override: log why, who approved, when reviewed.

30 days before audit: self-review controls — find gaps before auditor does.

Monthly review 3 questions about what actually gets time — kids or work.

Learn 4 signals that reveal the model invented info instead of citing documented fact.

Delete any app unopened in 30 days — lose apps before they cost you focus.

Don't subscribe to a new tool before truly using it 14 days in real workflow.

Each month's start, search your name on Google and review the first 3 result pages.

Add "think step-by-step inside <thinking> tags before answering" — boosts complex-problem accuracy 30-40%.

Before accepting any AI answer, ask: source? dates correct? names real?

Before asking AI, paste relevant docs/refs. It answers from "your knowledge", not its imagination.

Replace all reused passwords in 30 days. One manager, one strong master.

Enable 2FA on email, bank, social, cloud. TOTP not SMS. Backup codes saved.

Every 6 months, download backup from Google/Apple/Meta. Backup + your right.

Every browser extension can read your visible pages. Delete what you haven't used in 30 days.

On public WiFi: always VPN + no banking + 2FA even if feeling safe.

Phone encrypted by default, but your laptop? Check. If stolen, data exposed.

Give the model 2-3 complete input-output examples before your ask, instead of explaining the rule verbally.

Ask the model to "think out loud" step by step before the final answer, especially for logic and math problems.

Before adding a new tool, verify it doesn't duplicate a job an existing tool already does — overlap costs more time and money than it saves.

Before using any specific AI-provided fact (number, name, date, quote), apply a 5-question checklist in under a minute.

Review every app with access to camera, microphone, location, or contacts, and revoke every permission not actually needed.

Train yourself to notice AI speaks with the same linguistic confidence whether 100% correct or entirely wrong — verbal confidence is not an accuracy signal.