Clay Shirky  |  James Love  |  Lisa Strausfeld Blur Brainstorming: Experiments in Conference Form Prof. Clay Shirky, NYU/ITP clay@shirky.com The Blur 02 conference, "Power At Play in Digital Art and Culture," was co-organized by Creative Time, The New School and Parsons School of Design, and took place at The New School. The theme of the conference was the shifting power relations we are witnessing between a variety of entities—individuals and organizations, corporations and the state, the for-profit and non-profit sectors—and addressed ways of creating aesthetically or politically engaged work by playing with these changes. I rarely participate in these sorts of conferences as I usually find them a waste of time, but I was particularly interested in Blur because my current area of inquiry is in the way social patterns are occasioned or shaped by the internet. Though my background was in the theater, I long ago left that world first to be the Chief Technical Officer (CTO) of an internet design firm, and then to start writing and teaching on network technologies at the Interactive Telecommunications Program (ITP) at New York University (NYU). My presentation at Blur was on the grossly unequal distribution of popularity in large systems, and the forces that produce that distribution. As interesting as the theme of the conference was to me, I assumed that I would stand out, not being an artist or a theorist of aesthetics. To my surprise, the conference actually had a remarkably heterogeneous attendee list, including several attendees outside the Aesthetic-Academic complex, and the conversations that resulted were fascinating and wide-ranging. This was in large part because the conference designers undertook considerable experimentation with the conference form itself, in an attempt, as the software designers say, to "eat our own dog food." By making the structure of the conference both subject and object, Blur invited the participants to talk not just about the outside world, but also to reconsider how theorists and practitioners communicate with one another, an opportunity most of the participants embraced with relish. One particularly successful piece of experimentation was the second day of the conference, a Saturday, which was structured around a single day-long brainstorming exercise. Many conference attendees reported the Saturday session as the most valuable and important of the conference, and its combination of standard conference design and brainstorming techniques make it a good model for future work. This report principally concerns the design and execution of that brainstorming session, noting the aspects of the structure that made it such a positive experience, as well as detailing some of the pitfalls. It ends with speculation about future work extending these efforts to other arenas, including future versions of Blur. Set-up The set-up for the Saturday session combined several simple strategies. The day began with a plenary session where the conference organizers led the group in listing many of the themes we were interested in, some of which had been implicit and others of which had arisen during the previous day's panels. Some of these were quite general—intellectual property—and others were quite specific—search engine design. After the general list was made, a sub-set of these items that were interesting to a broad cross-section of the assembled group was chosen. The attendees were clustered into half a dozen working groups of 7 or 8 people, and charged with developing a project that could be launched in the real world, whose goal would be to engage the themes of the conference. Each group then made further selections from that sub-set of themes, and over the course of a few hours including a working lunch, each group designed a proposal that took their chosen themes as a starting point. They also spent time creating visual representations of their project. In the afternoon the groups convened, and each made a short presentation of their proposal, followed by a general discussion on the issues the work brought up. The proposals designed and presented by the various groups took on many different themes and imagined quite varied outcomes. Ministry for Art and Social Change (MASC) A proposal to create a kind of "Post Hoc NEA" who would award artists grants after the fact, based on statistical evidence used as funding criteria. This proposal stimulated a long discussion about both whether social change was or should be an explicit goal of the artist, whether the artist should announce social change as a goal in advance of the work, and in any case, how such change might be best measured. National Downgrade Day and the Beta Fund National Downgrade day was a proposal to improve the quality of software by encouraging people to downgrade to previous versions of software, in order to stem feature creep. The Beta Fund was a proposal related to alternate funding models for software creation. These proposals received vigorous assent from the participants, leading to a discussion of code as cultural statement, and what the planned obsolescence of software says about our society, and whether the art world, with its rhythm of "What's now? What's next?" fell into the same patterns. Interactive Art Media Generation Engine (IMAGE) A proposal for a machine—generated judgment—free system for solving other peoples problems without requiring human decision. The engine would take a list of potential resources, and output a project. The imagined use for IMAGE would be to create a Blur brainstorming proposal. The discussion that followed was relatively flaccid, as the degree of meta-ness about