Imagine that everyone straps a little box on their foreheads. These little boxes produce tiny invisible marbles at some rate, say: 10 marbles every second. While you are wearing the box, it shoots invisible marbles out at the objects you happen to be looking at. Those objects along with everything else in the environment are equipped with little devices that register and absorb the incoming marbles, so that all your marbles get absorbed by something. These marbles are a crude approximation of the attention you pay. Every time you pay attention to some object, it gets bombarded with the marbles shooting from your forehead.
The idea seems silly because it is. I’d never suggest we actually fling high speed projectiles in arbitrary directions from boxes mounted on people’s foreheads, that would be dangerous and irresponsible. If this is to be implemented at all, it would of course be rendered digitally and transparently as best as our technology will allow. Moreover, the direction a person’s head is facing is a terrible indicator of where their attention is being paid; to do this precisely, we’d need something far more sophisticated. But leave these technical details aside for the moment. This is a toy model, and I’m describing it in some detail to help us think about what the attention economy is doing, and what we are doing in it. So boxes on foreheads with marbles shooting out with some frequency and getting absorbed by other objects. Still with me?
If not, refer back to the meme diffusion model from Weng et al. in Post 1:
The meme diffusion model is a model of the “use” of Twitter. The frequency of use is described by the probabilities, but the character of the user’s behavior itself is “minimal”. There’s lots of things that users do with Twitter-- some of which might matter for the way memes spread-- but the study reduces all those behaviors to the minimal model it describes. That’s good enough for Twitter because the act of retweeting is literally as simple as the push of a button and we can track the propagation of the content over time with the model. But to run an entire social economy we’ll need something more flexible, something that will account for the variety of use behaviors that characterize our daily lives. Even if we just stick with Twitter, the minimal model doesn’t describe anything the user actually does; they might retweet immediately without a second thought or they might concentrate intently, contemplating every last character. The marble network I’m describing gives this interaction more granularity so we catch these kinds of differences. The difference between spending 5 seconds or 15 second reading a tweet might not seem like much, but in the attention economy every quanta of attention counts. The marble network will let us count them.
To be absolutely clear, the attention economy is already online; we are already deeply immersed in the task of allocating our finite attention resources across our various projects. The marble network I’m describing is a model for understanding, simulating, and ultimately managing the economy of attention. Thinking about the flow of marbles in this network will help us discuss how the attention economy works and what we can do with it. Consider once again the specific diagrams being used to represent the behaviors in Twitter. In this network, each user is a node on a graph, and edges represent the connection (“following”) between them. When a user sends a tweet to its followers, think of this as opening an “attention channel” between the user and the follower, connected by the tweet itself as both an object of the follower’s attention and a product of the tweeter’s labor. If our marble network is working properly, opening this attention channel would allow marbles to flow between the two nodes. The flow of attention is a directed graph; in this case, the marbles will flow from the foreheads of every follower, back to the user who first sent the tweet. The followers are “paying attention” to the content producer by looking at the content they produced, and that attention flow is captured by the number of marbles that flow across that channel as long as the follower’s head is pointed at the tweet.
Weng et al. gives us a minimal model of attending behavior in Twitter. A fully realized attention economy would involve models that characterize all other user behaviors, for all other uses of all other objects. Every time I use a pen, or drive on a highway, or take my heart medication, or smell a rose, I am paying attention to the products of the labor of others in many indirect ways. The attention economy accounts for these complex interdependencies that form the material basis of social organization. In exactly this sense, these attention models areeconomic models: they allows us to model the collective production, distribution, and consumption of real-world labor and resources.
So here's the trick, and what makes this flow of attention different from every financial economy we are familiar with: you can't store attention. You can't stockpile attention or reserve a bank of attention units. There is no debt in an attention economy and there can be no surplus of attention. There is just the total amount of attention being produced, and the many ways we allocate that attention among all the things we spend our time doing. Attention must always be paid as it is produced or acquired, and there are no leftovers. So when people pay attention to some node in the network, the recipient of that attention can’t put marbles in a jar for a rainy day. Instead, when you are the target of incoming marbles, your rate of marble production increases. Increasing your rate of marble production is a bit like “compensation” for the attention you’ve attracted, but the metaphor with money will only go so far so let’s stick with marbles. My followers only pay attention to the meme I created for a few seconds, but those brief moments of attention result in at least some marbles flowing across the network that are aimed at me. I can’t put those marbles in my wallet, though, because you can’t store attention; it is as if the marbles flow right through me, increasing the rate of marbles I’m producing from the box on my forehead. Say that, for every 10 people paying attention to my meme, the rate at which I produce marbles increases by one marble a second for some short duration of time; I was producing marbles at 10 marbles/second, and now because of the attention I’ve attracted the box on my forehead begins producing 11 marbles/second. The more attention I get, the faster my rate of marble production.
