Platformer - Introductory White Paper - Working Draft V0.61

Platformer Open Protocol and Technology Suite for Distributed Coordination and Exercise of Individual Political and Economic Power

Introductory White Paper Working Draft v. 0.61 Noel Bush <[email protected]> Berlin, September 2009

Contents A Note About This Document...........................................................................................................................1 1. Overview......................................................................................................................................................1

1.1. What is Platformer?..................................................................................................................................................1 1.2. Why is it needed?.....................................................................................................................................................1 1.3. How does it work?....................................................................................................................................................1

2. Background...................................................................................................................................................2

2.1. A survey of the problem...........................................................................................................................................2 2.2. Approaches to a better model..................................................................................................................................3

3. Key Concepts................................................................................................................................................4

3.1. Constituencies..........................................................................................................................................................4 3.2. Platforms..................................................................................................................................................................5 3.3. Positions...................................................................................................................................................................6 3.4. Coalitions..................................................................................................................................................................7 3.5. Pledges.....................................................................................................................................................................7 3.6. Deep Profiles............................................................................................................................................................8 3.7. Campaign Architects................................................................................................................................................9

4. Technical Overview of System......................................................................................................................9

4.1. Fully distributed.......................................................................................................................................................9 4.2. Open source (free software).....................................................................................................................................9 4.3. Secure and private..................................................................................................................................................10 4.4. An open platform for application development.....................................................................................................10 4.5. Re-implementable..................................................................................................................................................10 4.6. Verifiable................................................................................................................................................................10 4.7. International, multicultural....................................................................................................................................10

5. Technology Implementation Plan...............................................................................................................10 5.1. Protocol Specification.............................................................................................................................................10 5.2. Node Reference Implementation............................................................................................................................10 5.3. Application Reference Implementations................................................................................................................10

6. Community Development Plan...................................................................................................................11 7. Expected Impact.........................................................................................................................................11

may possess only a tiny quantity of political or economic power.

A Note About This Document

This document describes a system that is as-yet unimplemented. For the sake of clarity, however, we often speak herein as though the system already exists. As the system is actually built, it is anticipated that this document will evolve alongside it, so that the description of “what is” according to this document will increasingly come to match what, in reality, is.

1.3. How does it work? Platformer provides the means of coordination in a form that truly belongs to everyone. All of the Platformer protocols and technologies are open source, meaning that they can be freely distributed 1 and modified. Platformer is not a single piece of software, nor is it a web site or web service. Platformer can be most readily compared with two of the key technologies that underpin the Internet and World Wide Web—HTTP, and HTML. Like HTTP, Platformer describes how comput2 ers can talk to one another: web browsers use HTTP to get content from web servers; Platformer specifies how nodes can request and exchange information related to identifying constituencies and organizing actions. Like HTML, Platformer supplies a format for the exchange of information: web servers deliver web pages to browsers as HTML; Platformer specifies the means for capturing and encoding information about political positions, intentions, planned actions, and results. Generally, people do not use Platformer directly, just as they do not use HTTP or HTML direct3 ly. Instead, most people use applications that communicate via Platformer. So, for example, someone who has previously established a link among several online identities, such as a Facebook account, a blog, and an instant messaging identity, might answer a poll on Facebook that asks for a position on a war. Later that day, the same user might receive a proposal via instant messaging to participate in a boycott of a popular coffee chain—which that person frequents —with the explanation that this action will have a measurable effect on military policy. Since the connection will likely not be immediately apparent, the user will be invited to click a link which opens a page explaining how the seemingly unrelated boycott of the coffee shop will lead to legislative action on war policy—in this case, perhaps, the corporation that runs the coffee shops will receive a message stating that a significant

