If this proposal were implemented, I might only upvote comments and articles that I think the top HN users would like, not necessarily the ones I like, turning

If this proposal were implemented, I might only upvote comments and articles that I think the top HN users would like, not necessarily the ones I like, turning HN into a cliché of HN. "How to bootstrap your minimum viable product using Node.js". "Scala, Clojure, or Erlang?". "LISP for Bayesian A/B testing." I'd much prefer a system which correlated my votes with other users and preferentially showed me articles and comments which matched my own tastes. Sure, if I only upvote to match my own biases, I'll get more biased articles. But if I also upvote good but contrarian opinions (and I would) I'll also get more good and contrarian opinions. Best of all, this encourages non-strategic voting--so, later on, if you find a good use for someone's voting record, you can trust the veracity of that record.