Refactor high-complexity React components in Dify frontend. Use when `pnpm analyze-component...
npx skills add defi-naly/skillbank --skill "think-again"
Install specific skill from multi-skill repository
# Description
Adam Grant's framework for intellectual humility, rethinking, and updating beliefs.
# SKILL.md
name: think-again
description: "Adam Grant's framework for intellectual humility, rethinking, and updating beliefs."
dimensions:
domain: [decision-making, psychology, communication, leadership]
phase: [reflection, debate, culture-building, belief-updating]
problem_type: [changing-minds, intellectual-humility, productive-disagreement, unlearning]
contexts:
- situation: "holding a strong belief"
use_when: "need to audit whether belief is still justified by evidence"
- situation: "trying to persuade someone"
use_when: "need techniques that actually work (questions > arguments)"
- situation: "team has groupthink"
use_when: "building culture of productive disagreement"
- situation: "received information that contradicts your view"
use_when: "deciding whether and how to update beliefs"
- situation: "debate is becoming adversarial"
use_when: "shifting from prosecutor mode to scientist mode"
combines_with:
- thinking-fast-and-slow # diagnose biases before updating
- how-to-win-friends # influence techniques complement persuasion
- hard-thing-about-hard-things # changing minds during crisis
- never-split-the-difference # negotiation with belief-change
contrast_with:
- skill: thinking-fast-and-slow
distinction: "Think Again is about WILLINGNESS to change; TF&S is about ABILITY to see clearly"
- skill: how-to-win-friends
distinction: "Carnegie is about likability and influence; Grant is about truth-seeking and intellectual honesty"
Think Again: The Power of Knowing What You Don't Know
Core Philosophy
We spend too much time thinking and not enough time rethinking. The ability to unlearn and relearn is more valuable than the ability to learn.
The paradox: The smarter you are, the more vulnerable you are to falling in love with your own ideas—and the harder it is to let them go.
The Four Mindsets
We default to three mindsets that prevent rethinking. The goal is to think like a scientist.
| Mindset | Mode | Problem |
|---|---|---|
| Preacher | Delivering sermons | Protect and promote ideals |
| Prosecutor | Winning arguments | Prove others wrong |
| Politician | Seeking approval | Campaign for support |
| Scientist | Testing hypotheses | Seek truth through evidence |
The Preacher Mindset
"My beliefs are sacred truths."
Symptoms:
- Deliver sermons to the converted
- Defend beliefs with passion, not evidence
- Treat questioning as heresy
The Prosecutor Mindset
"Your beliefs are wrong and I'll prove it."
Symptoms:
- Focus on winning, not understanding
- Attack weak points in others' arguments
- Feel satisfaction when "winning" debates
The Politician Mindset
"I'll say what gets approval."
Symptoms:
- Adjust views to match the audience
- Avoid unpopular positions
- Prioritize acceptance over accuracy
The Scientist Mindset
"Let me test this and update based on evidence."
Characteristics:
- Treat beliefs as hypotheses
- Actively seek disconfirming evidence
- Update views when evidence warrants
- Find joy in being wrong (because it means learning)
Confident Humility
The sweet spot between arrogance and imposter syndrome.
CONFIDENCE IN ABILITY
HIGH │ LOW
────────────────┼────────────────
HIGH │ Armchair │ Imposter
CONFIDENCE quarterback syndrome
IN BELIEFS (arrogant) (paralyzed)
────────────────┼────────────────
LOW │ Confident │ Self-doubting
│ humility │ and ineffective
│ (optimal) │
Confident humility: High confidence in your ability to learn and figure things out. Low attachment to your current beliefs.
The goal: Strong opinions, loosely held.
The Joy of Being Wrong
Reframe being wrong as a positive experience.
The ego threat: Being wrong feels like an attack on our identity and intelligence.
The reframe: Being wrong means you're now less wrong than before. That's progress.
Practice:
- When proven wrong, say "I was wrong" out loud
- Thank people who correct you
- Celebrate updates to your thinking
- Track beliefs you've changed
The detachment principle: Separate your ideas from your identity. You are not your beliefs. Changing your mind is not changing who you are.
Binary Bias
We oversimplify complex issues into two opposing camps.
Reality: Most issues exist on a spectrum with multiple dimensions.
Examples:
- Not "capitalism vs socialism" but a spectrum of market/state mixes
- Not "nature vs nurture" but complex gene-environment interactions
- Not "pro vs anti" but many nuanced positions
Complexifying
Instead of picking sides, explore the complexity:
- What's the range of positions on this issue?
