Yesterday Brendan proposed a new idea which is to have a fork of Leaflet that would only be used to comment on Wikipedia articles. This is a really great idea because Wikipedia is one of few internet based public platforms that works towards informing the public, with the help of public input. The goal of this Leaflet fork should be to increase the quantity and quality of public input about Wikipedia articles.

In order to get quality input, you would need to banish the "anything goes" form of online posting and ask people to follow policies and guidelines similar to the ones Wikipedia created. Any post that isn't helpful, constructive or informative might be hidden or deleted.

The scale of this type of internet moderation isn't really possible without crowdsourcing some of the work of moderation, which is why Wikipedia crowdsources so much of the work of editing its articles. Since this is just for commentary on Wikipedia articles and not the actual editing of the articles, you don't need the super complex back-end system Wikipedia uses for version control and editing. If you want the commentary to carry real world weight though, it would be necessary to allow for voting up or down on the article's comments. If one article generates 8000 comments, then you could sort and filter the comments by which received the most support.

One thing you could do with this kind of voting system is use the voting data to find out who is really an expert on what topics, and what their level of expertise is. You could create leaderboards for people who received the most support for their ideas in certain topic areas.

Another thing to consider regarding the voting system is its level of complexity. I don't think a simple thumbs up or down like on Reddit is the best way to do it. This could take some experimentation, but maybe there are ways for readers to give more nuanced feedback to nuanced ideas with just one to three clicks on some feedback buttons. Maybe the simplest way to get slightly more nuanced feedback is to ask the reader at the bottom of the comment, "To what extent do you agree with this comment?" and let them use a slider to choose on a -5 to +5 scale where 0 is unsure and +5 is fully agree. Maybe this specific method wouldn't work out in practice very well, I am just imagining one of many possible innovative ways to capture feedback beyond a simple "like".

Improving the commentary on Wikipedia articles could in the long run improve the quality of the Wikipedia articles themselves. This is a goal worth working towards.