The Nature Of Relationships
The Nature Of Relationships
Some very good thoughts here, well expressed.
Originally shared by Julian Bond
Some thoughts on ways of categorising Social Networks along different dimensions.
Network effects
There is power in size. The question is, how much power? Metcalfe looked at it via the number of 1-1 connections possible and decided it was proportional to the square of the total number of nodes. This is one reason why Facebook or Twitter have more potential network power than G+. So the utility rise not just with N but N^2
Reed took this a stage further and looked at group forming and the total number of potential groups. The total number of possible groups is 2N - N - 1 suggesting that Utility rises in group forming networks at 2^N. However in real world networks nowhere near all possible groups actually form. In Social Groups it seems to me you need a minimum of 5 noisy people for the group to be self-sustaining. On the usual basis that 90% lurk, 9% Comment and 1% post, that means the real lower end is more like 50 people. At the upper end, Dunbar's number steps in as groups become too big for everyone to know each other. Somewhere around 150-250 people, groups tend to fragment. So as well as Social Networks gaining utility with N^2, if they support groups, their utility rises at somewhere between N^2 and 2^N.
So for a Social Network to have long term utility we need lots of people able to form 1-1 relationships. And the ability to form small (5-250), long lived groups around common interests. The open question right now is whether we can build another truly large scale system (tens of millions of users, tens of thousands of groups) on top of a federated architecture or if this requires Google/Twitter/Facebook scale centralisation.
https://en.wikipedia.org/wiki/Metcalfe%27s_law
https://en.wikipedia.org/wiki/Reed%27s_law
One-Few-Many
A long time ago now, I had the idea of categorising communication systems and platforms according to the numbers of creators and the number of consumers. Or authors and readers. Break each side into "One", "Few", "Many" and you have a 3*3 matrix of One-Few-Many authors communicating with One-Few-Many readers. So Private Blogs are typically One-to-Few. Big Media broadcast outlets are One-To-Many. Discussion forums are Few-To-Few. a person's experience of Feed aggregators is Few-To-One. Private messaging is generally One-to-One. eBay is Many-to-Many. Filling out all the cells in the matrix is an interesting intellectual exercise. The point though is that there are fundamental differences here between the various styles of comms.
https://www.voidstar.com/node.php?id=2977
https://www.voidstar.com/images/onefewmany.gif
Sync-Async
Or Sync-real time, nearly real time, Async. How is the style of our comms defined by time? Chat and group chat (like IRC) happen almost in real time. People sometimes try and use Twitter or Email like this but it really doesn't work that way. We then bolt notification systems on top of this to try and speed up the responses and encourage people to respond immediately. But of course this breaks people's natural workflow. It also encourages a lack of thought in the reply. Often what's actually needed is a conscious decision to make the comms Async. Don't expect a reply in 24 hours and the reply you get is likely to be more thoughtful.
Media
Text, voice & image, video. There's a real problem with communications that use a tool that is too high up the stack here. This also relates back to the Sync-Async problem. Don't interrupt me with a voice message when I could skim read the text in my own time. But also don't use a screen shot just because of platform limitations (Twitter!). And don't do a 5 minute video to camera when 300 words would do that I could skim read in 20 seconds. But each media type as it's place. Your text review is no substitute for listening to the music or watching the film. But don't force me to watch the 45 minute TED talk to grok your 5 page idea. TL;DR. is bad, but TL;DV. is worse.
Post Size
Micro, Small, Long. There are fundamental differences in style between 140 chars, one screen of text and a 5 page essay. Every time we impose a limit the users will try and find ways round it. Hence Unroll, and reply-to-self /N conventions. Ideally, any one platform should support all these and not differentiate. In practice, dropping a 1000 word block into a chat doesn't work. But that's actually obvious. What isn't obvious but should be is that long form text is important. And somewhere in social networks there should be support for this. A long form post with semi-threaded short form comments below it is a highly successful pattern so should be supported.