1. Overview 1.1. What is Platformer? Platformer is a set of technologies that enable collaborative, distributed exercise of individual political power in coordinated actions with explicit and measurable goals. Platformer works primarily as a kind of infrastructure that allows individuals around the world to coordinate activities in ways that were previously unthinkably complex. Platformer allows any individual to make use of political and economic power that might otherwise seem so negligible as to be useless, by coordinating with others across the globe, and across ideological, socioeconomic, cultural and even linguistic barriers. Furthermore, Platformer preserves individual control of personal information—personally-identifying data as well as information that is relevant to a person’s political points of view and economic status and priorities. Instead of submitting data to a campaign organizer or political party organization and hoping that the data stays private and isn’t used for nefarious purposes, an individual has an iron-clad guarantee that his or her information is used only to advance the causes of the individual’s choosing. In a reversal of the present-day situation where a “marketing”-oriented approach defines political organizing, Platformer puts control of personal information back where it belongs. 1.2. Why is it needed? We all know that any two people may differ from one another on many important ideological points, yet share a common concern for a single cause. But ideological divisions can often be so overwhelming that they prevent shared action on such common causes. Add to this the difficulty of coordinating across geographical distances, political boundaries, and cultural and language barriers, and it is easy to understand why so many matters on which we know that many people have common views simply cannot be addressed in a coherent manner. Yet these difficulties in coordination are, at base, often just that: logistical issues. History and contemporary experience both show us that, whenever it is possible for groups of people to mobilize on issues of importance, change can be effected, regardless of the fact that each individual

Platformer: Introductory White Paper (Draft)

1

2

3

1

More precisely, the technology is available under the GNU General Public License (http://www.fsf.org/licensing/ licenses/gpl.html), and hence should be termed “free”. However, frequent misunderstanding of “free” in English (“free as in freedom or free as in beer?”) tends to lead to the colloquial use of “open source” even when “free” is what is meant. The bottom line is that the source code for the software is available to anyone for free, that anyone can modify the software freely, and that these same rights must be extended to users of all derivative works—in other words, they cannot be revoked. More correctly, HTTP describes how certain programs can talk to one another—the communication between actual machines happens at lower levels of the “protocol stack”— but for this analogy we are not concerned with the details. Even though people do sometimes write HTML “by hand”, rarely does someone read HTML directly—instead, the markup tells a browser how to display text, images and other content. http://platformer.org

quantity of regular customers pledge to boycott the coffee shop unless the corporation uses some of its lobbying funds to press the legislature to vote in the desired direction. While there currently exist web sites that approximate some of this functionality, a critical problem is that the data on these sites is managed by a single entity. Legal protections such as privacy policies, and IT security policies, presumably protect user data from unwanted exploitation. But they also prevent the use of that data for many constructive ends. The centralization of the data as well as the application that manipulates it means that a tremendous amount of personal information is spread around, and duplicated, across many sites, yet its collective political and economic power can only be leveraged by any one individual to the extent that any given site owner imagines and offers such possibilities. One key characteristic of Platformer, then, is that the communication and coordination services it provides are not centralized. There is no single entity that houses the data that Platformer manipulates, nor any single location where its various component programs run. Rather, all storage and processing is distributed, and anyone may participate in operating the infrastructure of the network by running a Platformer “node”. Anyone may also develop an application that makes use of this data, and anyone may use tools that permit the creation and management of issue-based campaigns. However, critically, Platformer builds in unassailable security mechanisms that make it impossible for anyone to access anyone else’s personal data. While it may seem paradoxical that anyone can run a node, and that the nodes together hold the data, and yet the data cannot be accessed by any given individual, this is possible thanks to the manner in which all data is distributed. Essentially, every piece of data, including the identity of the individual it belongs to, is broken up into several pieces, and those several pieces cannot be assembled by anyone except the owner of that data. So an application developer can specify, for example, that a message should be sent to all people who favor a given ballot initiative, without ever knowing the identities of any of those people. The security mechanisms are vital for making Platformer useful for people under political regimes that suppress or punish dissenting opinions. Data collection and profiling by authoritarian states becomes much more difficult with Platformer, because it is impossible to seize a physical (or a virtual) asset and uncover individual identities. That the Platformer technology set is open source is key to ensuring the security of the system, because this allows anyone with adequate expertise—or anyone with access to someone with adequate expertise—to evaluate the software in

Platformer: Introductory White Paper (Draft)

use and satisfy, personally, any concerns about 4 safety.

