A guess: if LessWrong implemented this, onboarding lots of new users at once would be easier to do without ruining the culture for the people already here.
Another guess: this tool will accentuate political divide among any group that use it without acute awareness of this effect and a well chosen set of countermeasures.
The countermeasures? That’s a difficult question, but it should start by measuring the effect. I’d probably go with an ICA, then compute some ratio to test for increased polarisations for « hot » topics following the introduction of this scoring method.
But maybe you are asking why this effect tends to happen in the first place? Depending on your background, one of the two following explanations might best suit you:
-On a common sense level, the evidences for polarisation from social network are overwhelmingly clear, so any tool that looks like it can help construct a social network is at risk of being dangerous.
-On a more rationalist-seeking level, I think the key thought is to notice we can replace the label « trust » in propagating trust by the label « ingroup » as in propagating (feeling of belonging)
A guess: if LessWrong implemented this, onboarding lots of new users at once would be easier to do without ruining the culture for the people already here.
Another guess: this tool will accentuate political divide among any group that use it without acute awareness of this effect and a well chosen set of countermeasures.
Can you elaborate as to how you see this happening?
The countermeasures? That’s a difficult question, but it should start by measuring the effect. I’d probably go with an ICA, then compute some ratio to test for increased polarisations for « hot » topics following the introduction of this scoring method.
But maybe you are asking why this effect tends to happen in the first place? Depending on your background, one of the two following explanations might best suit you: -On a common sense level, the evidences for polarisation from social network are overwhelmingly clear, so any tool that looks like it can help construct a social network is at risk of being dangerous. -On a more rationalist-seeking level, I think the key thought is to notice we can replace the label « trust » in propagating trust by the label « ingroup » as in propagating (feeling of belonging)
Do you mean appearance or appurtenance?
A bug in my internal translator, thanks for signaling it. :-)
(I also added a link for the ingroup concept, for the today’s lucky ten thousand)