The Machine, the Expert, and the Common Folks

Human decision-making is subject to "noise," or random variability, which can lead to inconsistent outcomes like a judge's sentence being affected by hunger. This noise comes in several forms, including individual biases and transient factors like mood, and it impacts both objective predictions and subjective evaluations. Studies show that even simple, consistent mathematical models can outperform human experts in predictive tasks because they are free from this noise. However, humans still excel in situations requiring contextual understanding, like the classic "broken leg" problem where a model lacks crucial, last-minute information.