"This seems great for some d..."

https://arbital.com/p/1yz

by Eric Bruylant Feb 12 2016 updated Feb 12 2016


This seems great for some domains (especially rapidly evolving scientific fields and many political/policy things). How clear an idea of how this would work do you have? What provisions for updating predictions would there be? And how would you incentivise early predictions? A full prediction market, with some structural things to make people not run out of points entirely?

Edit: Moved the rest of this comment to 14s.


Comments

Alexei Andreev

For now we'll just do Arbital on invite-only basis. This feature is still pretty far off in the future, so I'm hesitant to say anything publicly, but we can discuss this privately if you want.

If Arbital sees that your votes accurately predict experts' votes, then you'll get more dominion. So in some cases it's actually really advantageous to vote first.

Eric Bruylant

Okay. I'm happy to be a sounding board/give input when you feel like this is one of the next steps.

And, I see. That helps give rapid feedback, though could cause distortions (people trying to predict expert predictions rather than the thing) if it was highly weighted. And even with that you'd need a way to make it worthwhile to not optimize by waiting until just before the deadline before voting (or, in many cases, waiting until there is actual confirmation of something and getting points for "predicting" with very high certainty).