This guide introduces newbies to some basic lectures and resources on the attention economy. Far from being comprehensive, this guide focuses on recent, cutting edge contributions to this great conversation, sorted into rough categories for ease of use. I hope that this primer sketches a picture of the social, political, and economic stakes of perhaps the most radical restructuring of social organization that humanity has ever dared to undertake, and of the science that has made it possible.

As the Wikipedia page notes, Herbert Simon first suggested attention management as a method for dealing with information abundance in the 1970’s, as part of his research program into complexity and cybernetic organization. “The Attention Economy” secured its place in mainstream business and marketing jargon after Davenport and Beck’s 2001 book by the same name. Since then, attention management has played a central role in the basic principles of web and game design, and is fundamental to social media management and internet advertising. Overviews have been written to keep people on track, like this 2007 overview from ReadWriteWeb or this 2011 link repository at On the Spiral. Such overviews tend to treat the attention economy (and it is always the attention economy, never an attention economy) as a mix of business strategy and design philosophy.

The complexity sciences have matured a great deal since Simon’s pioneering work. We are in a better position today to model, predict, visualize, and indeed manipulate the dynamics of complex systems than we were even a decade ago. These technological advances come on the heels of incredible progress in mathematics and computer science, a paradigm which has come to be called “Big Data” by the media and has attracted significant government and research interest. This paradigm has broad application, from modeling the dynamics of climate change to building computational models of the brain. Big data is helping to understand the structural properties of proteins, ecosystems, and the cosmic background radiation. Big data is also helping us understand the complex dynamics of human social, political, cultural, and economic relationships. When big data is leveraged against human organizational structure, the result is a model of the attention economy. 

A model of meme diffusion

Attention models describe the dynamics of self-organized human social structures. The above attention models from Weng et al (2012, see below) visualizes a time slice of a meme diffusion network, using the Twitter firehose as source data for modeling the memetic spread of socially important hashtags during 2011. Each node in these networks represents a single Twitter user; each connection represents a retweeted meme containing the corresponding hash tag. These models are suggestive and compelling even to a novice observer: think about what the two lobes in the #GOP network represent, and why such graphs mean for the state of the political discourse. Such models not only provide quantitative measures of the self-organized networks of participants in these conversations, but serves as a basis for anticipating their future organizing behavior. 


A particularly stark demonstration of this power occurred during the 2012 American Idol finale. Ciulla et al published a model to track the mentions of the contestants in the run up to the final vote. Using a very simple analysis, these scientists were able to predict the outcome of the American Idol final vote with a high degree of confidence. The paper concludes its demonstration with the following humbling result:

“On a more general basis, our results highlight that the aggregate preferences and behaviors of large numbers of people can nowadays be observed in real time, or even forecasted, through open source data freely available in the web. The task of keeping them private, even for a short time, has therefore become extremely hard (if not impossible), and this trend is likely to become more and more evident in the future years.”

These results suggest that the organizational structure of human populations can be anticipated and even predicted in advance, without knowing specifics about the individual human agents in the networks. Such predictions in turn give rise to techniques for shaping, manipulating, and controlling those networks. Above all, such predictive models will feature prominently in any systematic attempt at planning for the future of the species and the planet. Given the dire future we anticipate for ourselves, and the incredibly poor progress that our existing institutions have made towards resolving these problems, any alternative would be welcome.

The attention economy is not merely an organizational alternative. It is a framework for representing the hopes, concerns, and interests of world’s digital communities, with all their creativity, tireless collaboration, and selfless contributions. Internet communities will eventually use attention models to anticipate the distribution of resources and division of labor with far more stability and humanitarian care than any financial market or business plan could ever hope to accomplish. Perhaps more importantly, attention models will also let us measure with great precision the consensus of the people, with far more accuracy and confidence than any vote could ever muster. Taken together, attention models provide the essential feedback loops for allowing human communities to self-organize at a global scale beyond the limitations and failures of government and financial institutions. The problems we face systematic institutional changes, and we are rising to the challenge. 

Internet is our only hope

The resources, lectures, and articles below will help introduce you to the progress we’ve made, and hopefully will give you some insight into what comes next.

Competition among memes in a world with limited attention (Weng et al, 2012)

We propose an agent-based model to study the role of the limited attention of individual users in the diffusion process, and in particular whether competition for our finite attention may affect meme popularity, diversity, and lifetime. Although competition among ideas has been implicitly assumed as a factor behind, e.g., the decay in interest toward news and movies, to the best of our knowledge nobody has attempted to explicitly model the mechanisms of competition and how they shape the spread of  information. In particular, we show that a simple model of competition on a social network, without any further assumptions about meme merit, user interests, or explicit exogenous factors, can account for the massive heterogeneity in meme popularity and persistence.