the project made it difficult to propose any clear questions or statements about it. Ceci n'est pas une search engine A proposal for a search engine that would work on cultural principles. To allow for richer and more idiosyncratic or serendipitous views of the Internet than current search technology allows or encourages. It would also provide a mechanism (similar to that imagined by MASC above) to allow funds to be dispersed to creators based on traffic. The discussion that followed was principally concerned with the degree to which search engines shape their users' world views, especially unconsciously, and the degree to which interface designers can force their users into or out of certain behaviors. BLURLIFE An online multiplayer game that would illustrate social dimensions of power on a large scale. The proposal principally concerned the design of a planet whose economy/ecosystem would involve players in both micro- and macro-level decision-making. As many of the participants were interested in games, and especially massively multi-player games, the discussion centered around the difficulty of designing a game that is both pedagogical and fun. Cultural Security Distribution Options This group imagined a dramatic change in intellectual property laws with the introduction of a compulsory licensing scheme and a mechanism for collecting user fees and distributing them to creators based on traffic. A system that this group imagined would reduce the hold the content owners have on the conversation between artist and audience. The conversation sparked by the Cultural Security proposal was a mini-conference in itself, with heated debates about the role of the artist, the nature of popularity, and the role of the state. This was easily the most vigorous hour of debate at Blur. The effect of the presentations was cumulative and multiplicative as we became aware of themes that emerged from more than one group's work, thus making the presentations and Q & A both a way of introducing new material into the conference and helping manifest what the Quakers call "the sense of the meeting," the set of thoughts on everyone's mind that may not have been articulated in advance, or spoken by any one person. Desirable Effects Though the strategies that went into the design of the Saturday session weren't unique to Blur, they were used in a way that created especially effective working environment. Having been to a number of conferences with the kinds of goals embraced by Blur—surfacing of new and interesting material, creating connections between participants, seeding future work—I can say that Blur was unusually effective in four regards. First, it created an environment where theorists and practitioners actually engaged in valuable and constructive work together. Though bridging the chasm that separates these two groups is a stated goal of many conferences, the two groups tend to clump like the Jets and Sharks at the school dance. By designing the brainstorming, group presentation, and subsequent conversation around projects that had both a practical and theoretical component, Blur created actual dialogue. Second, the Saturday session actually advanced the state of play. The quality of some of the projects that were presented provided a better jumping off point for discussion than many of the panel presentations of the previous day. Indeed, the only time the discussion became really heated, and the profound disagreements as to point of view or goals among the participants became clear was during the discussion of the group proposals. Third, it helped solve the "talk time" dilemma. All conferences suffer from the same structural problem: how to balance talking and listening. In any group, the time available to any participant falls as group size grows, and the grim math of this equation makes it imperative that the hosts provide some sort of structure for making the most of the scarce resource of "air time." This problem is worst for conferences of 30-50 people. Below a couple dozen participants, the conversation can be relatively free-flowing. At a hundred or more participants, the solution is equally clear: Someone holds the floor, and then that position is formally passed along. By breaking into groups, each participant was able to engage in real dialogue in the kind of small-group setting conducive to real conversation, while the presentation and discussion session made the highlights of the group conversations available to everyone. Finally, the brainstorming sessions created work with effects that extended outside the conference. This is the brass ring: one of the Saturday proposals, "Cultural Security Distribution Options," involved a proposed non-discriminatory license fee for intellectual property in return for a sharp reduction in copyright restrictions. In turn, the proceeds from this fee would be dispersed to the artists whose work existed in the system. The conversation that followed this proposal was easily the most vigorous of the conference, and the presentation plus discussion crystallized many of the emergent themes of the conference that had arisen during the panel discussions of the day before. In addition, because the attendees of the Cultural Security group included theorists like Ted Byfield and activists like Alan Toner and James Love, the ideas were both conceptually engaging and politically grounded. The proposal has gone on to a life of its own after Blur, having been presented in alternate guises to other communities, in addition to informing the work of several of the attendees on practical issues relating to intellectual property. Taken together, the product of Saturday's