3 ms·

Uh what ?

• chalst 5y ago
  Taleb deploys Bayes theorem and the base rate "fallacy" [fn] to show that Snowden is more likely to be a traitor than, uh, what he thought before he started studying Snowden. https://medium.com/incerto/snowden-phony-hence-traitor-86ee41197578?source=social.tw https://medium.com/incerto/snowden-phony-hence-traitor-86ee4... It's really as bad as the literal URL suggests. [fn]: To be fair, the idea that base rate bias is a fallacy is asserted in Wikipedia, so that's not originally his mistake. What base rate bias is, is a misconception that leads us to make poor estimates. It's not a fallacy, since fallacies are classes of inferences and, while there's a lot of debate as to what biases really are, I know of no logician who thinks they are a kind of inference.
  • orwin 5y ago
    This article is really a long-winded way to say "it's not me, it's you". The first to paragrph hold no purpose, I'm not on twitter but this is exactly the pettiness level i expect. Still, interesting article, validated my experience in the skeptic community and why i couldn't mesh well with the hard liners.
    • chalst 5y ago
      I guess that's right. I can't say I thought "interesting article", more that in writing it he condemns himself. "I can prove he's a traitor", followed by an argument that whose fundamental weakness becomes more apparent the longer it is reflected on, is not the noble exercise in patriotic duty he is trying to pass off his score settling as. It's the curse of the social media age: people whose great virtues might in another time have led themselves to universal admiration, in the hustle of self promotion allow their ugly vices to be seen.