Conditional Spam (Something we could use a better word for but this will do for now)
In short: Conditional Spam is information that is valuable to 99 percent of people.
Huge proportions of the content generated and shared on the internet is in this category, and this becomes more and more the case as a greater percentage of the population outputs to the internet as well as reading it. In this category are things like people’s photos of their cats, stories about day to day anecdotes, baby pictures, but ALSO, and importantly, things like most scientific studies, news articles, and political arguments. People criticize twitter for encouraging seemingly narcissistic pointless microblogging, but in reality it’s the perfect engine for distributing conditional spam: Anyone who cares about your dog can follow you, and anyone who doesn’t can NOT.
When your twitter or facebook or RSS is full of things you’re not informed (or entertained, since this applies to fun as well as usefulness) by, this isn’t a failing of the internet. It’s a failing of your filter. The internet is a tool optimized to distribute conditional spam as widely as possible, and you can tune your use of it to try and make the less than 1 percent of it you’ll inevitably see something you WANT to see, and your less than 1 percent of it you’ll MAKE go to the people who actually care about it.
I don’t like the phrase conditional spam both because it’s not CREATED with sinister motives and it also presents it as a bad thing. 99 percent of all things are not for YOU but that doesn’t mean it’s not good that they’re created. I think coming up with good terminology for this can also help us start to create actual mechanisms by which to optimize it. You can sort of shortcut the information filter process by paying attention to only people who pay attention to similar things as you, but is there an efficient way to set up eg, a news source that only gives you news you are likely to be interested in reading? It might be tuneable by tracking how likely you are to finish reading the articles.
It would be nice to have some way of adding tags to the information, so that we could specify what information we need, and avoid the rest. Unfortunately, this would not work, because the tagging would be costly, and there would be incentives to tag incorrectly.
For example, I like to be connected with people I like on Facebook. I just don’t have to be informed about every time they farted. So I would prefer if some information would be labeled as “important” for the given person, and I would only read those. But that would only give me many links to youtube videos labeled “important”; and even this assumes too optimistically that people would bother to use the tags.
I missed my high-school reunion once because a Facebook group started specifically to notify people about the reunion gradually became a place for idle chat. After a few months of stupid content I learned to ignore the group. And then I missed a short message which was exceptionally on-topic. There was nothing to make it stand out of the rest.
In groups related to one specific goal, a solution could be to mark some messages as “important” and to make the importance a scarce resource. Something like: you can only label one message in a week as important. But even this would be subject to games, such as “this message is obviously important, so someone else is guaranteed to spend their point on it, so I will keep my point for something else”.
The proper solution would probably use some personal recommendation system. Such as: there is an information, users can add their labels, and you can decide to “follow” some users which means that you will see what they labeled. Maybe something like Digg, but you would see only the points that your friends gave to the articles. You could have different groups of friends for different filters.
Conditional Spam (Something we could use a better word for but this will do for now)
In short: Conditional Spam is information that is valuable to 99 percent of people.
Huge proportions of the content generated and shared on the internet is in this category, and this becomes more and more the case as a greater percentage of the population outputs to the internet as well as reading it. In this category are things like people’s photos of their cats, stories about day to day anecdotes, baby pictures, but ALSO, and importantly, things like most scientific studies, news articles, and political arguments. People criticize twitter for encouraging seemingly narcissistic pointless microblogging, but in reality it’s the perfect engine for distributing conditional spam: Anyone who cares about your dog can follow you, and anyone who doesn’t can NOT.
When your twitter or facebook or RSS is full of things you’re not informed (or entertained, since this applies to fun as well as usefulness) by, this isn’t a failing of the internet. It’s a failing of your filter. The internet is a tool optimized to distribute conditional spam as widely as possible, and you can tune your use of it to try and make the less than 1 percent of it you’ll inevitably see something you WANT to see, and your less than 1 percent of it you’ll MAKE go to the people who actually care about it.
I don’t like the phrase conditional spam both because it’s not CREATED with sinister motives and it also presents it as a bad thing. 99 percent of all things are not for YOU but that doesn’t mean it’s not good that they’re created. I think coming up with good terminology for this can also help us start to create actual mechanisms by which to optimize it. You can sort of shortcut the information filter process by paying attention to only people who pay attention to similar things as you, but is there an efficient way to set up eg, a news source that only gives you news you are likely to be interested in reading? It might be tuneable by tracking how likely you are to finish reading the articles.
It would be nice to have some way of adding tags to the information, so that we could specify what information we need, and avoid the rest. Unfortunately, this would not work, because the tagging would be costly, and there would be incentives to tag incorrectly.
For example, I like to be connected with people I like on Facebook. I just don’t have to be informed about every time they farted. So I would prefer if some information would be labeled as “important” for the given person, and I would only read those. But that would only give me many links to youtube videos labeled “important”; and even this assumes too optimistically that people would bother to use the tags.
I missed my high-school reunion once because a Facebook group started specifically to notify people about the reunion gradually became a place for idle chat. After a few months of stupid content I learned to ignore the group. And then I missed a short message which was exceptionally on-topic. There was nothing to make it stand out of the rest.
In groups related to one specific goal, a solution could be to mark some messages as “important” and to make the importance a scarce resource. Something like: you can only label one message in a week as important. But even this would be subject to games, such as “this message is obviously important, so someone else is guaranteed to spend their point on it, so I will keep my point for something else”.
The proper solution would probably use some personal recommendation system. Such as: there is an information, users can add their labels, and you can decide to “follow” some users which means that you will see what they labeled. Maybe something like Digg, but you would see only the points that your friends gave to the articles. You could have different groups of friends for different filters.
SEO is the devil.