The Glory System: A Model For Moral Currency And Distributed Self-Moderation

post by Darklight · 2021-02-19T16:42:48.980Z · LW · GW · 16 comments

Contents

16 comments

One of the most important changes in our civilized modern society over the past historical tendencies has been the state's monopoly on the use of violence.  Most everyone today aside from the various strains of anarchist agree that this is a necessary and important requirement for modern life, but perhaps a problematic side effect of this change is that the use of force by individuals as a negative reinforcement tool to shape behaviour is lost.  Whereas in the distant past, a slight against one's honour could be challenged by the threat of a duel, today, positive reinforcement through transfers of money are the main way human behaviour is shaped.

Markets rely on these positive reinforcement signals to aggregate information about people's preferences.  But in some sense, traditional currencies are one-sided.  In effect, you can only vote Yea or Abstain with your dollar.  What if, you could also vote Nay?

Certainly the government can enforce fines as a way to provide negative reinforcement to bad market actors.  This punishment of negative externalities is one of the strongest justifications for government intervention.  But government interventions are often both heavy handed and slow to respond.  What if individuals could spend money to destroy someone else's money?

This somewhat radical and odd idea forms the basis of what I propose to be the glory system.  In the glory system, individuals can both transfer glory to others as a form of positive reinforcement, and spend glory to destroy other people's glory as a form of negative reinforcement.  It essentially functions as a system of distributed reward and punishment, as a true method of social credit.

Previously I had considered if there was a way to implement this idea in the form of a cryptocurrency.  But realistically, I have doubts that an actual currency that can disappear or be destroyed by others legally would have much interest as a store of value.  So what can this idea perhaps be more reasonably applied to?

One obvious comparison to be made with the the glory system is the karma system used on Reddit and here on Less Wrong.  Both allow both upvotes and downvotes in a sense.  The glory system as implemented in a forum reputation system would have some interesting characteristics.

Let's assume that glory is accumulated by upvotes on your posts and removed by downvotes.  This would allow users to bank upvotes and spend them later to add additional upvotes or downvotes, as much as they were willing to pay, beyond the first free vote.  Coupled with rules that automatically delete or hide posts that are downvoted to zero, and prevent users with zero or negative glory from posting, this can function as a kind of distributed self-moderation system.

It also provides a natural way to protect against spammers, by setting it so that new accounts have zero glory, and must be gifted glory by existing users to be able to post.  This would effectively function as a kind of application or invitation system, to ensure that new accounts were vouched for by existing members of the community.

Probably the most important feature however is that it allows not just the direction but also the magnitude of individual's preferences to be signaled.  Thus, for instance, a contrarian who really cares about a post, can spend glory to defend it and keep its rating above zero in proportion to how much they are willing to pay.  Conversely, if someone sees a particularly objectionable post, they can immediately attempt to bury it.  In aggregate the result functions like a kind of market of ideas, with the potential for bidding wars over particularly controversial notions to be played out over time.

A potential issue that could arise if the exchange rate of glory to votes is one-to-one, is that two accounts could farm upvotes by trading glory back and forth indefinitely.  The simplest solution to this is to set the exchange rate to be 2-to-1 or greater.  This means you have to spend two glory to make one glory, which leads to diminishing returns.

An interesting side effect of the exchange rate is that defending your own posts with upvotes is cheaper than attacking other posts with downvotes.  This could mitigate the potential issues of harassment that plague other forum reputation systems.  Adjusting the exchange rate also determines the relative weighting of the old guard versus newbies and the crowd.  In essence, it allows fine-grained control over to what extent the system balances meritocracy and democracy.

I am tempted to try implementing this system.  Given that moderation on Less Wrong is already formidable and effective enough, I don't think it's necessary to implement it here, but as an interesting experiment, I'm considering forking the Less Wrong 2 codebase and building my own little test forum and deploying it somewhere and inviting friends to join.  Before I do so however, I want to put the idea out to the Less Wrong community to ask for opinions and feedback on the feasibility of the glory system, and any potential pitfalls or critiques I may have overlooked in my enthusiasm for the idea.

16 comments

Comments sorted by top scores.