Questions for you:
- When something unexpected happens at work, which of the terms in the story do you and your colleagues reach for most readily? What does that default vocabulary reveal about how your organisation thinks about uncertainty?
- Can you think of a situation that was described as bad luck but was actually epistemic uncertainty — where more information was available and would have changed the outcome?
- How precise is the language your organisation uses when discussing risk, probability, and chance? Do different people mean different things by the same words? Across teams? Across cultures?
Organisational applications:
The cost of imprecise language about uncertainty: Sherman Kent’s Words of Estimative Possibility were invented because intelligence analysts were using words like “probable” and “likely” to mean completely different things, with operational consequences. The same problem exists in most organisational risk conversations. When a project manager says a deadline is “likely” to be met, and a board member hears “almost certain” while the project manager meant “better than even,” the gap between those interpretations can drive genuinely different decisions.
Establishing shared definitions for probability language — not necessarily Kent’s exact scale, but any agreed-upon framework — is a low-cost intervention that substantially reduces the noise in risk communication. The alternative is continued reliance on terms that feel precise but are systematically ambiguous.
Distinguishing aleatoric from epistemic uncertainty changes what you do: The page’s most practically useful distinction is between aleatoric uncertainty, where the outcome is genuinely random and no further information can resolve it, and epistemic uncertainty, where the outcome appears random but is in fact knowable with more data or analysis. Organisations that treat epistemic uncertainty as aleatoric stop investigating when they should be researching.
Those who treat aleatoric uncertainty as epistemic waste resources on analysis that cannot produce certainty. A supply chain disruption caused by a genuine random event requires resilience and contingency planning; one caused by insufficient supplier monitoring requires better information systems. Misidentifying which type of uncertainty you face leads directly to the wrong response.
Luck versus serendipity as a strategic distinction: The story distinguishes luck, which is unintentional and happens to you, from serendipity, which is active and can be cultivated. This is not merely a semantic difference. Organisations that attribute positive, unexpected outcomes entirely to luck have no practical concept of improving the frequency of those outcomes.
Those who understand serendipity as the product of surface area — exposure to diverse inputs, openness to unexpected connections, investment in relationships outside the immediate task — can deliberately increase the conditions that make fortunate discoveries more likely. The distinction matters most when deciding whether to systematise a successful accident or treat it as a one-off.
Further reading
On precision in probabilistic language and communication:
Superforecasting: The Art and Science of Prediction by Philip Tetlock and Dan Gardner. Tetlock’s account of forecasting practice is substantially about the discipline of precise probability language — assigning numbers rather than words, and being held accountable for them. The Sherman Kent framework the story references is a direct predecessor to the approach Tetlock advocates.
Noise: A Flaw in Human Judgement by Daniel Kahneman, Olivier Sibony and Cass R. Sunstein. The chapters on communication and calibration cover the organisational consequences of imprecise probability language in professional contexts, with evidence for how much variability results from people using the same words to mean different things.
On the distinction between types of uncertainty:
The Black Swan: The Impact of the Highly Improbable by Nassim Nicholas Taleb. Taleb’s distinction between Mediocristan and Extremistan maps onto the aleatoric/epistemic divide in useful ways, particularly for understanding which domains reward more analysis and which require accepting irreducible randomness.
Thinking, Fast and Slow by Daniel Kahneman. Kahneman’s treatment of the difference between risk, where probabilities are known, and uncertainty, where they are not, provides the cognitive science background to the terminological distinctions the story draws.
On serendipity, luck, and what can be cultivated:
The Luck Factor by Richard Wiseman. Wiseman’s research directly addresses the luck/serendipity distinction, demonstrating empirically that the behaviours associated with experiencing more fortunate outcomes are learnable, which is the practical implication of treating serendipity as distinct from pure luck.
Against the Gods: The Remarkable Story of Risk by Peter L. Bernstein. Bernstein’s history covers how the vocabulary for talking about uncertainty developed over time, providing useful context for why the terms in the story carry the meanings they do and where those meanings came from.
About the image
The CIA Seal in the entrance lobby of the CIA headquarters.
Photo montage by Matt Ballantine, 2026 Photo: https://www.cia.gov/stories/story/the-history-of-cias-seal/
Random the Book 