Good judgment begins by guarding against self-deception

Today’s reflection

In his 1974 Caltech commencement address, physicist Richard Feynman called research that imitates the procedures of science while missing its essential integrity ‘cargo cult science.’ He argued that we must disclose not only results that support our case, but also every relevant fact that could cast doubt on it. Here, honesty means more than refraining from falsehood; it means actively revealing how we might be wrong. Today's English sentence is a short direct quotation from the transcript preserved by Caltech, and the Korean sentence is a natural translation.

1. Sincerity alone cannot prevent self-deception. Because we can honestly believe an interpretation that favors us, we need a process that deliberately searches for contrary evidence.
2. Examine the strength of the test, not the strength of the confidence. A polished explanation that has not faced serious objections may be hope with excellent formatting rather than judgment.
3. A failed prediction is data for revising a model, not a threat to one's dignity. When reality differs from the forecast, record which assumption failed instead of rewriting the story after the fact.

The clearer our preferred outcome is in strategy, leadership, or marketing, the greater the risk of self-deception. We may exaggerate the promise of a project after investing in it, remember only favorable customer reactions, or mistake a team's silence for agreement. AI can also produce plausible support for the assumptions embedded in our prompt. The quality of judgment depends less on how persuasively we explain a conclusion than on whether it survives serious contrary evidence.