What Is Rationality as Survival? (Taleb Definition)
Rationality as survival is Nassim Taleb's alternative to the conventional economic definition of rationality (maximizing expected utility, following Bayesian updating). Taleb argues: what is rational is what allows you — and the collective — to survive.
The problem with conventional rationality:
Standard expected utility maximization breaks down under the ergodicity problem: when a sequence of events can hit ruin — an absorbing state you cannot exit — maximizing expected value can systematically lead to extinction. The "rational" actor who takes positive expected value bets that carry ruin risk will eventually go bankrupt. That's not rational by any meaningful definition.
What survival-rationality implies:
-
Overestimating low-probability catastrophic risks is rational. If underestimating them leads to ruin, and ruin is non-recoverable, the cost of over-caution (some missed opportunities) is asymmetric with the cost of under-caution (extinction). The apparently "irrational" fear of catastrophe is the correct response.
-
Following ancestral taboos without understanding them may be rational. If a prohibition has survived across many generations under real conditions, it has passed a Lindy test. The survival is evidence. The 40-person study that claims to refute it has passed a peer review, which is much weaker evidence.
-
Never risking ruin is rational. Even when the expected value is positive, if the failure mode is non-recoverable, the bet should not be taken. The Kelly Criterion implements this: size bets as a fraction of total resources so that ruin is never possible.
The grandmother who avoids certain behaviors she can't fully explain may be more rational than the statistician who recommends the opposite — because the grandmother's constraint is calibrated to multi-generational real-world testing, while the statistician's is calibrated to a single study.
For the full framework, read Ergodicity, Ruin, and Rational Risk-Taking.