Updated
Updated · The Guardian · Oct 31
Psychologist Warns AI Overuse Erodes Cognition, Urges 30-Minute Friction Rule
Updated
Updated · The Guardian · Oct 31

Psychologist Warns AI Overuse Erodes Cognition, Urges 30-Minute Friction Rule

1 articles · Updated · The Guardian · Oct 31

Summary

  • Carly Dober says heavy AI use can weaken learning and problem-solving by encouraging cognitive offloading—letting tools do the mental work needed to build and maintain neural pathways.
  • That over-reliance also feeds automation bias and complacency, she argues, making users trust chatbot output too readily even though large language models can still produce inaccurate answers.
  • Meta’s new Muse agent—designed to handle forms, calendars, bills and reservations—arrives as Dober warns convenience-driven tools remove the mental struggle the brain needs to stay sharp.
  • Dober recommends adding deliberate barriers: wait 30 minutes before asking AI for help, write first drafts yourself, log out of apps, remove AI tools from phones and schedule screen-free time.
  • She links the same convenience logic to privacy risks, saying users often trade data for ease despite “consent fatigue” and limited awareness of how aggregate data can shape buying and political behavior.

Insights

Can an AI assistant built for convenience protect privacy and human judgment, or does frictionless design make overreliance inevitable?
Will Meta’s Muse save people time—or quietly train them to trust AI with money, decisions, and personal data they barely understand?
If users must approve every sensitive action, who is really accountable when an AI agent makes the wrong move?