Nassim Nicholas Taleb (born 1960) is a former options trader and probability researcher whose work concerns how systems behave under rare, high-impact events and what follows for decisions made in ignorance of them.

Black swans

The Black Swan (2007) describes events that are unpredictable in advance, carry extreme consequences, and are explained confidently after the fact. The third property is the dangerous one: retrospective explanation creates the impression that the event was foreseeable, which supports the belief that the next one will be.

His target is the use of models assuming a normal distribution in domains where outcomes are not distributed that way. In finance, technology and social systems, a single observation can exceed the sum of everything recorded before it. A risk model calibrated on the observed range is at its most confident immediately before the observation that breaks it.

Antifragility

Antifragile (2012) introduces the property he argued has no name in English. Fragile systems are harmed by volatility. Robust systems resist it. Antifragile systems gain from it, within limits — muscle under load, a business that takes share during a downturn.

The practical form is asymmetry: arrange exposures so the downside is capped and the upside is not. His barbell strategy puts most resources in the safest available position and a small portion in speculative bets, avoiding the middle where a moderate loss is possible and a large gain is not. This connects to optionality and to decision under uncertainty.

Skin in the game

Skin in the Game (2018) argues that people making decisions should bear the consequences of them. Advice from someone insulated from the outcome is unreliable for a structural reason rather than a moral one: when they are wrong, nothing reaches them, so nothing corrects them. See antifragility.

His combative style has drawn criticism, and statisticians have disputed parts of his treatment of fat tails. The core observation — that models calibrated on the observed range fail exactly when it matters — is widely accepted.