This is too good to be true. Yesterday I read Sebastiaan’s write-up of how he graphically [made] a link between two individuals who both liked the same thing on the internet, and how, by doing that, he could alert himself to things he might like.
Today I finally see, in my reader, an earlier post from Kicks Condor, in which he talks about surfacing other readers who have linked to things he has linked and how that might help him to discover interesting things to read. That could even be the basis of a self-organising discovery engine.
Clearly, they ought to know about one another. Maybe this post of mine will trigger that.
Cool, yes, the alert worked! That alone is very worthwhile and goes a long way toward discovery. In a way, I think this is the most idealized form—you’ve just done the equivalent of “Hey, check this out” and I am very fortunate that I get to read your reasoning rather than to simply see a like in my box.
I like that Sebastiaan’s end goal is to discover a person and not just CONTENT. To some extent the networks do this: mostly they promote trending squares of blurbs and images, but sometimes you see a note: “Follow these three people.” But you have no idea why and it’s not always based on similarity of our link neighborhoods, but based on geographical closeness or crossing some popularity threshold or your search terms and so on.
I don’t want to be so allergic to social networks that I can’t see the positive tools—bubbling up blurbs and images can be good fun, liking things is effortless nudging—but I think the Indieweb has already improved on this because its protocols are so light that it forces the human connections. (The ‘homebrew website’ clubs are the opposite of viral marketing.) You could see these as counterproductive—but the problem with ‘productive’ protocols is that they become so saturated as to be useless. Google, for instance, is so good that it is useless.
I still think algorithms are tremendously useful, particularly when the hypertexter controls the algo. And Sebastiaan is toying with this. I wonder to what degree his query language could simplified as to be more widely useful. Perhaps there is an Excel-type language that could become the dials for the ‘archivist’/‘librarian’/‘curator’ role.