See also:

• Simon (1962), “The Architecture of Complexity“
• Bar-Yam (2004), “A Mathematical Theory of Strong Emergence using Multiscale Variety“
• RSAnimate (2012), The Power of Networks
• Bruno Gonçalves’s G+ stream (actively posts on recent developments in complexity science)
• Curated list of online talks related to complexity science (Scoop.it)

The Dissemination of Culture (Axelrod, 1997)

The methodology of the present study is based on three principles: 

l. Agent-based modeling. Mechanisms of change are specified for local actors, and then the consequences of these mechanisms are examined to discover the emergent properties of the system when many actors interact? Computer simulation is especially helpful for this bottom-up approach, but its use predates the availability of personal computers (e.g., Schelling 1978). 

2. No central authority. Consistent with the agent-based approach is the lack of any central coordinating agent in the model. lt is certainly true that important aspects of cultures sometimes come to be standardized, canonized, and disseminated by powerful authorities such as church fathers, Webster, and Napoleon. The present model, however, deals with the process of social influence before (or alongside of) the actions of such authorities. It seeks to understand just how much of cultural emergence and stability can be explained without resorting to the coordinating influence of centralized authority. 

3. Adaptive rather than rational agents. The individuals are assumed to follow simple rules about giving and receiving influence, These rules are not necessarily derivable from any principles of rational calculation based on co sts and benefits or forward­looking strategic analysis typical of game theory. Instead, the agents simply adapt to their environment.

Supplimental Java applet

See also:

• Axelrod (1981), “The Evolution of Cooperation“
• ——, Online Guide to Newcomers on Agent-Based Modeling in the Social Sciences
• West (2011) Why Cities Keep Growing
• Strogatz (2011) “Social Networks that Balance Themselves”
• Kohler (2012) 10 Lessons from 20 Years in Building a Complexity-Informed Archaeology

Complexity and the Economy  (Arthur, 1999)

Conventional economic theory chooses not to study the unfolding of the patterns its agents create but rather to simplify its questions  in  order to  seek analytical solutions. Thus it  asks what behavioral elements (actions, strategies, and expectations) are consistent with the aggregate patterns these behavioral elements co-create? For example, general equilibrium theory asks what prices and quantities of goods produced and consumed are consistent with (would pose no incentives for change to) the overall pattern of  prices and  quantities in  the  economy’s  markets.

Game theory asks what moves or choices or allocations are consistent with (are optimal given)  other agents’  moves  or  choices  or allocations in  a strategic situation. Rational expectations economics asks what forecasts (or expectations) are consistent with (are on average validated by)  the  outcomes these forecasts and  expectations together create. Conventional economics thus studies consistent patterns: patterns in behavioral equilibrium  that would  induce no  further reaction. Economists at the  Santa Fe Institute, Stanford, MIT, Chicago, and other institutions are now broadening this equilibrium approach by turning to the question of how actions, strategies, or expectations might react in general to (might endogenously change with) the aggregate patterns these create.  The result  complexity economics  is not an adjunct to standard economic theory but theory at a more general, out-of-equilibrium level.

See also:

• Taking stock of Complexity Economics (2012 video lecture series, various lecturers) 
• Kirman, (2012) Participatory computing and its implications for the economy

Cognitive Democracy (Shalizi, 2012)

We start by explaining social institutions should do. Next, we examine sophisticated arguments that have been made in defense of markets (Hayek’s theories about catallaxy) and hierarchy (Richard Thaler and Cass Sunstein’s “libertarian paternalism”) and discuss their inadequacies. The subsequent section lays out our arguments in favor of democracy, illustrating how democratic procedures have cognitive benefits that other social forms do not. The penultimate section discusses how democracy can learn from new forms of collective consensus formation on the Internet, treating these forms not as ideals to be approximated, but as imperfect experiments, whose successes and failures can teach us about the conditions for better decision making; this is part of a broader agenda for cross-disciplinary research involving computer scientists and democratic theorists.

See also:

• 2012 Personal Democracy Forum (various video lectures)
• Estrada (2012) Digital Politics
• Estrada (2012) Systems of Organization

The Power of Fear in Networked Publics (Boyd, 2012)

And here’s where we see fear entering back into the picture.  Because fear is a biological mechanism to get people’s attention, we see people turning to fear as a tool to get people’s attention.  Fear is an extraordinarily effective emotion to leverage.  Fear is especially powerful in an environment where the available attention is limited. If you pay attention to threatening stimuli, fear emerges.  At the same time, the presence of fear gets your attention.  The two – fear and attention – work hand in hand.  This is why the attention economy provides fertile soil for the culture of fear.

