2026-07-23
Guidelines on How to Think
- Pause before you decide: Give yourself time to avoid impulsive errors (combat availability and recency biases)
- Seek disconfirming evidence: Look for reasons you might be wrong (avoid confirmation bias)
- Quantify outcomes: Use numbers, probabilities, and base rates instead of gut feelings (counteract insensitivity to sample size and vividness)
- Simplify choices: Remove irrelevant options and frame decisions clearly (reduce choice overload and framing effects)
- Compare to appropriate baselines: Use reference classes and benchmarks, not isolated anecdotes (avoid survivorship and selection bias)
- Separate short-term noise from long-term trends: Ignore irrelevant fluctuations and focus on underlying processes (prevent recency and extrapolation errors)
- Use checklists and pre-commitment: Rely on rules, algorithms, or written plans to limit biases and emotional reactions
Kahneman and Tversky (System 1 vs System 2)
What you can affect (train, nudge, or manage)
- Deliberate reflection (System 2): improve through practice, education, and tools (e.g., logic, statistics, decision rules)
- Decision procedures and habits: use checklists, algorithms, rubrics, and pre-commitment to bypass automatic errors
- Statistical thinking: teach base rates, probabilities, and sampling principles to reduce mistaken intuitions
- Choice architecture/nudges: change environments, defaults, and incentives to steer System 1 responses
- Attention allocation: train to notice when to engage System 2 (trigger cues, pausing routines)
- Use of external aids: calculators, decision trees, and written plans that constrain impulsive choices
What you generally cannot directly change (hard to retrain in System 1)
- Automatic intuitions and heuristics: fast, unconscious rules (e.g., anchoring, availability, affect heuristic) are persistent and resistant to direct willful correction.
- Emotional reactions and immediate framing effects: visceral responses and gut feelings are difficult to eliminate through instruction alone.
- Innate cognitive limits: working memory capacity and processing speed are largely fixed.
- Slow attenuation of deep biases: mere awareness of a bias rarely removes it from System 1; reductions are modest without structural changes.
Focus training on strengthening System 2, creating rules and environments that compensate for System 1, and using nudges or procedural safeguards rather than trying to eradicate automatic intuitions.