2. Background 2.1. A survey of the problem Platformer began as a meditation on the inordinate influence of corporate lobbies in governmental functions such as legislation and oversight. This has long been acknowledged as a growing problem in some parts of the world. In most cases where reform is attempted, efforts focus on regulatory limitation of corporate influence. However, since the exercise of political power by corporations is so deeply intertwined with the rest of the economic ecosystem, there is almost inevitably an organic counter-response to any regulation—that is, when a channel of influence is closed off by new regulation, corporate interests simply develop new means of exerting influence that are not limited by laws. The situation is further complicated by the fact that the regulations limiting corporate influence are themselves targets of influence by corporations. In any market-based economy, then, it seems inevitable that corporations will exercise tremendous power, over many aspects of political, social and economic life. But admitting this reality does not mean that we must consign ourselves to accept interminable postponements and reversals in ongoing efforts to extend justice, equality, human rights, quality of life and education to peoples around the world, in deference to profit-based motives of corporations that almost always defer higher aims to “later”. Instead, we might look at ways to harness the existing power of these corporations in the service of the people who, after all, ultimately keep them in business. The de facto method of mobilizing for change has generally been the “grassroots” approach. When an issue has no organized advocacy, organization must start from scratch. In the case that no existing agency (political party, union, etc.) can be persuaded to take an interest in a cause, individuals must organize “on the ground” in efforts that often demand enormous amounts of time and energy. Each new cause must compete for attention in a world saturated with messages—most of which are consumption-oriented, not action-oriented. In a world full of entreaties to “buy and enjoy”, it is extremely difficult to reach enough people with a “join and take action” message in order to achieve even the simplest of goals. And for each person who is successfully recruited to a given cause, the pool of people available for another cause has likely diminished by one—most of us are 4

2

This perhaps surprising utility of open source to security concerns has been acknowledged by none other than the National Security Agency of the United States, which maintains an open source version of Linux called Security Enhanced Linux (see http://www.nsa.gov/research/ selinux). http://platformer.org

simply unable to commit to more than one or two causes (and many cannot even dream of committing to a single one). We do, of course, tend to think most readily of political parties when looking for ways to address causes of importance to us. But the political party system embodies a crippling paradox: almost always, the effectiveness of a political party is inversely related to the breadth of its platform. In other words, a single-issue party that draws passionate supporters, and may even enjoy wide membership, finds difficulty in effecting change because it must compete on a playing field with parties whose complex platforms involve them in a wide variety of issues, and thus provide numerous levers of influence. On the other hand, these larger, more mainstream parties, whose overall power may be much greater, have attempted to include so many constituencies that they cannot afford to pursue most of the causes under their purview with the energy that one would desire—and they often end up failing to represent any particular constituent adequately. This leads to apathy on the part of constituencies, and tends to make successful political parties stay successful by relying more on emotional appeals—on marketing, essentially— than on substantive, issue-based engagement. The result is a weakening of the political system, in which prominent political parties fail to represent people’s interests adequately, and “fringe” parties expend huge energy just to try to be heard. Since many corporations cultivate expertise in marketing to thrive, they can offer this expertise to political parties in exchange for favorable representation in government. Corporate influence helps to keep the fringe on the fringe, and to preserve the status quo in government. To the extent that “consumers” (as corporations term human beings) may express dissatisfaction with a given regime, it is a small matter to incorporate “dissent” into the overall marketing portfolio. But in no case can one see an alternative to the paradigm itself. Efforts to overturn this state of affairs—that is, revolutions—risk death in the cradle because, first, they can be coopted by the very same corporateorganic forces that attempt to assimilate any social foment into market-fed “culture”, and second, because explicit threats to the fundamentals of the established order are perceived with alarm, and addressed with efficient and thorough censure.