Producing marbles at a faster rate doesn’t mean I suddenly have more attention to give; nothing about my abilities (cognitive or otherwise) have changed by increasing my marble production. Instead, what changes are the objects I’m attending to. Because I’m producing more marbles, the objects I’m attending to are now receiving more incoming marbles than they were just moments before. My higher rate of marble production means I have slightly more influence on the flow of marbles across the network. I’m paying attention to a lot of objects, and each of those objects are only getting a small fraction of the marbles shooting from the box on my forehead. Still, they each have a chance at getting slightly more marbles now that my own rate of marble production has increased. Since more marbles are flowing my way, slightly more marbles flow down the network through the various objects I use. Of course, all those objects in turn are also “memes” that were produced by other people who are the target of my now-slightly-increased rate of marble production. My “popularity” as a producer has made my use salient against the background of users. So in the attention economy, if your goal is to maximize the marbles headed your way (and if it isn’t obvious already, we will see shortly why this isn’t always the case!) it is “more important” for to see your memes than for to see them, because Lady Gaga is herself the focus of far more attention and is going to have far more ripples in the overall network than me. This is just what it means to have more influence in the network. It’s not just because Lady Gaga has more money (though she has that too), but rather that she occupies the attention of far more people than I do. The marble network not only captures the dynamics of social attention but also the dynamics of social influence. The more marbles I produce, the more influence I have over the flow of marbles throughout the network. And I can change my level of marble production by producing things that attract the attention of others.
For this reason, it will be helpful to treat attention as the inverse of influence, and to treat your role as a node in the attention economy as something like a switch, turning the collective attention you attract into influence at how the network develops in the future. When you "pay" attention to something, you are effectively “trading” that attention for influence. This dynamic should be familiar from even non-technical treatments of the Attention Economy, and even in non-digital contexts. The movie I pay attention to, for at least the time I am paying attention to it, has some degree of influence over my thoughts, experiences, desires, and interests. When I pay attention to my work, I have influence over anyone who benefits or interacts with the products of my labor. I say that attention is “traded” for influence, but this transaction doesn’t look like the kinds of trades that take place in a traditional market, even while they might be just as competitive and lively. I "pay" attention, and though my attention is a finite resource and can only be split so many ways, it is also a resource that I produce constantly, simply in virtue of being conscious; from my perspective it seems like I have an unending supply of attention and I never seem to run out. I can have difficulty paying attention or keeping my attention focused on some particular thing, so managing my attention is no trivial task. Some people are better or worse at it for a variety of complex cognitive and social reasons, and either way distractions abound. Still, paying attention to something is a task that everyone is constantly engaged in, naturally, as a fundamental part of being alive, and each act of attention has material consequences for the production, consumption, and distribution of our collective labor and resources. The model of attention I’m describing connects each of us through the vast network of technological infrastructure that we together have created for ourselves. This network not only models the development of that infrastructure over time, but also allows us to distinguish each node uniquely and directly in proportion to their individual contributions to the collective.
From Part 1, I described the attention economy as a way of “modeling attention behavior in a complex, organized system of attenders”, and hopefully we are getting a better sense of what this means. In the attention economy, you are anattender. We all are. In the attention economy, there is no sense in distinguishing between producers and consumers, because “consumers” themselves contribute to production in the very act of consumption. Some attenders are more productive than others, and those attenders may tend to attract the marbles flowing around the network and attain significant influence in shaping its growth. This marble network is decentralized and competitive in many of the ways that are fiercely defended by proponents of traditional capitalist markets. But regardless of anyone’s productive abilities, and indeed regardless of whether any of us can pay for those uses, we are all participating in the same networked system of use. The attention economy is a way of modeling our collective attending behavior, and it will let us account for the variety of objects we use, consume, produce, and share with each other as we organize ourselves.
The central thesis of this essay, as stated in Part 1, is that “using attention models will increasingly be preferred to other kinds of economic models (especially financial models) as the primary tools for social organization”, and we can now explain what this means in somewhat more detail. Attention models are models of user-behavior; attention models give us information about what people are actually concerned with and how they spend their time. As we collectively confront problems that require social, coordinated action, human societies will increasingly appeal to attention-based models rather than other kinds of models for solving coordination problems. This is not an idealistic prescriptive claim that we should use these models, or a futurist prediction that we will use these models. This is a descriptive, observatuonal claim: We are already doing it, and we do it with more deliberate self-confidence by the day. My argument for this thesis is that the attention economy is a unified model of social organization, and so we therefore have some reason to prefer this model over the existing models. The attention economy unifies what have traditionally been considered the “separate magisteria” of human social organization: the domains of economics, of governance, and of culture, each of which are traditionally assumed to operate by their own internal dynamics. In fact, these domains are deeply interconnected, and an attention economy will allow us to visualize these relations directly. On my view, the attention economy is literally a model of culture (of the networks of practices of individuals) and these models have direct and clear consequences for questions of economics (the distribution of labor and resources) and governance (ensuring a just, consensus-based society). By explicitly understanding the attention economy in this way, we can start to see what a new social organizational structure might look like. In this post I described a simple thought experiment for thinking about where you fit into this complex, interconnected network of attenders. There is a lot to say about the consequences this model has for questions of governance and economics, and hopefully the marble network helps us think about those questions carefully when we deal with them in future posts. But in the next post we’ll need to do a bit more groundwork, to set us up for those discussions. We need to talk about what it means for a collection to be self-organized, and what reasons we have to use attention as a metric for self-organization.
Original Post: https://plus.google.com/u/0/117828903900236363024/posts/Pd5evDVHkGy