work—real dialogue that advanced the understanding of the attendees and lived on after the conference was over—combined to make that session a model of conference design. (Indeed, it was so successful that I borrowed much of the format for a conference I hosted later than year on the design of social software. More on this below.) Even more remarkably, the work felt effortless as we were doing it—the presenters had done such a good job of designing the format that it was only after the session was over that many of us realized what a remarkable day it had been. It's tempting to look for magic bullets, but the success of the brainstorming at Blur was a mix of preparation and happenstance (though chance favors the prepared mind). Designing future sessions that are as successful, whether at Blur or elsewhere, will involve identifying both the good aspects of the design and the lucky breaks. Ground Work The success of the Saturday session created several results worth trying to copy in the future. Small group work creates valuable social dynamics that are literally impossible to create in larger groups; shared effort leading to a public presentation, no matter how ad hoc, can focus the minds of the participants. The quality of the conversation about work created in the room can be of a fundamentally different character than work done elsewhere and presented after the fact. Anyone wanting to create a brainstorming session with similarly valuable results will need to understand the elements that made the Blur session work. The Right Mix of People Every good conference starts with the invite list. Though there is always an official theme, the real work is getting the right mix of people to show up. Blur started with a good list of participants, including theorists (Byfield, Shirky), artists (Zurkow, Cosik), activists (Love, Toner,) coders (Pesce, Jennings) and founders of new media companies (Syman, Hourihan). This mix provided real diversity in outlooks. In addition to the number of different fields represented, there were a number of entrepreneurs who had spent little or no time in the non-profit sector, and the political spectrum ran from communitarian to libertarian, with representatives from several countries. (The international representation was somewhat limited by travel funds—more representation from Asia and Africa at conferences generally are desirable, but expensive.) The Right Degree of Clustering In addition to a good mix of views, the conference started with the right degree of clustering among the attendees. While everyone knew someone, no one knew everyone. Each attendee knew someone else at the meeting, but there was no one person who was familiar with all the attendees or their work. Getting this kind of clustering right is a balancing act—when too many people know one another, the conference can turn cliquish, and splits into insiders and outsiders. If it is too atomized, only the overtaxed hosts can provide introductory context and nothing gels. When everyone knows someone but no one knows everyone, on the other hand, you can get a virtuous circle of introduction between people who should meet but don't know one another. This happened at Blur. Prior Conference Material By making the brainstorming the last day's work, the entire conference up to that point—pre-conference readings, the introductory party, the first day's panel sessions and dinner—all provided background material to draw from. Because the brainstorming session involved both the official themes of the conference and the unplanned themes that emerged from presentations and conversation, making the brainstorming the last piece of substantive work maximized the shared inputs of the group. Two particular efforts from the conference prior to the brainstorming deserve special mention: The Survey and Awareness of the Social Network During the preparatory period, Blur attendees were asked to fill out a form indicating which other conference participants they knew and how well. Then, just before the conference, this information was turned into a huge map of the Blur social network showing participants as nodes and drawing the reported connections between us. This map was presented early the first day. By showing us an image of ourselves as a group, and by doing it early on, the organizers made us aware both of the general importance of the social fabric of the meeting, but also aware of some specific clusters (a dense cluster of New Yorkers, a dense cluster of Banff attendees) as well as outliers (Meg Hourihan, co-inventor of the Blogger weblog platform, was a new face to most). This social self-awareness made people more aware of the "I only talk to people I already know" problem and made them more willing to introduce themselves to other participants. (Indeed, the impression made by the social map was so large that several people took it upon themselves to annotate it with magic markers, in order to fill in lines they thought were missing.) The Blur Game and Awareness of Play Blur commissioned Marina Zurkow and Eric Zimmerman to design a game to be played during the conference. Rather than design a game to be played intensively in one session, they opted for a game that could be played at low levels throughout. The game consisted of "power cards," which gave each player the right to demand or enjoin certain behaviors of the others. The cards came in an appealing variety of flavors, from the personal (You Must Tell Me a Secret) to the conference-wide (This Card Allows the Bearer to Give Explicit Instructions to the Speaker), as well as a handful of meta-cards (This Card Can Take On The Powers of Any Other Card, This