Arbital: learning from Wikipedia

Wikipe­dia has done many things right, and we can learn from its his­tory to cre­ate a bet­ter plat­form. Let’s take a look at a some of the prob­lems Wikipe­dia has to­day and how Ar­bital is go­ing to re­solve them.

Learning

Wikipe­dia built its tech­nol­ogy and set up its poli­cies to recre­ate the ex­pe­rience of an en­cy­clo­pe­dia, but on­line. An en­cy­clo­pe­dia is great for get­ting ba­sic in­for­ma­tion about a sub­ject, but it’s not well-suited to ex­plain­ing com­plex con­cepts. Wikipe­di­ans have at­tempted to solve this prob­lem by cre­at­ing ad­di­tional pages, e.g. In­tro­duc­tion to gen­eral rel­a­tivity, but this ap­proach does not scale well. Ar­bital’s solu­tion to this is a net­work of req­ui­sites, which al­low the plat­form to dy­nam­i­cally cre­ate se­quences of pages tai­lored to each user, where each page it­self also ad­justs to user’s prefer­ences and knowl­edge.

Granularity

Wikipe­dia doesn’t al­low (and will ac­tively delete) con­tent that they think is not no­table, which ex­cludes many pages which would be use­ful to peo­ple look­ing for de­tailed in­for­ma­tion on a topic. This of­ten leads to the cre­ation of other wiki com­mu­ni­ties, which spe­cial­ize in their par­tic­u­lar area (e.g. World of War­craft wiki, Cam­er­a­pe­dia). On Ar­bital, we want to em­brace spe­cial­ized con­tent and foster com­mu­ni­ties with fo­cused in­ter­ests. The most in­ter­est­ing things in re­search, en­ter­tain­ment, and most other ar­eas are of­ten at the edges, where a small minor­ity (or ~5M peo­ple, in the case of WoW) care about that in­for­ma­tion deeply.

In the case of sci­ence, this prob­lem is even worse. Wikipe­dia also doesn’t al­low origi­nal re­search, but if you are a sci­en­tist, origi­nal re­search is the most in­ter­est­ing part. And all set­tled sci­ence was at some point un­cer­tain. By not al­low­ing that kind of con­tent on its plat­form, Wikipe­dia is fal­ling in­creas­ingly be­hind the rapidly ac­cel­er­at­ing sci­en­tific fron­tier.

Truth vs neutrality

Wikipe­dia’s offi­cial policy is “Ver­ifi­a­bil­ity, not truth”. On Ar­bital, we think truth is pretty great, and we should have more of it. We even have a guid­ing prin­ci­ple for how to find it. Ver­ify­ing based on re­li­able sources works pretty well for figur­ing out the pop­u­la­tion of a coun­try but does hot help with con­tro­ver­sial, un­set­tled top­ics.

Expertise

Wikipe­dia has the no­tion of a re­li­able source, but no method for track­ing ex­per­tise within the sys­tem. This of­ten re­sults in edit wars, which are re­solved by a gen­er­ally well in­ten­tioned and even-handed ad­minis­tra­tor, who nonethe­less has limited do­main ex­per­tise. Ar­bital is plan­ning to solve this prob­lem with an in­no­va­tive karma and do­minion sys­tem, which will al­low proven do­main ex­perts to help set­tle dis­putes. In the mean­time, for do­mains like math where ex­per­tise is rel­a­tively un­con­tro­ver­sial, we have re­view­ers check­ing over new pages and ed­its.

Get­ting answers

When was the last time you tried to re­search some­thing on Wikipe­dia that wasn’t a set­tled fact? Per­haps you were won­der­ing if you should sup­ple­ment your diet with Vi­tamin D. Did you get a satis­fy­ing an­swer? Prob­a­bly not. Wikipe­dia is good at pro­vid­ing the reader with a ton of in­for­ma­tion, but it’s not good at an­swer­ing ques­tions, and it’s even worse at quickly giv­ing an ac­cu­rate overview of a dis­puted topic. Com­pare this to your ex­pe­rience on Stack­Overflow, where it’s much eas­ier to tell what the best an­swer is. Ar­bital has many fea­tures that make find­ing the right an­swer quick and easy: vote bars, likes, com­ments, and in­stant search. Ar­bital also helps you know how much con­fi­dence you should have about the in­for­ma­tion, and al­lows you to dive deeper into dis­cus­sions if you want to un­der­stand the cur­rent state of the topic bet­ter.

com­ment: , and a beta ver­sion of ques­tions and answers

Etc…

You can take a look at all the other things Wikipe­dia is not. Many of these de­ci­sions make sense, but oth­ers re­strict Wikipe­dia from be­com­ing all it could be.

Parents:

  • More about Arbital

    Lots more in­for­ma­tion about Ar­bital vi­sion.

    • Arbital

      Ar­bital is the place for crowd­sourced, in­tu­itive math ex­pla­na­tions.