ple to direct them in ways that integrate with life as it currently operates. It has long been recognized that the boycott can be a spectacularly effective means of accomplishing what legislation cannot (or legislators will not). But generally, boycotts are aimed directly at the entity whose activity we wish to change. Activists boycott a fast food chain, say, because their livestock treatment practices are determined to be cruel. Success is measured by whether the corporation changes its practices. Somewhat indirect boycotts are sometimes implemented. Companies that did business with the former South African regime were boycotted, for example, and this indirectly helped to topple apartheid. Currently, there are efforts underway to spread boycotts of companies who do business in Israel, in an attempt to influence policies there. But still, in such cases as these, we can see the inherent limitation: it is only possible for someone who currently patronizes the target of a boycott to participate. Only people who bought wine could participate in a boycott of South African wine, for example. Further, the only situations that wine drinkers can influence, in the traditional boycott model, are those in which some part of the production of wine is an economic element. But there is no reason to accept these limitations. Virtually every person alive has some political and/or economic power. Even those people who do not earn income, who live in abject poverty, still possess the power intrinsic to being a human being—they can work, they can refuse to work; they can move, they can stay still; they can speak, they can be silent. And when we consider those of us fortunate enough to have some degree of economic agency, even the slightest amount, then we can see even more obvious power. Even the individual who works in a job that barely covers expenses, and who has little to no choice in those expenses, possesses the power to work or withhold work, to buy or to not buy (even necessities), to speak or to stay silent. If a person sees the possibility to effect real change, even difficult actions in difficult circumstances seem worthwhile. The problem, generally, is that for the vast majority of us—even those of us in the First World with disposable income—the amount of political and economic power we possess individually is so minuscule in that we cannot use it. We know that if we could get “everyone” together who felt the same way we do about a particular cause, we could make something substantial happen. But the seeming impossibility of contacting individuals who share our interest in just one given issue—especially if those individuals may be scattered across a country or a globe, may exist at different socioeconomic levels from us, may speak different languages, may hold radically different views on other issues—prevents us from even exploring this path.

2.2. Approaches to a better model So instead of plotting yet another “master plan”—really, a utopia—that dictates away all the problems, but falters (or explodes) on implementation, it makes sense to investigate developing parallel infrastructures that complement existing systems whenever possible, and address power imbalances not by seeking to regulate away (or obliterate) the powerful entities that so often work against the greater good, but rather to enable peoPlatformer: Introductory White Paper (Draft)

3

http://platformer.org

But it was from a consideration of this state of affairs, and from an immersion in the burgeoning technologies of “social networking” that began to mature and proliferate on the web several years ago, that the notion formed that it might now be possible to organize at this atomic level—to bring together people from radically different “constituencies” (from the perspective of a political party) to unite in common, effective action on specific causes, and to enable anyone to initiate such activities of organization, in a manner and with an ease comparable to that with which we now are able to retrieve and exchange information in multiple media on a limitless variety of subjects, using the technologies we now take for granted as “the web”. Social technologies have revealed some critical inadequacies of the mechanisms in use in many liberal democracies. Namely, individuals have a variety of concerns, prioritized in unique ways that cannot possibly be represented by a given party platform. Individuals have a limited amount of energy, usually very little, to devote to any given cause. Bad experiences with being misrepresented by political parties cause people to disengage. Bad experiences with seeing grassroots efforts not produce desired effects cause people to disengage. And fear of losing control of private information inhibits people from participating. The bottom line is that an individual’s actual political power, such as it is, is essentially not something that can be effectively wielded by the individual. Either they must surrender this power to some entity that is not likely to make use of it in a way that is maximally desirable for the individual, or they leave it primarily in the service of corporations whose products they buy (and whose causes they thereby indirectly support). So the notion, here, is to turn this situation around: Put the full amount of an individual’s political power at their disposal. Make an individual’s private data fully usable by that individual without surrendering privacy to a third party. Enable individuals to direct their power toward precisely the causes they care about. To accomplish this, we want to (i) reverse the vector of control so that individuals direct larger entities to advance causes on the individual’s behalf, and (ii) provide an alternate means of organization so individuals can easily coalesce around particular causes, without needing to “buy in” to causes with which they don’t agree. As we approach more concrete ideas of what might be done, some key concepts begin to emerge. These are outlined in the next section.

