[Question] How important is it that LW has an unlimited supply of karma?


LessWrong users can up/​down­vote posts and com­ments, which then re­ceive a karma boost (capped by the vot­ers own karma). There is no limit to how many differ­ent posts and com­ments one can do this to. In this sense there is an un­limited sup­ply of karma to be handed out. (This is also the case for Face­book, YouTube, In­sta­gram, Hack­erNews(?), Medium, …)

Is this im­por­tant? That is, does it have non-triv­ial medium or long-term effects on the “LessWrong econ­omy”—the kind and amount of con­tent that gets pro­duced?

Rough Thoughts

Here are some quick thoughts I wrote down. I pub­lish this ques­tion de­spite them be­ing un­finished, in­stead of let­ting them wither deep in my Google Drive.

Un­der the cur­rent sys­tem…

  • Over time, it’s not clear whether karma is in­fla­tion­ary or defla­tion­ary. It de­pends at least on whether the rate of growth of con­tent is slower or faster than the rate of in­crease of karma pro­duc­tion.

  • The only way to get a large amount of karma is to pro­duce con­tent that ap­peals to many users, or pro­duce a large amount of con­tent that ap­peals to at least some users. One can­not get high karma by pro­duc­ing a small amount of con­tent that a small num­ber of users likes a lot.

    • If the real econ­omy was like this, there wouldn’t ex­ist busi­nesses like SpaceX, Palan­tir or Boe­ing.

      • Some­thing seems very bro­ken about LW, if, were the big world to run on LW prin­ci­ples, peo­ple wouldn’t be able to fly as a means of travel. Lots of peo­ple want to fly. But very few are able to pay for the con­struc­tion of a 747. So we only have air­travel be­cause there can ex­ist in­ter­me­di­aries who can make that pay­ment, and in turn get re­warded by col­lect­ing all the lit­tle flight de­sires of very many peo­ple kind-of-keen to fly.

      • Cur­rently, there can­not be any such in­ter­me­di­aries on LessWrong. A con­crete ex­am­ple of a LessWrong Boe­ing might be some­thing like: CFAR re­ally wants some­one to write a 40-page liter­a­ture re­view of X. No one else re­ally cares, apart from the fact that were CFAR to get that re­view, their work­shops would im­prove pretty sig­nifi­cantly for most at­ten­dees.

  • There are fewer free-rider prob­lems. De­spite con­tent be­ing non-ex­clud­able and pub­li­cly available, users have an in­cen­tive to up­vote things, be­cause Alice do­ing so in­stead of Bob does not cost Alice any­thing (we’re as­sum­ing they both end up con­sum­ing the con­tent, so at­ten­tion and time costs are the same).

    • This seems very im­por­tant, and like some­thing that could offset the “Boe­ing prob­lem” men­tioned above.

If in­stead of the cur­rent sys­tem each karma point given was taken from your own score, then…

  • One could not in­definitely keep up/​down-vot­ing con­tent with­out pro­duc­ing new con­tent one­self. In prac­tice, one could do this if one had cre­ated one bea­con of amaz­ing work in the past.

  • Over time, as the same amount of karma gets spread across more and more con­tent, the value of a karma point in­creases (be­cause the op­por­tu­nity cost of what else that karma point could have been used for in­creases).

  • There might be even more defla­tion­ary pres­sure on karma if users pro­duce great con­tent but then leave the site.

  • There is a dis­con­nect be­tween con­tent karma and user karma. A user who has pro­duced much high-qual­ity con­tent might not have a cor­re­spond­ing amount of karma, hav­ing given it away.

Some uncertainties

  • A salient im­ple­men­ta­tion is that an up­vote costs ex­actly the amount of karma that’s be­ing awarded to the con­tent. But how much karma should down­votes cost?

  • It is un­clear how a limited karma sup­ply in­ter­faces with a limited max­i­mal up­vote size

  • There might be les­sons from macroe­co­nomics and mon­e­tary policy rele­vant to this. I don’t know, be­cause I know some­thing-that-rounds-to-noth­ing about those fields.