Post by Stanford University

What if the secret to human intelligence isn't processing more data — but knowing when to stop? New research from Douglas Guilbeault, assistant professor of organizational behavior at Stanford University Graduate School of Business, finds that people reach decisions and learn social conventions through "satisficing": gathering just enough information to make a confident, intuitive leap, rather than processing all available data the way a large language model does. In experiments with groups of 24 to 96 strangers, participants converged on a shared name for an unfamiliar face in roughly the same number of rounds regardless of group size. The best predictor wasn't more data — it was a simple threshold for when information becomes enough, the same principle linguists use to explain how toddlers learn grammar. The finding offers a clearer picture of what distinguishes human cognition from AI: the ability to make sense of incomplete information and reach an understanding others recognize as reasonable, even when the data is far from complete. Read the full story: https://lnkd.in/eUBk6Jie