Platformer: Introductory White Paper (Draft)

3. Key Concepts 3.1.

Constituencies

Illustration 1: You are at the center of your own constituency.

The term “constituency” generally refers a structure found in representative democracies, where an elected or appointed agent is responsible to a segment of a population. These segments may be defined by political boundaries, or by key characteristics or interests (e.g., elderly people, people with chronic diseases, people of color). If such a segment is identified and associated with an agent (usually by means of an election), then the people in that segment are referred to as “constituents”. But, generally, the existence of a representative agent is always implicit—it would seem strange to talk about “constituencies” without representatives. Yet, in an idealized sense, we can imagine that every individual is a member of multiple constituencies, with a different level of vested interest in each, and that the union of all these memberships constitutes a unique constituency with the individual as the center. If you, as an individual, are empowered as an equal participant in a negotiation of needs and policies, then you are essentially the representative of your own constituency. Other people belong to your constituency, to greater and lesser degrees. We may imagine those other individuals whose priorities and needs most closely match a given constituency’s central individual to be located closest to the center of a circular region, and those who are less and less close in overlap to be farther away. Likewise, you belong to numerous other constituencies. Of course, you are not and cannot be aware of all these constituencies, or of the memberships of others. These constituencies themselves may also be in constant flux, as new issues come to the fore in your life and others recede, and the same is happening for other people. But conceptually, at least, this seems to capture the reality of the 4

http://platformer.org

changing beliefs, needs, and priorities of individuals in relation to one another, far more accurately than do political parties. Note also the critical distinction between this conceptualization, and the more familiar notion of “interest groups”. Interest groups are generally defined externally, and individuals are sorted, or sort themselves, into these groups, and this is thought to be a way of understanding the needs and priorities of people en masse. But the weakness of this approach, aside from the need to continually review and refine the group definitions themselves, is that it omits the point that what is most important about these groups is how they relate to one another. It is not enough simply to register that this segment favors increased handgun controls, and this segment favors lower taxes, and that this segment is the intersection of the two—it is critical to understand the system from the viewpoint of the individual: individuals are not interchangeable, and the flux of influence exerted by different groups over one another owes its dynamism not simply to the numerical size of different segment memberships at any given snapshot in time, but to the history followed by each individual as they adjust their priorities, and positions, based on the flux itself. In other words, this is a classically nonlinear system, and analytic approaches that attempt to view it from top-down, outside-in, must always falter on the shores of statistical approximation. Political scientists may make predictions about trends in opinion, and these predictions may sometimes reach a useful degree of accuracy, but this model cannot translate adequately to a system that aims at empowering the individual. Hence our inverted approach, the “individual constituency” model, is a key concept for Platformer, because it assumes from the beginning that we care most about an accurate model of every individual. Note one other interesting characteristic of this model of constituency: there is not always a hard border delineating who is in and who is out of a constituency. As suggested by the illustration above. This characteristic becomes especially interesting when we consider the next concept, platforms.

from discrete exercises in compromise among party constituencies, the platform of an individual constituency undergoes continuous revision. It reflects the explicit beliefs and planned actions directed by those beliefs on the part of the members of the constituency. But a platform, in our new sense, must still delineate a discrete set of individuals—the “smoothly fading” constituency membership from above must be viewed slightly differently. We can imagine the same figure as in Illustration 1, but this time with boundaries identified as concentric circles around the center:

Illustration 2: Marking boundaries among members of a constituency.

Each of these boundaries marks off a different platform. The most focused platform, likely, will be the one in the center with the smallest membership; the broadest will be the one enclosed by the largest circle, with the largest membership.

