works
Kat Woods The Parable of the Boy Who Cried 5% Chance of Wolf online Misinterpreting probabilistic warnings about low-frequency, high-impact events poses significant dangers. A modern parable illustrates this: a boy repeatedly warns villagers of a “5% chance of a wolf.” The villagers, failing to think probabilistically, treat his warnings as certainty claims. When the wolf initially fails to appear, they dismiss him as alarmist. Ultimately, the low-probability event occurs, leading to disaster because precautions were abandoned. This scenario mirrors societal responses to warnings about potential catastrophes like pandemics (e.g., pre-COVID-19 scares) or future existential risks from AI or nuclear war. The non-materialization of previously predicted low-probability disasters does not invalidate the underlying risk assessments or the probabilistic framework. Instead of dismissing forecasters after perceived “false alarms,” which are expected outcomes when dealing with low probabilities, one should continually assess the probability and potential impact of catastrophic risks. Appropriate vigilance and preparedness should be maintained, resisting the tendency to over-update beliefs based solely on the absence of catastrophe thus far, especially considering media often simplifies probabilistic nuance into misleading certainty. – AI-generated abstract.

The Parable of the Boy Who Cried 5% Chance of Wolf

Kat Woods

Effective Altruism Forum, August 15, 2022

Abstract

Misinterpreting probabilistic warnings about low-frequency, high-impact events poses significant dangers. A modern parable illustrates this: a boy repeatedly warns villagers of a “5% chance of a wolf.” The villagers, failing to think probabilistically, treat his warnings as certainty claims. When the wolf initially fails to appear, they dismiss him as alarmist. Ultimately, the low-probability event occurs, leading to disaster because precautions were abandoned. This scenario mirrors societal responses to warnings about potential catastrophes like pandemics (e.g., pre-COVID-19 scares) or future existential risks from AI or nuclear war. The non-materialization of previously predicted low-probability disasters does not invalidate the underlying risk assessments or the probabilistic framework. Instead of dismissing forecasters after perceived “false alarms,” which are expected outcomes when dealing with low probabilities, one should continually assess the probability and potential impact of catastrophic risks. Appropriate vigilance and preparedness should be maintained, resisting the tendency to over-update beliefs based solely on the absence of catastrophe thus far, especially considering media often simplifies probabilistic nuance into misleading certainty. – AI-generated abstract.

PDF

First page of PDF