Hacker News new | ask | show | jobs
by helicalmix 891 days ago
> This is the correct Bayesian action: if I know that a false argument sounds just as convincing as a true argument, argument convincingness provides no evidence either way. I should ignore it and stick with my prior.

If two diametrically opposed arguments are equally convincing, it's not intuitive to me that the "correct" solution is a middle ground. Especially if both argument state that a middle ground is not possible.
1 comments
twelfthnight 891 days ago
I guess if I encountered two diametrically opposing arguments of equal validity I would increase my uncertainty, not just stick to my prior. That's super different to just ignoring new arguments, right? Like going from "I think X is the best candidate" to "I don't know who the best candidate is".
link
helicalmix 891 days ago
Oh, that's a good point, and I think you're idea is a better behavior model vs. the "stick with prior" idea.

What if the conclusion forces behavior though? For voting, you can abstain from voting if you're unclear, but what if you're in a position where you have to choose two distinctly different options.

In this case, I think you pretty much have to stick with your prior, but I'm curious if you have a more nuanced approach here.

link
twelfthnight 891 days ago
> I think you pretty much have to stick with your prior

Yeah, at least to me that makes sense if you're forced to make a choice. Although in some ways if you really have no clear evidence it doesn't matter what choice you make, since both are equally reasonable.

> I'm curious if you have a more nuanced approach here

Not really :D. I do think things are often more complicated than "you must choose two options", for example you might have to choose between two candidates (in the US at least), but if you are more uncertain maybe you don't put a giant poster in your front yard or get upset at neighbors if they have a different opinion.

link
pixl97 891 days ago
In general this is not how biological life has evolved and not how most decisions work within that framework.

Most decisions you'll ever make are "You have ___ seconds/minutes/hours to make a choice". And depending on the pressures you're under you will skip the expensive logical/rational processing as needed and fall back to heuristics.

link