What AI is (and what it is not)
Day 1 of 5 · AI Catch-Up Program (EN): Beginner Course
Tool, not magic
Learning goal
- See AI as a tool that needs clear input and QA.
- Spot risky use cases.
- Draft your AI use map.
Why it matters
- AI doesn’t know your intent; it reads text only.
- You own decisions and QA; AI does not accept liability.
- Good input → better output; iteration is normal.
Explanation
- AI suggests; you review.
- No company context unless you provide it.
- Guardrails: anonymize sensitive data; avoid critical decisions without human review.
Where it helps / where it doesn’t
- Helps: outlines, summaries, ideas, first drafts.
- Does not replace: legal/medical advice, password handling, confidential data processing.
Guided exercise (10–15 min) — AI use map
- List 3 tasks to accelerate with AI.
- For each: what you expect, what risk you see.
- Note if sensitive data appears; plan to anonymize.
Independent exercise (5–10 min) — Do-not-use list
List 5 tasks you won’t give to AI (passwords, critical decisions, sensitive clients) and what you’ll do instead (human review, manual step).
Self-check
- AI use map written.
- Do-not-use list created.
- You can explain why QA is mandatory.
Optional deepening
- OpenAI Usage Policies: https://openai.com/policies/usage-policies
- One Useful Thing (Ethan Mollick): https://www.oneusefulthing.org