Less Wrong is a community blog devoted to refining the art of human rationality. Please visit our About page for more information.

IlyaShpitser comments on LW 2.0 Strategic Overview - Less Wrong

47 Post author: Habryka 15 September 2017 03:00AM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Comments (294)

You are viewing a single comment's thread. Show more comments above.

Comment author: IlyaShpitser 17 September 2017 01:08:59AM *  4 points [-]

Google is using a much more complicated algorithm that is constantly tweaked, and is a trade secret -- precisely because as soon as it became profitable to do so, the ecosystem proceeded to game the hell out of PageRank.

Google hasn't been using PageRank-as-in-the-paper for ages. The real secret sauce behind Google is not eigenvalues, it's the fact that it's effectively anti-inductive, because the algorithm isn't open and there is an army of humans looking for attempts to game it, and modifying it as soon as such an attempt is found.

Comment author: Wei_Dai 17 September 2017 02:04:03AM 8 points [-]

Given that, it seems equally valid to say "this will work, for the same reason that PageRank worked", i.e., we can also tweak the reputation algorithm as people try to attack it. We don't have as much resources as Google, but then we also don't face as many attackers (with as strong incentives) as Google does.

I personally do prefer a forum with karma numbers, to help me find quality posts/comments/posters that I would likely miss or have to devote a lot of time and effort to sift through.

Comment author: IlyaShpitser 17 September 2017 03:05:01PM 2 points [-]

It's not PageRank that worked, it's anti-induction that worked. PageRank did not work, as soon as it faced resistance.

Comment author: John_Maxwell_IV 18 September 2017 07:54:42AM 0 points [-]

You really are a "glass half empty" kind of guy aren't you.

Comment author: IlyaShpitser 18 September 2017 01:51:58PM *  3 points [-]

I am not really trying to be negative for the sake of being negative here, I am trying to correctly attribute success to the right thing. People get "halo effect" in their head because "eigenvalues" sound nice and clean.

Reputation systems, though, aren't the type of problem that linear algebra will solve for you. And this isn't too surprising. People are involved with reputation systems, and people are far too complex for linear algebra to model properly.

Comment author: Lumifer 19 September 2017 07:36:36PM 3 points [-]

people are far too complex for linear algebra to model properly

True, but not particularly relevant. Reputation systems like karma will not solve the problem of who to trust or who to pay attention to -- but they are not intended to. Their task is to be merely helpful to humans navigating the social landscape. They do not replace networking, name recognition, other reputation measures, etc.