Foreword

Well you already read the topic and you probably managed to synthesize some idea what this post might about. But it's probably not what you thought. What you did though, is valid. It might be something more useful than what I'm currently thinking. And what I'm thinking is rather vague. As usual my post might be of a rather low quality but don't let it get to you. If there's clear errors then please do point them out.

 

Introduction

I happened to think about the scoring system for comments and posts on this website. I think it's pretty clever, although I don't really care about points. Perhaps that's just because I have not gotten addicted to the points yet. I might get hooked on the dopamine of achieving yet another level of having even more points.

User feedback

The comment and post structure that uses this user feedback based autonomical content filtering, which is essentially just hiding posts which are below some score threshold, is in my opinion very interesting. How could a system based on a similar dynamic produce something more involving?

 

Community projects

I  thought about.. Is it SDN? I don't know anything about, but there's this protocol of joint coding projects. Maybe you know what it is. Then I thought of wikipedia. Users are modifying the contents of wikipedia and it's a community project.

 

Hyperlinks

Then there's the posts by Eliezer on this website. In a sense I think it's amazing how the posts are structured and connected to eachother through those hyperlinks. The information has coherency and the hyperlinks produce some sort of incremental quality to the substance. I can't quite put my finger on it.

 

Lesswrong observations and things to pay attention to

However despite all these things, there's always "too many" comments. There's always mistakes. There are people correcting those mistakes. And there's misunderstandings. It takes time to read stuff. Our attention is limited by time. I think the point system for separating discussion from the mainposts, I don't know how it exactly works, is based essentially on the idea that first you need to proove that you're making sense, and then you get to post. This threshold then appears in the characteristics of the mainposts. At least I think that's the idea. I think that there's a lot of smart people reading these forums as well as contributing here, and I feel that people could contribute in a more effective way.

 

Some actual ideas

From these points I essentially came to think that there might be away of doing this down and upvoting business as well as these hyperlinks in a manner that would incorporate some qualities from the wikipedia, so that the entire substance of multiple posts that are linked together could be modified by userfeedback which would not only involve the votes but the comments would instead modify the substance in someway.

 

Examples

So for an example stuff that I happened to think about like the following kinds.. Categorizing comments to different types. For an example:

Language and grammar correction: Pointing out grammatical or linguistic errors separately. Anyone can select a portion of your post and correct the language. After doing that the community can vote if the correction is proper. If it is, then post itself is modified, but the original version can still be viewed. And some sort voting criteria should be established for whether or not the grammar correction modifies the substance rather than the spelling and grammar, to protect the substance.

But this grammar issue probably isn't all that interesting. It still functions to demonstrate what kind of potential there might be in systems alike. Perhaps better ones

User feedback quality and probabilities: To continue with the language example you could also rate users based on what they vote and suggest. So if you downvote a correct grammar modification that is used to estimate your error frequency, and vice versa, if you upvote a grammar correction that is really correct, then that is also used to estimate. And so forth. This is essentially the same as user feedback, but it might achieve a finer sensitivity by incorporating user weights.

Substance corrections for facts: Similarly it might be cool point out erroneous facts. If it something very simple like Kennedy was born in 1762 then it's easier to see that you could have the users modify this content in a wikipedia style. But doing this effectively might have some limitations, since the substance people write has a lot more content than just discrete facts.

Hyperlinks as argumentation and logical relationships: So following the correction of facts it might possible to go one step further. It might possible to point to a substance as a claim. Which then could be argued through hyperlinks in a following sense like if A is true then it follows that B because hyperlink. And then you could have users vote if the hyperlink really backs up that B follows from A. I can imagine those little red thumbs down icons or something like that.

 

 

 

Alright I might be able to write more ideas like this or might not, perhaps I could attempt go into more details but I think this should do. My challenge is for you to think how could something like this work. Like I wrote in the beginning, you probably immediately had some sort of an idea of what this post is about when you read the topic. Those words associate to a lot of content and when you put them together there somekind of idea should appear. So I put some of mine into the list of examples. The superorganism is maybe a little off as a term, since I was thinking ogranism in the sense of organization, but still it is about thinking people contributing.

So what do you think?