We pay attention to the emotion of fear because it helps us protect ourselves and those around us.  Our willingness to pay attention to fearful stimuli is precisely why it’s possible to create a culture of fear.  We are far too willing to consume information that makes us afraid because we feel as though we want that information in order to protect ourselves. Fear-mongerers leverage our willingness to pay attention to fearful stimuli in order to generate attention.  A fearful newspaper headline captures people’s attention.  This draws people into paying attention to the newspaper as a whole, which is precisely the intention of headlines.  Likewise, when TV anchors are spouting off fearful information, people are far less willing to turn the channel.  Again, this is of interest to the television network. With social media, the intersection is messier.  There are certainly broadcast messages being communicated from far off, but the majority of attention-seeking takes place in the world of user-generated content.  This creates an ecosystem where hysteria isn’t necessarily from on high, but, rather, all around us. 

See also:

• Danah Boyd at Webstock 2012 (video)
• Clay Shirky (2010), Cognitive Surplus
•  —–,  “Its not information overload, its filter failure”
• Leavitt (2012), “The ethics of attention”
• Huberman (2012) “Social Media and Attention“
• Culture Machine (2012) Paying Attention  (full issue of dedicated articles)

When they had docked at Mljet in their slow-boat refugee barges, they’d been given their spex and their ID tags. As proper high-tech pioneers, they soon found themselves humbly chopping the weeds in the bold Adriatic sun.

The women did this because of the architecture of participation. They worked like furies.

As the camp women scoured the hills, their spex on their kerchiefed heads, their tools in their newly blistered hands, the spex recorded whatever they saw, and exactly how they went about their work. Their labor was direct and simple: basically, they were gardening. Middle-aged women had always tended to excel at gardening.

The sensorweb identified and labeled every plant the women saw through their spex. So, day by day, and weed by weed, these women were learning botany. The system coaxed them, flashing imagery on the insides of their spex. Anyone who wore camp spex and paid close attention would become an expert.

The world before their eyeballs brimmed over with helpful tags and hot spots and footnotes.

As the women labored, glory mounted over their heads. The camp users who learned fastest and worked hardest achieved the most glory. “Glory” was the primary Acquis virtue.

Glory never seemed like a compelling reason to work hard-not when you simply heard about the concept. But when you saw glory, with your own two eyes, the invisible world made so visible, glory every day, glory a fact as inescapable as sunlight, glory as a glow that grew and waned and loomed in front of your face-then you understood.

Glory was the source of communion. Glory was the spirit of the corps. Glory was a reason to be.

Camp people badly needed reasons to be. Before being rescued by the Acquis, they’d been desolated. These city women, like many city women, had no children and no surviving parents. They’d been uprooted by massive disasters, fleeing the dark planetary harvest of droughts, fires, floods, epidemics, failed states, and economic collapse.

These women, blown across the Earth as human flotsam, were becoming pioneers here. They did well at adapting to circumstance-because they were women. Refugee women-women anywhere, any place on Earth-had few illusions about what it meant to be flotsam.

Vera herself had been a camp refugee for a while. She knew very well how that felt and what that meant. The most basic lesson of refugee life was that it felt bad. Refugee life was a bad life.

With friends and options and meaningful work, camp life improved. Then camp life somewhat resembled actual life. With time and more structure and some consequential opportunities, refugee life was an actual life. Whenever strangers became neighbors, whenever they found commonalities, communities arose. Where there were communities, there were reasons to live.

Camp user statistics proved that women were particularly good at founding social networks inside camps. Women made life more real. Men stuck inside camps had a much harder time fending off their despair. Men felt dishonored, deprived of all sense and meaning, when culture collapsed.

Refugee men trapped in camp thought in bitter terms of escape and vengeance. “Fight or flight.” Women in a camp would search for female allies, for any means and methods to manage the day. “Tend and befriend.”

So: In a proper modern camp like this one, the social software was designed to exploit those realities.

First, the women had to be protected from desperate male violence until a community emerged. The women were grouped and trained with hand tools.

The second wave of camp acculturation was designed for the men. It involved danger, difficulty, raw challenge, respect, and honor, in a bitter competition over power tools. It acted on men like a tonic.

Like any other commons-based peer-production method, an Acquis attention camp improved steadily with human usage. Exploiting the spex, the attention camp tracked every tiny movement of the user’s eyeballs. It nudged its everyware between the users and the world they perceived.

Comparing the movements of one user’s eyeballs to the eyeballs of a thousand other users, the system learned individual aptitudes.

A user who was good with an ax would likely be good with a water saw. A user quick to learn about plants could quickly learn about soil chemistry and hydrology. Or toxicity. Or meteorology. Or engineering. Or any set of structured knowledge that the sensorweb flung before the user’s eyes.

The attention camp had already recorded a billion things that had caught the attention of thousands of people. It preserved and displayed the many trails that human beings had cut through its fields of data. The camp was a search engine, a live-in tutoring machine. It was entirely and utterly personal, full of democratically trampled roads to human redemption. By design, it was light, swift, glorious, brilliant.

See also: 

• Sterling (2005), Shaping Things
• ——, Reboot 2009 (video)
• Augmented (hyper)Reality (2010)