Summary of "Range"

Core Idea

Breadth beats depth in uncertain domains: generalists with diverse experience outpredict narrow specialists, make better career choices, and solve novel problems
Expertise is dangerous: deep knowledge creates blind spots, functional fixedness, and overconfidence in your own domain
Organizations fail when they silence lateral perspectives and treat failure as personal weakness instead of information

Specialize only in "kind" domains: chess, golf, medicine with clear feedback loops and repeating patterns
Stay broad in "wicked" domains: geopolitics, innovation, strategy, forecasting—where rules shift and past success breeds blind spots
Delay specialization: explore multiple "possible selves" first; late specializers make better long-term career fits than early commitments

"Hedgehog experts" (narrow focus) fit any outcome to their worldview; their predictions worsen as they accumulate domain knowledge
"Foxes" (broad, integrative thinkers) outperform on forecasting by integrating diverse perspectives
Functional fixedness traps experts—they can't see novel uses for existing tools because their domain knowledge narrows possibilities
Overconfidence grows with specialization, not expertise—deeper knowledge doesn't equal better judgment

Use reference classes: find structurally similar past events instead of treating current problem as unique
Generate contrary information actively: seek perspectives that conflict with your instinct, not just confirmation
Treat prediction failures as feedback: adjust beliefs after losses, not just wins
Apply analogical thinking across distant domains to escape specialist intuitions
Hold your tools lightly—audit which methods you'd defend reflexively vs. abandon for evidence

Separate information flow from decision authority: lateral/upward communication stays free while hierarchy makes final calls
Seek informal one-on-one perspectives: don't rely on formal meetings alone; surface doubts people won't voice publicly
Reward problem-exposure, not just problem-solving: failures must be treated as valuable data, not personal weakness
Build deliberate cross-pressures: if hierarchical, encourage dissent; if autonomous, reinforce shared responsibility
Create offline safe spaces for dissent, then reintegrate those conversations into official decisions—don't let concerns stay hidden

Combine polymaths (broad experience + one deep expertise) with narrow specialists
Integrate diverse perspectives rather than deepening single specialization
Reward exploring "adjacent" domains alongside core expertise
Use reference classes and contrary evidence as team decision protocols

Audit your domain: Does it have clear feedback (specialize) or shifting rules (stay broad)? Adjust commitment accordingly
Delay major career decisions: Explore 2-3 "possible selves" before locking in; preferences change
Create a personal forecast log: Make predictions, record reasoning, adjust beliefs rigorously after outcomes—not just when you're wrong
Seek lateral input before deciding: Ask one-on-one for perspectives you'd miss in formal settings; actively surface dissent
Hold analogies from distant fields: When stuck in domain expertise, pull examples from unrelated industries or disciplines