TL;DR

Use AI for repetitive tasks, brainstorming, drafting, and data analysis. Don't use AI for final decisions on critical matters, personal/emotional tasks, or anything requiring up-to-date facts. Always verify important outputs.

When AI is great

Repetitive, time-consuming tasks:

  • Summarizing long documents
  • Drafting routine emails
  • Organizing data
  • Creating variations (headlines, slogans, designs)

Brainstorming and ideation:

  • Generating topic ideas
  • Outlining arguments
  • Suggesting approaches
  • Exploring possibilities

Learning and explanation:

  • Understanding complex topics
  • Breaking down jargon
  • Step-by-step tutorials
  • Practice problems

Quick answers to common questions:

  • "How do I convert units?"
  • "What's a good synonym for X?"
  • "How does Y work?"

When to be cautious

High-stakes decisions:

  • Medical advice (see a doctor)
  • Legal counsel (hire a lawyer)
  • Financial planning (consult an advisor)
  • AI can inform, but experts should decide

Anything requiring recent information:

  • News, current events, stock prices
  • Policy changes, new laws
  • AI knowledge has a cutoff date

Emotional or personal tasks:

  • Apologies, condolences, love letters
  • These need genuine feeling, not generated text

Creative work requiring originality:

  • AI remixes patterns; true innovation needs human vision
  • Use AI for inspiration, not as the final product

When NOT to use AI

Never for:

  • Sharing passwords, SSNs, bank details
  • Critical medical diagnoses
  • Legal documents without expert review
  • Replacing human connection (therapy, counseling)
  • Final decisions on important life choices

Avoid for:

  • Sensitive company information (unless approved tools)
  • Academic work where AI is banned
  • Tasks where errors have serious consequences

Decision framework

Ask yourself:

1. Is this high-stakes?

  • Yes → Don't rely solely on AI
  • No → AI might help

2. Does it require up-to-date information?

  • Yes → Verify with current sources
  • No → AI is fine

3. Is it personal or emotional?

  • Yes → Do it yourself
  • No → AI can assist

4. Can I afford to be wrong?

  • No → Verify AI output carefully
  • Yes → AI is low-risk

How to use AI responsibly

1. Start with AI, finish with you

  • AI drafts, you refine
  • AI suggests, you decide
  • AI assists, you lead

2. Verify important information

  • Cross-check facts with reliable sources
  • Don't trust AI blindly

3. Add your voice and judgment

  • Personalize AI outputs
  • Apply context and common sense

4. Use AI as a tool, not a crutch

  • Don't let AI replace skills you need to maintain
  • Use it to augment, not replace, your abilities

Examples

Good use:

  • "Draft an email declining a meeting" → AI drafts, you personalize

Bad use:

  • "Write my college essay" → Academic integrity violation

Good use:

  • "Explain how photosynthesis works" → Learn faster

Bad use:

  • "Should I have surgery for my condition?" → See a doctor

Good use:

  • "Give me 10 blog post ideas about gardening" → Ideation

Bad use:

  • "Tell me today's stock market performance" → AI may not know

The bottom line

AI is a powerful assistant for routine tasks, brainstorming, and learning—but it's not a replacement for human expertise, judgment, or connection. Use it thoughtfully, verify its outputs, and never let it make critical decisions for you.

When in doubt, ask: "Would I trust a very smart but occasionally forgetful intern with this task?" If not, don't trust AI alone.

What's next

  • Prompting 101
  • Evaluating AI Answers
  • AI Safety Basics