Revitalising Less Wrong is not a lost purpose

ohn_Maxwell_IV ar­gued that re­vi­tal­is­ing Less Wrong is a lost pur­pose. I’m also very skep­ti­cal about Less Wrong 2.0 - but I wouldn’t agree with it be­ing a lost pur­pose. It is just that we are cur­rently not on a track to any­where. The #LW_code_ren­o­va­tion chan­nel re­sulted in a cou­ple of minor code changes, but there hasn’t been any dis­cus­sion for at least a month. All that this means, how­ever, is that if we want a bet­ter less wrong that we have to do some­thing other than what we have been do­ing so far. Here are some sug­ges­tions.

Sys­tem­atic changes, not con­tent production

The key prob­lem cur­rently is the lack of con­tent, so the most im­me­di­ate solu­tion is to pro­duce more con­tent. How­ever, not many peo­ple are an Eliz­ier or a Scott. Think about what per­centage of blog are ac­tu­ally suc­cess­ful—now throw on the ex­tra limi­ta­tion of hav­ing to be on topic on Less Wrong. Note that many of Scott’s most pop­u­lar posts would be too poli­ti­cal to be posted on Less Wrong. Try­ing to get a group of peo­ple to­gether to post con­tent on Less Wrong wouldn’t work. Let’s say 10 peo­ple agreed to join such a group. 5 would end up do­ing noth­ing, 3 would do 2-3 posts and it’d fall on the last 2 to drive the site. The odds would be strongly against them. Most peo­ple can’t con­sis­tently pump out high qual­ity con­tent.

The plan to get peo­ple to re­turn to Less Wrong and post here won’t work ei­ther un­less com­bined with changes. Pre­sum­ably, peo­ple have moved to their own blogs for a rea­son. Why would they come back to post­ing on Less Wrong, un­less some­thing was changed? We might be able to con­vince some peo­ple to make a few posts here, but we aren’t go­ing to re­turn the com­mu­nity to its glory days with­out con­sis­tent con­tent.

Why not try to change how the sys­tem is set up in­stead to en­courage more con­tent?

De­cide on a direction

We now have a huge list of po­ten­tial changes, but we don’t have a di­rec­tion. Some of those changes would help bring in more con­tent and solve the key is­sue, while other changes wouldn’t. The prob­lem is that there is cur­rently no con­sen­sus on what needs to be done. This makes it so much less likely that any­thing will ac­tu­ally get done, par­tic­u­larly given that it isn’t clear whether a par­tic­u­lar change would be ap­proved or not if some­one did ac­tu­ally do it. At the mo­ment, what we have is peo­ple com­ing on to the site sug­gest­ing fea­tures and there is dis­cus­sion, but there isn’t any­one or any group in charge to say if you im­ple­ment this that we would use it. So peo­ple will of­ten never start these pro­jects.

Be­fore we can even tackle the prob­lem of get­ting things done, we need to tackle the prob­lem of what needs to be done. The cur­rent sys­tem of peo­ple sim­ply mak­ing posts in dis­cus­sion in bro­ken—we never even get to the con­sen­sus stage, let alone im­ple­men­ta­tion. I’m still think­ing about the best way to re­solve this, I think I’ll post more about this in a fu­ture post. Re­gard­less, prac­ti­cally *any* sys­tem, would be bet­ter than what we have now where there is *no* de­ci­sion that is ever made.

Below I’ll sug­gest what I think our di­rec­tion should be:


Less Wrong is the web­site for global move­ment and has a high num­ber of pro­gram­mers, yet some so­cieties in my uni­ver­sity are more ca­pa­ble of get­ting things done than we are. Part of the rea­son is that uni­ver­sity so­cieties have po­si­tions—peo­ple de­cide to run for a po­si­tion and this grants them sta­tus, but also cre­ates re­spon­si­bil­ities. At the mo­ment, we have *no-one* work­ing on adding fea­tures the web­site. We’d ac­tu­ally be bet­ter off if we held an elec­tion for the po­si­tion of web­mas­ter and *only* had that per­son work­ing on the web­site. I’m not say­ing we should re­strict a sin­gle per­son to be­ing able to con­tribute code for our web­site, I’m just say­ing that *right now* im­ple­ment­ing this stupid policy would ac­tu­ally im­prove things. I imag­ine that there would be at least *one* de­cent pro­gram­mer for whom the sta­tus would be worth the work given that half the peo­ple here seem to be pro­gram­mers.


If we want more con­tent, then an easy way would be to have a links sec­tion, be­cause post­ing a link is about 1% of the effort of try­ing to write a Less Wrong post. In or­der to avoid dilut­ing dis­cus­sion, these links would have to be posted in their own sec­tion. Given that this sys­tem is based upon Red­dit, this should be su­per easy.


The other easy way to gen­er­ate more con­tent would be to change the rules about what con­tent is on or off topic. This comes with risks—many peo­ple like the dis­cus­sion sec­tion how it is. How­ever, if a sep­a­rate sec­tion was cre­ated, then peo­ple would be able to have these ad­di­tional dis­cus­sions with­out im­pact­ing how dis­cus­sion works at the mo­ment. Many peo­ple have ar­gued for a tag sys­tem, but whether we sim­ply cre­ate ad­di­tional cat­e­gories or use tags would be mostly ir­rele­vant. If we have some­one who is will­ing to build this sys­tem, then we can do it, if not, then we should just use an­other cat­e­gory. Given that there is already Main and Dis­cus­sion I can’t imag­ine that it would be that hard to add in an­other cat­e­gory of posts. There have been many, many sug­ges­tions of what cat­e­gories we could have. If we just want to get some­thing done, then the sim­plest thing is to add a sin­gle new cat­e­gory, Open, which has the same rules as the Open Threads that we are already run­ning.

Halve downvotes

John_Maxwell_IV points out that too many posts are get­ting down­votes and crit­i­cal com­ments. We could try to change the cul­ture of Less Wrong, per­haps ask a high sta­tus in­di­vi­d­ual like Scott or Eliz­ier to re­quest peo­ple to be less crit­i­cal. And that might even work for even a week or a month, be­fore peo­ple for­get about it. Or we could just halve down­votes. While not com­pletely triv­ial, this change would be about as sim­ple as they come. We might want to only halve down­votes on ar­ti­cles, not com­ments, be­cause we seem to get enough com­ments already, just not enough con­tent. I don’t think it’ll lower the qual­ity of con­tent too much—quite of­ten there are more peo­ple who would down­vote a post, but they don’t bother be­cause the con­tent is already be­low zero. I think this might be worth a go—I see a high po­ten­tial up­side, but not much in the way of down­side.


If we could de­ter­mine that a par­tic­u­lar set of fea­tures would have a rea­son­able chance of im­prov­ing LessWrong, then we could crowd-source putting a bounty on some­one im­ple­ment­ing these fea­tures. I sus­pect that there are many peo­ple who’d be happy to donate some money and if we chose sim­ple, well defined fea­tures, then it ac­tu­ally wouldn’t be that ex­pen­sive.