- What are the best arguments for each?
- Where might I be oversimplifying?
- What evidence would change my view?
Rethinking Cycles
Build rethinking into your regular practice.
Personal Rethinking
Belief audit: Periodically review your core beliefs.
- When did I form this belief?
- What was the evidence then?
- Has new evidence emerged?
- What would change my mind?
Prediction tracking: Make predictions, write them down, check results.
- Forecasts force precision
- Outcomes reveal blind spots
- Review builds calibration
Interpersonal Rethinking
Argue like you're right, listen like you're wrong.
In conversations:
1. State your view clearly
2. Genuinely listen to their response
3. Look for the valid parts of their perspective
4. Update your view where warranted
5. Acknowledge updates explicitly
Changing Other People's Minds
You can't force people to rethink. But you can create conditions that make it more likely.
What Doesn't Work
| Approach | Why It Fails |
|---|---|
| More facts | Backfire effect—they dig in |
| Stronger arguments | Triggers prosecutor mode |
| Emotional appeals | Triggers preacher mode |
| Social pressure | Triggers politician mode |
What Does Work
1. Ask questions instead of making statements
Questions invite reflection. Statements invite defense.
- "What evidence would change your mind?"
- "How did you come to that view?"
- "What would it mean if this turned out differently?"
2. Acknowledge the valid parts of their view
People can only hear you after they feel heard.
3. Find common ground
Start with shared values, then explore where you diverge.
4. Motivational interviewing
Help people find their own reasons to change:
- Express genuine curiosity about their perspective
- Ask what they see as the downsides of their current view
- Explore their underlying values and goals
- Help them see gaps between values and current beliefs
5. Plant seeds, don't expect harvests
Changing minds takes time. Your job is to create doubt and curiosity, not immediate conversion.
Collaborative Arguments
Transform adversarial debates into collaborative truth-seeking.
The Task Conflict vs Relationship Conflict Distinction
Task conflict: Disagreement about ideas, decisions, approaches.
Relationship conflict: Personal friction, resentment, hostility.
Productive teams have high task conflict and low relationship conflict. They argue about ideas vigorously while respecting each other.
Creating Productive Conflict
- Establish shared commitment to truth over being right
- Separate ideas from people who proposed them
- Assign devil's advocates to argue against consensus
- Welcome dissent and reward people who disagree well
- Argue then decide, don't decide then argue
Building a Rethinking Culture
For Teams
- Psychological safety: Make it safe to admit "I don't know" and "I was wrong"
- Process accountability: Reward good thinking processes, not just good outcomes
- Routine skepticism: Regular reviews asking "What if we're wrong?"
- Diverse perspectives: Include people who think differently
For Organizations
- Knowledge networks: Systems to surface challenges to conventional wisdom
- Retrospectives: Regular review of decisions and their outcomes
- Forecasting tournaments: Track predictions to build calibration
- Learning from failure: Post-mortems without blame
The Rethinking Scorecard
Track your intellectual flexibility:
| Behavior | Score Yourself |
|---|---|
| Times you said "I was wrong" this week | |
| Beliefs you've updated this month | |
| Questions you asked before stating opinions | |
| Times you sought disconfirming evidence | |
| Arguments where you changed your mind | |
| People you thanked for correcting you |
Application Framework
When You Hold a Strong Belief
- What's my evidence for this belief?
- When did I last update this view?
- What would change my mind?
- Who disagrees, and what's their best argument?
- Am I in preacher/prosecutor/politician mode?
When Trying to Change Someone's Mind
- Am I trying to win or trying to help them think?
- Have I genuinely understood their perspective?
- What do they value that I can connect to?
- Can I ask questions instead of making statements?
- Am I willing to update my own view?
When Making Decisions
- What hypotheses am I testing?
- What evidence would prove me wrong?
- Who disagrees and why?
- Have I considered alternatives seriously?
- How will I know if I need to rethink?
Key Mantras
| Mantra | Meaning |
|---|---|
| "I might be wrong" | Default to intellectual humility |
| "What would change my mind?" | Stay open to new evidence |
| "Argue like you're right, listen like you're wrong" | Balance conviction with openness |
| "Strong opinions, loosely held" | Be decisive but updatable |
| "I'm not my ideas" | Detach ego from beliefs |
# Supported AI Coding Agents
This skill is compatible with the SKILL.md standard and works with all major AI coding agents:
Learn more about the SKILL.md standard and how to use these skills with your preferred AI coding agent.