Comprehensive notes on Daniel Kahneman's masterwork "Thinking, Fast and Slow" — the definitive guide to how we think, why we make mistakes, and how to make better decisions.
"Thinking, Fast and Slow" (2011) by Nobel laureate Daniel Kahneman summarizes decades of research on judgment and decision-making. It has sold millions of copies and fundamentally changed how we understand human cognition.
- Automatic, effortless, always on
- Pattern recognition, intuition, emotional responses
- Generates impressions, feelings, and inclinations
- Cannot be turned off
Examples: Reading text, detecting hostility in a voice, driving on an empty road, completing "2 + 2 = ?"
- Deliberate, effortful, requires attention
- Complex calculations, conscious reasoning
- Monitors and sometimes overrides System 1
- Lazy — uses System 1 whenever possible
Examples: Counting occurrences of "a" in a page, parking in a narrow space, comparing two products on multiple attributes
Most of what we think and do originates in System 1, but we believe it comes from System 2. We have an illusion of deliberation when we're actually running on autopilot.
The first number you encounter strongly influences subsequent judgments, even when the anchor is random.
Experiment: "Is the percentage of African nations in the UN more or less than 10%/65%?" People anchored on 65% guessed 45%; those anchored on 10% guessed 25%.
We judge frequency and probability by how easily examples come to mind.
Consequence: We overestimate risks of dramatic events (plane crashes, shark attacks) and underestimate risks of common events (heart disease, car accidents).
We judge probability by how well something matches our mental stereotype.
The Linda Problem: "Linda is a bank teller" vs. "Linda is a bank teller who is active in the feminist movement." Most people rate the conjunction as more probable — a logical impossibility.
We ignore statistical base rates when presented with vivid individual information.
People systematically underestimate time, costs, and risks while overestimating benefits. This applies to:
- Home renovations (average 30% over budget)
- Software projects (average 50% over schedule)
- New businesses (80% fail within 5 years)
Countermeasure: Reference class forecasting — look at how long similar projects actually took, not your optimistic plan.
System 1 builds the best story from available information and doesn't consider what information is missing. This creates:
- Overconfidence in thin evidence
- Failure to seek disconfirming data
- Narrative fallacy — creating coherent stories from random events
Losses are psychologically about twice as powerful as equivalent gains. Losing $100 feels worse than gaining $100 feels good.
Implications:
- People take irrational risks to avoid certain losses
- Endowment effect: valuing what you own more than identical items you don't
- Status quo bias: preferring the current state to change
How a choice is presented dramatically affects decisions:
- "90% survival rate" vs. "10% mortality rate" — same information, different choices
- "Save 200 of 600 people" vs. "400 will die" — same outcome, different preferences
- The experiencing self lives in the present moment
- The remembering self creates the story of your life
Peak-End Rule: We judge experiences by their peak intensity and how they end, not by their total duration or average quality.
Duration Neglect: A 30-minute painful procedure and a 60-minute equally painful procedure are remembered similarly if they have the same peak and ending.
- Slow down for important decisions — force System 2 engagement
- Use base rates — check statistical data before relying on intuition
- Consider the reference class — how did similar situations turn out?
- Reframe the question — would your answer change with different framing?
- Pre-mortem analysis — imagine failure and work backwards
Discover decision-making principles from Kahneman and other brilliant thinkers at KeepRule — where cognitive science meets practical frameworks for better judgment.
This project is licensed under the MIT License - see the LICENSE file for details.