Connection, Content, Action! – A View on the Internet

I was struck by Alexis Madrigal’s description of how the Internet functions – a human phenomenon, recreated every day, mediated by material machines and generative algorithms. It’s an evocative image, in line with anthropological work on smaller communities online (say, the ethnography My Life as a Night Elf Priest: An Anthropological Account of World of Warcraft, on World of Warcraft). But here Madrigal blows that up into a much larger view:

For just about every person, the Internet is not content brands that they return to mindlessly day after day. The Internet experience is composed of people (friends, famous people, Internet famous people, high school frenemies) and individual things (stories, items of clothing, pictures). These components get rearranged anew every single day into the idiosyncratic Internet that one knows as one’s own.

And because Google is built by ingesting human intelligence, the way its search work reflects those priorities. MacArthur wants the Internet to be a directory of brand names, but that’s not how it developed. And if you remember the hand-edited Internet directory of coherent, complete websites that Yahoo once was, you know why: It was impossible to find anything! For human and technical reasons, the fundamental unit that makes sense is not (the site) but (the page). Anyone who has used the Internet knows this, but MacArthur can’t admit that because it would mean agreeing that Google indexing pages is a good thing.

One last thought. Nowadays, most people see several versions of the hand-edited Internet: one is the stream of content their friends share, two is Wikipedia, and three is the way Google recommends search terms in real-time. Your Internet is increasingly shaped by other people’s judgment processed through machines’ ranking algorithms. With Facebook Graph Search, and Google’s Search Plus Your World, this trend is picking up steam.

I might rework Madrigal’s line that “Google is built by ingesting human intelligence.” Perhaps human activity is a better concept than “intelligence.” Just think I Can Has Cheezburger? – LOLCats, rather than The Atlantic or Harpers.

In this view, Madrigal’s analysis comes in lines with what Bonni Nardi’s analysis of World of Warcraft. Here is Rex on Savage Minds describing Nardi’s approach:

Nardi has a long background in studying how people interact with technology. If I understand this correctly, people originally studied usability: how people interacted with computers and how you could change computers to make them more usable. Then they realized that what people wanted to use technology for was affected by the form that technology itself took. Nardi was one of the people who took this insight and developed ‘activity theory’, a generalized approach which made action rather than the actors the center of its approach.

People’s actions, and the mediating content and technology, get rearranged anew every single day on the Internet. It’s an intriguing view, one that highlights its impermanence and its generativity.

Related Posts Plugin for WordPress, Blogger...
This entry was posted in Society, Technology. Bookmark the permalink.

One Response to Connection, Content, Action! – A View on the Internet

  1. Alexander Harvey says:

    If your interface to the is not to your liking, either you don’t matter to the internet yet, it hasn’t got around to you yet, or you are ignorant of the ways of the world. I do not mean this harshly.

    To begin somewhere, I will begin with AltaVista, and a much smaller virtual world.

    I would say that a searching strategy was a key input one could give to your experience. Knowing how information content, most simply the words, tend to congregate was an advantage. If you could infer some combination of words that would be likely to produce a relevant “whack”, a single hit, or perhaps a few hits, that could be to your advantage, particularly given that each search page could be slow to load, and digest. That process of inference could be a little bizzare and geeky, one could I recall use boolean algebra with ANDs and NOTs and parentheses! My point is that it favoured those that could make useful inferences about how web pages were constituted. There was also a level of confidence that given the correct instructions the engine would simply crank out a result, its preception of you was not a search parameter.

    Using natural language was, as I recall, not at that time the way to go, adding “how” do a search string could imply your desire to see the word “how” in the result.

    The Adding of some interpretation of natural language to the search engines, was a benefit but hardly the revolution that parameterising its perception of the searcher has been.

    Recording that I might be excessively interested in say all things to do with “dogs”, could lead to consequences should I search for say (types of bark) unless I have previously shared my prevailing interest in “trees”. One can get retaliation in early (types of tree bark) or the more geeky (types of bark -dog).

    The inversion of searching, matching the searcher to the content, has an inevitability to it. Indexing the inner pages of the populace (our wants) being equally or more important than the pages of the web.

    I think we are looking at two asymmetric systems of inference, cascading over each other. What I might infer about the engine, what it might infer about me, what I should infer as to its likely inferences about me, and so on. We can adopt strategies.

    My name is attached to this missive, information about me and my preoccupations is available, knowing who I am and what I am like is a valuable resource either being tapped or waiting to be tapped, an invitation to be indexed.

    If I matter sufficiently I will be learnt and my search wishes fulfilled no matter how I express them. There is web content right now that is just itching to search me out. If I don’t like that I will stategise, but in the end my powers of inference are quite limited.


Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>