The Conspiracy Glitch

The brilliant quantum computing theorist Scott Aaronson just got an MIT job offer, and points us to a cartoon on conspiracy:

Conspiracy theories represent a known glitch in human reasoning.  The theories are of course occasionally true, but their truth is completely uncorrelated with the  believer’s certainty.  For some reason, sometimes when people think they’ve uncovered a lie, the raise confirmation bias to an art form.  They cut context away from facts and arguments and assemble them into reassuring litanies.  And over and over I’ve argued helplessly with smart people consumed by theories they were sure were irrefutable, theories that in the end proved complete fictions.

Passionately expressed, but it is really clear that people suffer more from overconfidence and confirmation bias in their beliefs about lies than in other beliefs?

GD Star Rating
loading...
Tagged as:
Trackback URL:
  • http://profile.typekey.com/halfinney/ Hal Finney

    I think you can at least generalize to a large spectrum of non-mainstream views. I’ve known smart people who believed in ESP, cold fusion, abiogenic oil, Drexlerian nanotech, non-HIV AIDS, Peak Oil catastrophism, and more. Most of them not only believe that the evidence on balance leans their way, but that the evidence is overwhelmingly in their favor and that the mainstream is being stubborn and close-minded. It shades into conspiracy but often doesn’t cross that line.

    And of course the other side of the coin is that some conspiracy theories turn out to be rather plausible. Was there more than a lone gunman involved in Kennedy’s assassination? Congress found that it was likely, but it is still widely considered the canonical conspiracy theory.

  • http://profile.typekey.com/robinhanson/ Robin Hanson

    Hal, ESP, cold fusion, etc. are not theories about lies, so it seems you reject the lie bias theory proposed. Do you propose a “non-mainstream” bias, that we suffer more from confirmation bias when considering non-mainstream views?

  • http://profile.typekey.com/sentience/ Eliezer Yudkowsky

    There’s almost certainly a bias toward believing more strongly that which we have to argue with other people.

  • http://profile.typekey.com/robinhanson/ Robin Hanson

    Eliezer, do non-mainstream views make one have to argue more than mainstream views? If so, this would be a proposed cause of an anti-mainstream bias.

  • Douglas Knight

    My impression is that the evidence for ESP is extremely similar to the evidence that a minimum wage affects employment. I think the evidence for cold fusion and the effect of health care on health are pretty similar, too. Why does consensus respond differently?

  • _Felix

    Hey, I just thought of something. If faced with a liar in real life, acting confident and omniscient is a strategy for getting him to crack and reveal whatever tangled web he is concealing. So the same thing may apply with conspiracy theorists, who act intuitively as if anybody they speak to is part of the cover-up, or at least in denial.

    That’s one factor; another is that it is an art form, because conspiracy theories are (for some people) entertaining conversational fare, like ghost stories, best told under a veil of sincere belief.