carado.moe

tammy's blog about
AI alignment, utopia, anthropics, and more;

also on twitter, lesswrong, rss, and at

Overcoming Narratives

who drinks more alcohol, liberals or conservatives ?

if you have an existing opinion on both of those (such as: drinking alcohol is bad, and conservatives are better than liberals), then you're likely to just come to a default assumption (such as: liberals drink more).

if you want to be more reasonable about it, you may try to think of explanations. "maybe liberals drink more because this and this! that sounds like it makes sense." but it's pretty likely you could come up with an explanation for the opposite case (conservatives drinking more) just as easily!

and just like any piece of evidence: if it's as easy to come by in both cases, then it's not actually telling you anything.

the only thing you get by coming up with potential narratives that sound like they make sense but aren't evidenced by anything in the world, is to pre-entrench yourself in a position rather than be open to what may actually be the shape of reality; if there even is a shape at all to the topic at hand.

of course, just assuming to be true the narratives that other people tell you is worse, but even coming up with your own narratives is generally a bad idea.

in general, avoid narratives, unless evidenced; and when there's evidence, only make the minimum amount of assumptions from that narrative.