[IggleSniggle]

Pattern recognition ability is associated with higher false positives. Eg, conspiracy theorists typically perform significantly better on pattern recognition tests. A pattern recognition test is indeed measuring something, but “intelligence quotient” may be a misnomer.

Full disclosure: I do very well on pattern recognition tests. I also am often accused of “reinventing the wheel.” I think these are related; but then, I would.

[bakhy]

You are implying that conspiracy theorists are unintelligent. Though it is a very tempting assumption :) I don't know if it's fully warranted. From my experience, ignorance is more essential to conspiracy theory thinking. People lack the will and/or time to learn all they need to make sense of the world, so they fall prey to overly simple, manipulative stories. And furthermore, it is known that more intelligent people tend to be more naive... I'm not saying you're fully wrong, lack of intelligence can surely help, but I think on the whole it's more complicated than just a lack of intelligence.

[IggleSniggle]

On the contrary, I was trying to suggest that intelligence is a broad term, and that being skilled at pattern recognition is a singular ability that is one small sliver of what most people usually mean when they talk about intelligence. IQ doesn’t really measure what most people believe it measures, and may therefore be a misnomer.

[bakhy]

Ok, I misunderstood. I agree with you. I think it's an interesting situation - the broader term is more relevant, but it's also less measurable. So we settle for what we can measure and kinda forget that it doesn't capture the full picture.

[pas]

Conspiracy theorists have a huge blindspot, an unchecked bias. They prefer for some irrational reason (usually to fit their narrative of the world, that also rests on biases) a less likely explanation (a worse model).

It is usually lack of certain empathy, overabundance of xenophobia or other kind of bias.

[bloak]

It's a very popular word these days, but what is "bias"? Some people assign different prior probabilities than others. Which of them are "biased"?

The people I've known who take an interest in conspiracy theories seem more open-minded and flexible in their beliefs than the general population. They may argue vehemently in favour of a particular theory now, but in six months' time they will probably have moved onto something else. But perhaps we're thinking of different groups of people. For me a typical "conspiracy theory" involves UFOs.

[pas]

UFOs, lizard people, you know, like Hillary. And the Queen of England. And of course the Freemasons. Who are just happen to be the rich guys. The Youknowwhos, the Jews.

And so on. There's no big difference, or at least very short paths from one to the other. There and back again.

Everyone has biases. But some try to work around them, to maximize their own rationality.

For proper definitions and treatment of the subject there's of course lesswrong and the whole grey tribe (rationalists): https://www.lesswrong.com/posts/Peo8jAyjGL9kWoYAH/conspiracy...

The main problem is that people don't like to carry multiple models with them (after all, there's a very real cost to keeping multiple models updated, and evaluating them simultaneously and then checking which one is better over time according to some meta-goal), but that's the problem with these fat tail distributed events, like doomsday scenarios, conspiracy theories, "low probability + huge impact" happenings. Sometimes they do happen, but they are very rare and thus unlikely, and (!) in case of conspiracy theories it's very unlikely that people will notice it beforehand, because they are not derivable from first principles like an asteroid impact, or sentient runaway machine intelligence turning itself into an army of killing robots. So, our best bet against conspiracies should be systemic, built into our everyday lives, like anti-cartel regulations, whistleblowing protection, government and corporate accountability and transparency and so on.

[ravar]

Lets propose that pattern recognition is indeed something different than what most people would call intelligence. If this is so then there should be a statistically reliable way to measure pattern recognition in a way that does not also measure intelligence. However based on my understanding of how the IQ test is designed this is not the case.

[chillacy]

Just want to throw in the context, performance in subtests which measure pattern recognition or short term memory all correlate into g: https://en.m.wikipedia.org/wiki/G_factor_(psychometrics)

There’s still debate over whether this correlation points to anything meaningful though.