Card Allows You to Take All of Another Person's Cards). Like the social map, the cards made the social structure of the conference itself both more interesting and more explicit, and the heightened awareness by the participants of the kinds of social operations that were possible within the conference made the brainstorming environment richer. In addition, the cards played a role during the Saturday session. They were used during various parts of the day: the introductory comments were short-circuited by playing a card on the conference organizer, who was forced to cut to the chase; cards were played to 'steal' a member from another group whose expertise was needed for a few moments; and during lunch, the game provided a kind of backdrop to the work at hand. Finally there were the choices made about the set-up of the brainstorming itself. These choices were the least complex, but it is remarkable, in conference design, how often the simple choices are overlooked or gotten wrong. The Right Groups Diversity alone is not enough, as diversity plus cliquishness is nothing more than segregation. The conference designers took care to design the groups in advance, rather than letting people self-segregate, in order to take advantage of the diversity of views they'd brought together (no letting all the programmers join a single group) and to create clusters of people who should meet but might not otherwise have a chance to do so. The Right Kind of Space The space Blur took place in was ideal for this kind of loose-jointed work—a large, open high-ceilinged room, with natural light and an outdoor area. The Right Amount of Time The brainstorming session was given the whole day to unfold, from the first identifying of possible themes to focus on through the final discussion. Because of the dilemma of talk time, noted above, the tendency for organizers is to schedule conferences quite densely. While Blur had such dense scheduling on its first day, this was nicely balanced by time for more serendipitous discovery on Saturday. A Mix of Structure and Freedom The brainstorming exercise had a good mix of structure (the themes to be used as inputs were listed in advance; visual representation of the idea was a requirement for the presentation) and freedom (groups could add new themes; the project could take any form). This allowed groups of people working together for the first time to get past the ground rules and on to the more valuable work, while giving them the freedom to astonish or amuse one another with ideas well outside the proverbial box. Problems No format is perfect, of course. There were some unavoidable problems with the Saturday session, and other problems that arose from the way it was executed. If future conference work, either at Blur or elsewhere, draws on this form of conference design, the former problems should be addressed up front with the participants, and the latter set of problems should be avoided where possible. There were two principle problems in the unavoidable category: the variability of group work, and the size vs. time tradeoff. Variability of Group Work Some groups gel, others don't. There are things the conference organizers can do to mitigate this problem as much as possible, but collaboration is not like cooking; there is no recipe for making group work come out perfectly every time. This problem was lessened at Blur by putting the collaborative session on the last day. The attendees who came to speak but not to listen simply opted out of the group work, thus removing the damaging element of the group member who feels forced to participate against their will. Even with that danger largely removed, however, the difference between the most and least engaged groups was significant. Though this is an inevitability in group work in general, in the future conference presenters should make the dangers of this problem clear. Size vs. Time Though the small group work was itself a way of managing the size vs. time tradeoff in the conference format itself (see the reference to talk time, above), here too there are issues of scale. Research into small group work suggests that the ideal sized for focused teams is 5 or 6 members, which effectively suggests that the number of small groups at a conference should have been number of attendees divided by 6. However, when each group needs to present, having more than half a dozen groups may mean that several hours must go to presentation. Blur opted to have 6 groups of 8 or more members, rather than 8 or more groups of only 6 members, and even with that format, we had too little time at the end of the conference for presentation and discussion. Size vs. time tradeoffs are endemic to conference design, and present no ideal solution, but the necessary premium on group focus suggests keeping the half-dozen member small group as a goal where possible. In addition, there were some issues of conference format during the Saturday session. Balance of Time Though the long period of relatively free-form time for creating the ideas was valuable both for the presentations themselves, and for allowing conferees to have the kind of hallway conversations that are the making of any successful conference, the presentation and discussion time at the end of the day was too compressed. Indeed, there was a mini-revolt (led in part by me) when the organizers attempted to truncate the discussion of the Cultural Security Distribution Options proposal. The organizers were so mindful of clock time and the schedule that they almost short-circuited exactly the sort of discussion that made Blur worth doing. In future versions of this work, the organizers should