Risk vs Uncertainty¶

Why It Matters¶

Insurance works for quantifiable risks (actuarial tables exist). It fails for novel events where the distribution is unknown. Financial models that treat uncertainty as risk blow up spectacularly when unprecedented events occur.

The Key Move¶

Ask: do I know the probability distribution? If yes, expected value/utility calculations apply. If no, you need different tools - robustness, optionality, scenario planning, heuristics. Don't pretend to know distributions you don't know.

Classic Application¶

Insurance pricing. Actuarial tables work for quantifiable risks - car accidents have stable statistical properties. They fail for unprecedented events (9/11, COVID) where there's no historical distribution to reference.

Surprising Application¶

Career planning. Some career domains have quantifiable risk - sales has a known distribution of outcomes. Others have Knightian uncertainty - founding a company in a new market has no reference class. The appropriate planning strategies differ fundamentally.

Failure Modes¶

• Pretending uncertainty is risk - assigning fake probabilities to unknowable events
• Paralysis when facing uncertainty - inaction is also a choice
• Confusing risk aversion with uncertainty aversion - they're different
• Treating all uncertainty as equal - some unknowns are more unknown than others
• Using expected value when you should use robustness

Go Deeper¶

Knight's "Risk, Uncertainty, and Profit" for the foundational distinction. Taleb's "Antifragile" and "Black Swan" for practical implications.