Eigenkarma should be rooted in the trust of a few accounts that are named in the LW configuration. If this seems unfair, then I strongly encourage you not to pursue fairness as a goal at all—I’m all in favour of a useful diversity of opinion, but I think Sybil attacks make fairness inherently synonymous with trivial vulnerability.
I am not sure whether votes on comments should be treated as votes on people. I think that some people might make good comments who would be bad moderators, while I’d vote up the weight of Carl Schulman’s votes even if he never commented.
Thank you all so much for doing this!
Eigenkarma should be rooted in the trust of a few accounts that are named in the LW configuration. If this seems unfair, then I strongly encourage you not to pursue fairness as a goal at all—I’m all in favour of a useful diversity of opinion, but I think Sybil attacks make fairness inherently synonymous with trivial vulnerability.
I am not sure whether votes on comments should be treated as votes on people. I think that some people might make good comments who would be bad moderators, while I’d vote up the weight of Carl Schulman’s votes even if he never commented.
The feature map link seems to be absent.
What are Sybil attacks?
Vote manipulation with sockpuppets, apparently.
Feature roadmap link fixed!