newsfilter.io
Interview, Fireside Chat

Highlights: Hugo Mercier on why gullibility and misinformation are overrated

  • Predicts humans have evolved to accept vast amounts of social information but risks communication system collapse if truth-fiction discrimination fails.
  • Anticipates AI will generate overwhelming low-quality or misleading content, increasing noise and making credibility cues easy to fake.
  • Foresees cumulative information clutter causing practical reasons for the public to abandon strong views and stop efforting to learn about events.
  • Notes that AI-driven confusion could replicate government strategies to induce public disengagement and prevent rebellion.
  • Observes current mistrust regarding specific events, such as the Russian invasion of Ukraine, is already causing reduced effort in understanding complex issues and predicts this will worsen.
  • Expects the public will continue relying on curation and source trust for news evaluation, assuming content remains plausible.
  • States uncertainty exists over whether LLMs will fundamentally hinder journalist verification, though human source checking remains necessary.
  • Argues that discrediting trustworthy agents to increase mistrust is only effective if the initial trust level in the system is very low.
  • Suggests individuals turning to conspiracy theories often lack professional or social status they feel deserved.
  • Warns that without nurturing institutional contexts, individuals creating their own epistemic environments are likely to adopt elaborate but incorrect theories.
  • Identifies the motivation behind conspiracy work as comparable to the feeling of a journalist uncovering a story, despite negative outcomes.
  • Anticipates vaccine hesitancy will persist as a human nature trait, with a few percent staunchly opposing and 10–20% or more hesitating across societies.
  • Notes that while vaccine justifications vary by culture, the underlying intuitive resistance to the concept is universal.
  • Reports that in the overwhelming majority of experimental cases, providing strong factual arguments leads to the expected belief changes.
  • Cites only two documented cases of the "backfire effect": one regarding the Iraq War and another involving a specific vaccine in a specific population.
  • Suggests people with intuitive conspiracy beliefs often fear repercussions or death if they speak out, unlike those with reflective beliefs.
  • Expects conspiracy theories concerning current powerful actors are less likely to be true the more vocal and widespread they are.
  • Concludes that people generally make better trust judgments by erring on the side of distrust rather than over-trusting.