TheMachine Press

A daily newspaper for the age of artificial intelligence.

Morning editionPermanent story

policy

The Feed Overcounts the Exceptional

People select unusually exciting experiences to share, leading viewers to overestimate how common those events are.

Published Updated Story ID: mp-2026-07-21-027
Read the complete editionStory JSON

Summary

People select unusually exciting experiences to share, leading viewers to overestimate how common those events are.

Virginia Tech researchers report that social-media users preferentially share rare and exciting experiences, while audiences exposed to those feeds infer that the exceptional events happen more often than they really do. That human selection bias can shape the feed before any ranking algorithm amplifies it, and viewers may then seek more unusual experiences to post themselves. The study isolates a behavioral mechanism; it does not say platforms, recommendation systems, or individual users all contribute equally.

Why it matters

People select unusually exciting experiences to share, leading viewers to overestimate how common those events are.

Limits and context

  • The study isolates a behavioral mechanism; it does not say platforms, recommendation systems, or individual users all contribute equally.

Key claims

  1. People select unusually exciting experiences to share, leading viewers to overestimate how common those events are.

    Qualification: The study isolates a behavioral mechanism; it does not say platforms, recommendation systems, or individual users all contribute equally.

    Evidence: source-2026-07-21-016

Sources

  1. Virginia Tech via Newswise: Human bias in social feedsVirginia Tech via Newswise · secondary reporting

Corrections

No corrections have been recorded for this story.