Some very good thoughts here, well expressed.
Originally shared by Julian Bond
Some thoughts on ways of categorising Social Networks along different dimensions.
Network effects
There is power in size. The question is, how much power? Metcalfe looked at it via the number of 1-1 connections possible and decided it was proportional to the square of the total number of nodes. This is one reason why Facebook or Twitter have more potential network power than G+. So the utility rise not just with N but N^2
Reed took this a stage further and looked at group forming and the total number of potential groups. The total number of possible groups is 2N - N - 1 suggesting that Utility rises in group forming networks at 2^N. However in real world networks nowhere near all possible groups actually form. In Social Groups it seems to me you need a minimum of 5 noisy people for the group to be self-sustaining. On the usual basis that 90% lurk, 9% Comment and 1% post, that means the real lower end is more like 50 people. At the upper end, Dunbar's number steps in as groups become too big for everyone to know each other. Somewhere around 150-250 people, groups tend to fragment. So as well as Social Networks gaining utility with N^2, if they support groups, their utility rises at somewhere between N^2 and 2^N.
So for a Social Network to have long term utility we need lots of people able to form 1-1 relationships. And the ability to form small (5-250), long lived groups around common interests. The open question right now is whether we can build another truly large scale system (tens of millions of users, tens of thousands of groups) on top of a federated architecture or if this requires Google/Twitter/Facebook scale centralisation.
https://en.wikipedia.org/wiki/Metcalfe%27s_law
https://en.wikipedia.org/wiki/Reed%27s_law
One-Few-Many
A long time ago now, I had the idea of categorising communication systems and platforms according to the numbers of creators and the number of consumers. Or authors and readers. Break each side into "One", "Few", "Many" and you have a 3*3 matrix of One-Few-Many authors communicating with One-Few-Many readers. So Private Blogs are typically One-to-Few. Big Media broadcast outlets are One-To-Many. Discussion forums are Few-To-Few. a person's experience of Feed aggregators is Few-To-One. Private messaging is generally One-to-One. eBay is Many-to-Many. Filling out all the cells in the matrix is an interesting intellectual exercise. The point though is that there are fundamental differences here between the various styles of comms.
https://www.voidstar.com/node.php?id=2977
https://www.voidstar.com/images/onefewmany.gif
Sync-Async
Or Sync-real time, nearly real time, Async. How is the style of our comms defined by time? Chat and group chat (like IRC) happen almost in real time. People sometimes try and use Twitter or Email like this but it really doesn't work that way. We then bolt notification systems on top of this to try and speed up the responses and encourage people to respond immediately. But of course this breaks people's natural workflow. It also encourages a lack of thought in the reply. Often what's actually needed is a conscious decision to make the comms Async. Don't expect a reply in 24 hours and the reply you get is likely to be more thoughtful.
Media
Text, voice & image, video. There's a real problem with communications that use a tool that is too high up the stack here. This also relates back to the Sync-Async problem. Don't interrupt me with a voice message when I could skim read the text in my own time. But also don't use a screen shot just because of platform limitations (Twitter!). And don't do a 5 minute video to camera when 300 words would do that I could skim read in 20 seconds. But each media type as it's place. Your text review is no substitute for listening to the music or watching the film. But don't force me to watch the 45 minute TED talk to grok your 5 page idea. TL;DR. is bad, but TL;DV. is worse.
Post Size
Micro, Small, Long. There are fundamental differences in style between 140 chars, one screen of text and a 5 page essay. Every time we impose a limit the users will try and find ways round it. Hence Unroll, and reply-to-self /N conventions. Ideally, any one platform should support all these and not differentiate. In practice, dropping a 1000 word block into a chat doesn't work. But that's actually obvious. What isn't obvious but should be is that long form text is important. And somewhere in social networks there should be support for this. A long form post with semi-threaded short form comments below it is a highly successful pattern so should be supported.
Comments
Post a Comment