assume that the material produced in energetic brainstorming will be of considerable interest to the group as a whole and that the most interesting topics for the discussion will have to be discovered as the conversation unfolds. This in turn will require scheduling to provide more time for plenary discussion. Balance of Rules Though the overall balance of freedom and constraint provided fertile ground for brainstorming, some of the rules we were given were far too specific. In particular, though the general scope of the proposals was quite variable (as evidenced by the results, where different groups proposed software, game environments, events, and even government agencies), many of the instructions were so specific as to be detrimental. The overlay of narrow instructions can kill collaboration. Future versions of this work should strive for instructions related to intent (Make a visual representation that will help you explain your work to your colleagues) rather than to output (Make a drawing of the user interaction within your proposal). Anything You Can Do, I Can Do Meta The mirror image of undue narrowness is overwhelming breadth; give people enough scope, and they'll hang themselves. Though the brainstorming rules were intended to provide a simple and useful framework for conversation, the question immediately came up as to whether the group could, as part of their work, subvert those rules. This was a typical request, as creative people will naturally try to subvert rules in their favor, and was made inevitable in this context, as the subject of the conference was itself power relations in aesthetic and cultural work. The organizers' answer was, "Yes, you can change the rules as you go." This turned out to be the wrong answer. The only presentation of the six that could be deemed a failure was in the group that asked the question, and then spent their time building an elaborate set of self-referential explanations for why they weren't going to produce any concrete thoughts to present to the group. This wasn't because breaking the rules was wrong per se. The Cultural Security Distribution Options group, the most energizing proposal, also broke the rules. The difference was between the letter and the spirit of the law. Both groups thus avoided the narrower dictates of the rules, but the IMAGE group also avoided the spirit of the exercise, and used the rule-changing rule as an excuse. The theme of the conference wasn't just power, it was power at play, and play involves some sort of rules. Zurkow and Zimmerman's game demonstrated the possibility of having a game with rules whose outcome was both playful and surprising; brainstorming has many of the same requirements. An alternate answer, more in keeping with the goal of playfulness, would have been, "You can change the rules, but only in the direction of having a more engaging proposal." Since the construct created by the rules was somewhat arbitrary, designed to provide some shape to a quick exercise among people who didn't all know one another, overthrowing the rules for its own sake, instead of as an alternate route to the goal of interesting thought, was a fairly pointless exercise. Critiquing the conference by opting out is always a possibility for individuals, but takes on a different cast in group work, where one or two people can dampen the experience for people who might otherwise be inclined to engage with the exercise. In future versions, the participants should be held more closely to the rules (provided those rules themselves are sufficiently general, see above), with the goals of the conference providing a guide to when and how to break those rules. Future Work The Blur brainstorming session was sufficiently well thought out in its design, and produced such good results, that it merits descent with modification. Perhaps the key insight is that if conference organizers assume that this kind of brainstorming session will produce interesting work that could not arise any other way, its inclusion will begin to affect the design of the whole conference, from the early planning phases to follow-on work after the conference is over. Here are several possible pointers to future versions of this method of working: Extensibility The Saturday session of Blur was so effective in its overall outcome, even with the problems listed above, that I copied the form for a two-day meeting I held in November on the design of social software. Though the attendees of the conference were all software designers or theorists, I borrowed a number of the blur design principles. I arranged the invite list so that everyone knew someone but no one knew everyone (not even me). The conference was also a two-day conference with presentations on the first day and structured brainstorming on the second day. I designed the groups to maximize fruitful conversation, and the groups doing the brainstorming created visual representations of their work and reported back to the group. Though the domain of inquiry was not aesthetic, the best of the results mirrored Blur, in that several of the conversations started then have continued outside the conference, and several of the attendees have gone on to build and launch the pieces of software whose basic outlines were suggested during that session. Though this is no guarantee of the general extensibility of the Blur form, it is at least an existence proof that the explicit and implicit design principles that drove Blur have applicability in at least some other environments. Social Prosthetics Both the game and the social map functioned as social prosthetics, tools for extending group work. We are all accustomed