Brainstorming Ideas

Give a seed idea (a product, story premise, campaign, feature, etc.) and specify which mode you need:

1. Expand — turn one idea into 5-10 divergent directions
2. Associate — generate unexpected metaphors/connections to enrich an idea
3. Critique — run devil's advocate analysis to find weaknesses

Example request: "Expand this idea: a coffee shop that only serves coffee at the temperature outside."

Output: 7 directions ranging from literal (weather-synced café) to absurd (a coffee shop that argues with you about the weather) to business-model twists (subscription-based temperature-coffee).

Progress:

• Clarify the seed idea and the goal (more ideas? validation? both?)
• Choose mode(s): Expand / Associate / Critique
• Generate raw output using the relevant technique below
• Filter/rank results by novelty and usefulness
• Present with brief rationale for each

Mode 1: Expansion (1 idea → 5-10 directions)

Use divergence axes to force variety — don't just generate 10 similar riffs. Pull from different axes each time:

• Genre swap: sci-fi, horror, comedy, noir, fairy tale version
• Scale shift: micro (personal/intimate) vs macro (global/systemic)
• Audience shift: for kids, for experts, for skeptics
• Inversion: what if the opposite were true?
• Medium shift: as a product, a story, a service, a ritual, a game
• Constraint injection: zero budget, one day only, no technology
• Combination: merge the seed with an unrelated domain

Label each output with the axis used, so the thinking is transparent.

Mode 2: Association / Metaphor Generation

Technique: force connections between the idea and unrelated domains (nature, mythology, machinery, human relationships, games, weather, cooking).

Process:

1. Strip the idea to its core function or feeling ("what does this do for someone?")
2. Ask "what else in the world does this same thing?"
3. Generate 5+ candidate metaphors, keep the 2-3 most generative (ones that suggest new features or angles, not just decoration)

Mode 3: Devil's Advocate Critique

Attack the idea from multiple angles, not just one generic "what could go wrong":

• Feasibility: what resource/technical/time constraint breaks this?
• Audience: who would actively dislike or ignore this, and why?
• Novelty: is this secretly already done? By whom?
• Incentives: who loses if this succeeds, and would they sabotage it?
• Edge cases: what's the worst-case misuse or failure mode?
• Steelman rebuttal: for each critique, suggest one way to address it

Always end critique with a short list of "ideas that survive scrutiny" vs. "ideas that need rework."

Example 1 (Expansion): Input: "A library where silence is forbidden." Output:

1. Genre swap – horror: a library where silence summons something.
2. Audience shift – kids: a library that's basically a playground with books.
3. Inversion: a library where only whispering is allowed, amplified by design.
4. Medium shift: not a building — a podcast series recorded inside real libraries.
5. Scale shift – macro: a city-wide network of "loud libraries" replacing noise ordinances downtown.
6. Combination with theater: scheduled "read-aloud hours" turn it into performance space.
7. Constraint – zero budget: pop-up version in a park, using megaphones and donated books.

Example 2 (Devil's Advocate): Input: "A subscription box that sends a mystery ingredient each week for home cooks." Output:

• Feasibility: perishable ingredients + shipping delays = spoilage risk. → Mitigate with shelf-stable rotation.
• Audience: time-poor cooks may resent the obligation to use it before it spoils. → Add a "skip week" option.
• Novelty: Blue Apron/HelloFresh already do curated-ingredient models. → Differentiate via single-ingredient creative constraint, not full meal kits.
• Incentives: grocery delivery apps compete directly and have more capital. → Position as entertainment/game, not grocery replacement.
• Survives scrutiny: the "creative constraint" framing (game-like) is the differentiator worth keeping.