Less Wrong is a community blog devoted to refining the art of human rationality. Please visit our About page for more information.

jsalvatier comments on Money: The Unit of Caring - Less Wrong

95 Post author: Eliezer_Yudkowsky 31 March 2009 12:35PM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Comments (126)

You are viewing a single comment's thread. Show more comments above.

Comment author: jsalvatier 22 November 2010 06:38:07AM 1 point [-]

I think charities already do this.

Comment author: wnoise 22 November 2010 07:02:17AM 10 points [-]

Yes, and it's widely regarded as a problem -- for someone with rare skills or knowledge, it is usually far more valuable for them to donate money to buy time from others, rather than to donate their own time. A computer programmer really should not be making and serving soup at a homeless shelter. The same amount of time spent coding could pay for several people capable of doing the same thing.

Wikipedia can directly harness those with rare knowledge, and can do so piecemeal, in five-minute intervals, rather than by taking days at a time as even extremely short employment would require. For them it doesn't make sense to pay someone to write an article on an obscure topic. It does seem to make sense for them to pay for servers and sysadmins.

(It's true that their treatment of experts really could be better. They have managed to drive several experts away because dealing with some of the editors is just not worth the time.)

Are there other areas where it actually makes sense to have volunteered time rather than donating money?

Comment author: David_Gerard 22 November 2010 05:16:59PM *  11 points [-]

The expertise problem is one of Wikipedia's perennial headaches. Whereas most actual experts are happy to talk about their field with interested amateurs and even take their ideas seriously, structurally it's just about impossible to keep an indefinitely renewed supply of idiots out of the experts' faces on Wikipedia.

Just giving experts primacy was tried by Citizendium and failed badly - the token of expertise used was credentials and it turns out that cranks care a lot more about credentials than actual experts do, so the cranks moved in and CZ became known for pseudoscience.

I don't know of other ideas for how to keep idiots out of experts' faces, and I've been around English Wikipedia for seven years now. Any are welcomed, even if we or someone has likely tried them already.

Comment author: Risto_Saarelma 22 November 2010 06:00:00PM 5 points [-]

No good ideas for Wikipedia, but I've been wondering a bit about this in general. If we could somehow get a largish group of people generally agreed to be experts actively using a large-scale online reputation system where they start out marked as such, we could write an algorithm that gives stuff and people get expert cred when they are upvoted by the established experts. Could the algorithm be designed to set up a perpetuating and non-decaying reputation cluster of actual experts, while not being swamped by populist crap despite some of it getting tons of non-expert votes?

Comment author: garaman 03 December 2011 06:01:00PM 4 points [-]

" we could write an algorithm that gives stuff and people get expert cred when they are upvoted by the established experts"

Sounds like StackOverflow (http://stackoverflow.com)

It is a domain-specific community (everything about software development), in the form of questions and answers. Reputation is earned by upvotes from other community members. Naturally, the established experts emerge as high-reputation members, being those with a long history of giving good answers.

Comment author: kpreid 06 December 2011 01:32:12AM 5 points [-]

Stack Overflow is just one (and the original) of a set of sites/communities on the same software, Stack Exchange. Each site has its own reputation scores for users, on the principle that someone who is knowledgeable (and sensible) on a given topic isn't necessarily so on another (though there is a +100 cross-signup bonus, presumably on the “OK, you're not an unknown fresh pseudonym” basis).

Avoiding populism problems: Stack Exchange specifically avoids “general interest” in favor of sites with specific topics, and moderation discourages “populist crap” of the “What are your favorite X” / “How do you feel about X” sort.

The format draws a sharp distinction between questions and answers and meta-discussion thereof (either in the form of “comments” or the full-scale meta-discussion site) in order to increase signal-to-noise.

(In case you haven't noticed, I think they're doing a lot of things right and right things, in terms of creating a valuable resource and community.)

(Should there be a top-level post about SO/SE? I don't know that I could write it.)

Comment author: jhuffman 06 December 2011 08:58:52PM 2 points [-]

StackExchange makes no distinction between up-votes from experts and up-votes from idiots. The way I read Risto's suggestion is that the people get cred only from up-votes from experts. This is why it has to be seeded with experts.

Comment author: Risto_Saarelma 04 December 2011 08:04:04AM 1 point [-]

Stack Overflow and other forums dedicated to a specific topic don't have quite the same populism problem as general interest forums do.

Comment author: homunq 09 December 2011 03:17:31PM 0 points [-]

What about Quora?

Comment author: jsalvatier 22 November 2010 04:41:29PM 4 points [-]

I agree completely.

Possibly in other software projects where the donors consider it a hobby.

Comment author: David_Gerard 22 November 2010 05:11:11PM 1 point [-]

Oh, excellent example. Yes, that's what Wikipedia is analogous to: software projects are charities that need applied expertise more than they need money. A project can run on very little money indeed if it has sufficient dev brilliance. (Though a company that pays said brilliant devs to work on the project for a living is nice.)