to personal prosthetics at conferences—microphones extend our voice, projectors extend our visual presentations, and so on—but social prosthetics are rare and valuable. The map could be extended to include more than one axis of affiliation between people—worked with, studied with, friends with—or could include non-human nodes—connections not just between people but institutions and events as well. Likewise, the game could be made more interactive and integral by fusing game rules and brainstorming rules, so that certain "moves" in the design or presentation of a proposal would also be moves in the game world. And the map and game both point to the possibility of other kinds of social prosthetics such a software or other tools to make the social dynamic a formal part of future conferences. Wireless Tools One type of social prosthetic that existed in partial form at Blur was the presence of the Wifi network. By having wireless Internet connectivity, hosted by The New School, Blur participants with Wifi-enabled laptops were able to connect to the Internet throughout the conference. During the brainstorming session, the Wifi network was used by some of the groups to research or gather material as input for their presentations. These uses were ad hoc and unevenly applied, however, in part because Wifi was such a new technology at the time, so relatively few people had Wifi-capable browsers. For the social software conference, I copied Blur's inclusion of the Wifi network, but put it in the position of the Zurkow/Zimmerman game—a constant "2nd channel" presence. I wrote about the results of this experiment at length here: http://www.openp2p.com/pub/a/p2p/2002/12/26/inroom_chat.html For the organizers of future conferences, Wifi offers a way to create wireless social prosthetics, technological extensions to the group as a whole. I believe this will be a particularly fruitful area of inquiry for any conference interested in group dynamics, gaming, or aesthetic collaboration. Rules and Meta-Rules Both the social network map and the game demonstrated that by making a group aware of social structure and forces can encourage creative responses. As noted above, the Blur work occasionally erred on the side of too much specificity (micromanaging of the visual presentation) and at others in the direction of too much generality (granting carte blanche to break the proposed rules). A middle way might be possible and fruitful: design a brainstorming session along the lines of the game. One could create a set of rules all participants must adhere to, but make those rules sufficiently general, and provide enough meta-rules, that the participants could stretch the boundaries of the task without rupturing it. Iterative Brainstorming The end of the presentation and discussion session on Saturday was an exercise in frustration—like many valuable experiences, many of the participants felt we were just scratching the surface even as the conference was ending. One possible area of future work would be a conference designed more centrally around such brainstorming. Though it would take more time, probably a third day, iterative brainstorming might well produce work of greater depth and sophistication, as well as making the value of small group conversation more core to the event itself. The Blur structure was plenary/small group/plenary, with a short period of identifying shared themes up front and in the large, a long period of designing the presentations, and a final plenary for presentation and discussion. However, further themes of interest arose in the presentation and discussion, and some of the proposals worked especially well to crystallize certain points of view (or sometimes to crystallize two opposing points of view). Given the plenary/small/plenary structure, it is easy to imagine ways to do this iteratively, where certain proposals were selected for further iterations, group memberships were shuffled, new proposals were created from emergent themes, or groups were asked to make proposals that acted as critical responses to the earlier work. The point of these sorts of iteration would be to continue surfacing new or emergent themes, or groups were asked to make proposals that acted as critical responses to the earlier work. The point of these sorts of iteration would be to continue surfacing new or emergent material, and to approach important themes or problems from a variety of angles. Conclusion The design of the Blur conference produced real dialogue between individuals and groups who would not ordinarily meet, and the brainstorming session allowed those people to work together faster and more fruitfully than any form of presentation or panel conversation would have. The presenters did such a good job of creating the list of invitees, designing the format, and managing the session that the conferees were able to do good interesting work without being too bound by structure or too lost in structurelessness. This was no mean feat, and while the content of those conversations was quite valuable, the formal construction put in place at Blur may be the more important outcome. The critical question is whether the Blur design is useful and flexible enough to take on a life of its own. My own subsequent experiments suggest that the Blur brainstorming model can work well in other environments and can produce work that has effects outside the framework of the event itself. It would be fitting if one of the outcomes of a conference entitled "Power at Play" was a conference design that was powerful, playful, and valuable enough to be applied to future work.

All rights reserved © New School University // 2002