Brainstorm Ideas with AI
Module 2 · Practice Using AI

Brainstorm Ideas with AI

Run a structured brainstorm with an AI assistant: go wide to generate lots of options, then narrow down with criteria you choose — and keep the final call your own.

Why Brainstorming Is a Great First Use

Brainstorming is the ideal place to start with AI because the risk is low and the payoff is high. An idea doesn’t have to be correct — it just has to be worth considering. If the AI suggests ten ideas and eight are mediocre, you’ve still gained two you might not have thought of, in about fifteen seconds.

AI is particularly good at the part of brainstorming humans find hardest: generating volume without getting stuck on the first idea.


Go Wide, Then Go Narrow

Good brainstorms have two phases, and AI helps with both.

Phase 1 — Diverge (go wide). Generate lots of options without judging them. Quantity beats quality here.

Phase 2 — Converge (go narrow). Apply criteria, compare, combine, and pick.

your pick1. Diverge — go wide2. Converge — go narrow20+ ideas · change the lens · no judgingcriteria → top 5 → pressure-test → 1

The most common mistake is collapsing both phases into one prompt (“Give me the best idea for…”). You get one safe, middle-of-the-road answer instead of a range to choose from.

Diverge Prompts

Give me 20 ideas for [goal]. Include some safe ideas, some
unusual ones, and at least 3 that seem a little ridiculous.
Just list them — no explanations yet.

Useful variations:

  • Change the lens: “Now give me 10 more ideas as if you were a [kindergarten teacher / startup founder / skeptical accountant].”
  • Add a constraint: “Now 10 ideas that cost nothing.” Constraints often produce the most creative results.
  • Flip it: “What are 10 ways to guarantee this fails?” Then reverse each one.

Converge Prompts

From the list above, pick the 5 ideas that best fit these criteria:
[budget under $500], [can launch within a month], [appeals to
[audience]]. Show them in a table with columns for the idea,
estimated effort (low/medium/high), and the biggest risk.

Then keep going:

  • “Combine ideas 3 and 7 into one stronger idea.”
  • “What would a skeptic say about my top pick?”
  • “What’s the smallest version of this we could test next week?”

You’re the Editor

AI supplies raw material. You supply judgment about what fits your team, your budget, your culture, and your customers — things the AI can’t see. Treat its list as a starting point, not a shortlist.

🧪 Try It: Brainstorm a Real Workplace Challenge

Pick a real challenge you’re facing. If nothing comes to mind, use this one: “Our team’s weekly meeting runs long and people seem disengaged.”

  1. Diverge. In a new chat, send:
    I lead a team of [number] people. Challenge: [your challenge].
    Give me 20 possible ideas to address it — mix practical, bold,
    and a few unconventional ones. List only, no explanations.
  2. Change the lens. Ask for 10 more ideas from a completely different perspective of your choosing.
  3. Converge. Give it 2–3 criteria that matter to you and ask for the top 5 in a table with effort and risk.
  4. Pressure-test. Ask: “What’s the strongest argument against your #1 pick?”
  5. Decide. Pick one idea to actually try. Write one sentence on why you chose it — that reasoning is the part only you could supply.
⚠️ Common Pitfall: Stopping at the First List

Common mistake: asking for ideas once, skimming the list, and concluding “nothing new here.”

The first list is usually the most predictable, because the AI leads with the most common answers. The interesting ideas tend to appear after you change the lens, add a constraint, or ask for the unconventional options. Push for at least two or three rounds before judging.

🧠 Knowledge Check

1. Why is 'Give me the best idea for improving team morale' a weak brainstorming prompt?

2. Which follow-up is most likely to surface genuinely new ideas after a first list?


Up Next

You’ve now used AI hands-on. Module 3 goes under the hood: how AI actually learns, why it gets things wrong, and how to use that knowledge to get better results.