corneliu-iancu Ask better questions
Transform shallow questions into powerful ones by running them through 7 Socratic lenses.
Quick start
Take the user's question or problem and apply the 7 lenses:
- Challenge — What's wrong with my reasoning?
- Unconstrained — What if I had infinite resources?
- Meta — What didn't I ask?
- Uncertainty — Where am I wrong?
- Inverse — What's the opposite?
- Reframe — How would a [domain expert] see this?
- Assumptions — What am I optimizing for implicitly?
Example:
User's question: "How do I build better retention?"
After 7 lenses:
- "What flaws are in my retention model?"
- "What if I had unlimited budget to keep users?"
- "What signals am I ignoring that predict churn?"
- "Where's my retention strategy failing that I haven't admitted?"
- "How would I maximize churn instead?"
- "How would a behavioral psychologist reframe retention?"
- "Am I optimizing for retention or engagement?"
Pick 1-2 that resonate. Ask those instead. Get better answers.
How to use
Approach 1: AI-guided (interactive) — default
- User gives their problem: "I'm trying to..."
- Apply all 7 lenses and present reformulations
- User picks which ones matter
- Explore the chosen lenses deeper, then synthesize
Approach 2: Self-guided (manual)
- Present the 7 lenses as a framework
- User applies each lens themselves
- Help them refine the 1-3 that feel most important
Approach 3: Deep dive (iterative)
- Start with one lens (e.g., Inverse)
- Explore that lens thoroughly
- Move to the next lens
- Synthesize across lenses for comprehensive insight
The 7 lenses — quick reference
| Lens | Reframes | Best for |
|---|---|---|
| Challenge | Weaknesses in your approach | Catching blind spots |
| Unconstrained | What's truly wanted, not constrained | Vision-setting, discovering what matters |
| Meta | What you're missing | Completeness checks, edge cases |
| Uncertainty | Admitting what you don't know | Risk awareness, intellectual humility |
| Inverse | Opposite direction | Discovering root causes, hidden assumptions |
| Reframe | Fresh perspective from another field | Novel solutions, pattern recognition |
| Assumptions | What you're actually optimizing | Alignment checks, goal clarification |
For detailed guidance, prompts, and examples for each lens, read references/lenses.md.
When this skill helps most
- Before asking AI anything important — refine the prompt first
- When stuck on a problem — a different lens often unsticks you
- Decision-making — expose hidden trade-offs
- Project planning — catch unspoken goals
- Learning — go deeper than surface answers
- Creativity — challenge assumptions to unlock new ideas
Common patterns
Surface question into 3 refined questions
Surface: "Should I hire more engineers?"
Lens 1 (Challenge): "What are the real constraints on our output — is it headcount?"
Lens 4 (Uncertainty): "Where am I wrong about engineering capacity?"
Lens 7 (Assumptions): "Am I optimizing for velocity or code quality?"
Ask these 3 instead of the surface one.
Stuck decision
Stuck on: "Do I pursue this project or not?"
Lens 6 (Reframe): "How would a CFO see this? A designer? A customer?"
Lens 5 (Inverse): "What if I actively avoided this? What's the fear?"
Lens 3 (Meta): "What am I not asking that would help?"
Run through these lenses first; the decision often clarifies.
Brainstorming breakthrough
Brainstorming: "How do we improve onboarding?"
Lens 5 (Inverse): "How would we maximize onboarding failure?"
Lens 2 (Unconstrained): "What if onboarding took 10x longer but was perfect?"
Lens 6 (Reframe): "How would a game designer think about this?"
These unconventional angles unlock creative solutions.
Lens combinations
Trio for clarity:
- Lens 7 (Assumptions) + Lens 3 (Meta) + Lens 1 (Challenge)
Trio for creativity:
- Lens 5 (Inverse) + Lens 6 (Reframe) + Lens 2 (Unconstrained)
Trio for risk:
- Lens 4 (Uncertainty) + Lens 5 (Inverse) + Lens 3 (Meta)
Tips
- Start with 1 lens, not all 7 — too many at once overwhelms
- Some lenses matter more than others for any given problem — lean into the 2-3 that feel obvious
- Iterate — first refinement begets second refinement
- Trust your gut — if a lens doesn't resonate, skip it