Will the EA Forum or LessWrong start using EigenKarma by 2025 (public experiments included)?
➕
Plus
26
Ṁ3528
2025
5%
chance

Here's a description of what EigenKarma is: https://www.lesswrong.com/posts/Fu7bqAyCMjfcMzBah/eigenkarma-trust-at-scale

A small-scale but public (i.e., visible to every forum user) experiment is sufficient to make this question resolve "Yes", regardless of the outcome of the experiment.

person inside a network of trust, 8k, beautiful, golden, illustration, trending on art station, picture of the day, poetic

Feb 10, 3:12pm: Will the EA Forum or LessWrong start using EigenKarma by 2025? → Will the EA Forum or LessWrong start using EigenKarma by 2025 (public experiments included)?

Get
Ṁ1,000
and
S3.00
Sort by:

I propose a possibly better system called EigenFlow (I want an impact cert if it's actually implemented):

Users can manually set the portions of their feed
1) to display whatever is equal to someone other's feed
2) to display someone's posts

I think this is an uncanny valley between rss/substack/mastodon and social media recomendation algorithms.
People are used to reddit/hackernews/stack-overflow style upvoting, so if they see a number next to some arrows they'll think it's the same number for everyone. If you customize the score for everyone should probably be hidden ... but then it's just an algorithmic feed that recommends you content based on your actions, like tiktok, except worse because its more restricted. May as well go full magic at that point

@Sinclair Thanks! I used the ideas in your comment to inform my comment on the LW post https://www.lesswrong.com/posts/Fu7bqAyCMjfcMzBah/eigenkarma-trust-at-scale?commentId=W5rJZbEsfR4FgLJet

© Manifold Markets, Inc.Terms + Mana-only TermsPrivacyRules