3.2. Platforms Political parties and similar organizations have platforms. A platform is generally understood to be an explicit list of beliefs, and of projects or planned actions directed by those beliefs. Thinking again about the traditional notion of constituencies, we may recognize that platforms often represent the result of a process of compromise among various constituencies represented by a party. The strength of a constituency’s representation in a party will generally determine the degree of the party’s fidelity to that constituency’s priorities. In the individual constituency model, the notion of platform falls out rather directly. Again, instead of being an explicit “product” that results Platformer: Introductory White Paper (Draft)

Illustration 3: Deriving platforms from constituencies.

What do platforms contain? Previously we considered “beliefs, projects, and planned actions”. In

5

http://platformer.org

Platformer, we may revise this characterization and talk about positions and pledges.

ate this process of articulating his or her own specific position, we may say that it would follow this path: 1. Look at the existing variations on the position. 2. Does one of these variations match my beliefs? If yes, choose it. 3. If not, add a variation at the appropriate place in the tree. The job of the system, then, is to enable people with varying degrees of skill, interest, and expertise to participate in this sort of process. People with appropriate skill and interest may also take on the role of proposing edits to existing trees, and reviewing edits proposed by others. In a wiki-style collaborative mode, we can envision positions being continually articulated and refined by this sort of process. As individuals locate themselves on different position trees, these choices determine their membership in “constituency space”. These choices, in turn, produce the “platforms” that we may view by drawing boundaries within a constituency.

3.3. Positions In Platformer, “position” means essentially what it does in usual political discourse—a statement of belief or principle. A position may be broad in scope or very specific. In usual political activity, positions are stated in natural language that is considered carefully by the group of people who write the platform, and voted on by party membership. In Platformer, since we have inverted the situation, we recognize that different individuals will subscribe to different variations on a given position. Rather than offering a simple binary yes/no choice for a given position, we want to allow individuals to truly represent the nuances of their points of view on a given issue. The process through which an individual chooses a variation on a position is in fact how that individual locates him or herself within a constituency. It is easiest to think of this as a sort of (inverted) tree. We may begin with a simple statement:

3.4. Coalitions So we see now how the position-based parts of dynamic platforms may be derived. But what is to be done with these positions? Just as a party platform pledges certain actions that derive from the party’s positions on various issues, and just as the party then relies on the commitment of its members to supporting those planned actions, so too should our system incorporate the notion of planned projects and actions. But due to the special aspects of how our model is constructed, we can offer a much more powerful kind of coordinated action than traditional political organization. Consider, again, the virtual platforms derived from our banded constituencies. At any level of scope (that is, “tightness” of focus around the center of the constituency space), we can identify a unique platform—essentially, a “flash political party” (like the flash mobs of the early 2000s). Although the individuals contained within this grouping may differ in numerous respects, we know that they have a certain basis of commonality—namely, the platform that has been identified by selecting them from the given constituency space. Now, imagine repeating this process of platform identification, centered on different individuals. Each individual, having a unique constituency space, will also generate unique platforms. But the platforms can be compared with one another. They can be ordered by similarity. And it is here that the full power of all these concepts come together. (See Illustration 7.) By finding similar platforms derived from different constituency spaces, we can identify coalitions. The total set of members of any given coalition may be very surprising—it may include individuals who would never imagine the degree of

Illustration 4: A basic position.

An individual may wish to qualify this statement. We link the qualified version of the statement with the original:

Illustration 5: A qualification added to a position.

As more individuals encounter this statement, they may wish to add further qualifications. Sometimes, the same qualification will be added in two different locations, each addition representing a distinct variation on the position:

Illustration 6: A qualification added to two branches of a position tree.

This process may continue extensively. If we imagine an ideal way for an individual to negotiPlatformer: Introductory White Paper (Draft)

6

http://platformer.org

commonality that they share with others in the set. But share they do, and it is these unlikely alliances that now offer the possibility to achieve wholly unprecedented organizing results.

available to anyone who wants to dig into it. What it can do is simply state, to both constituencies, that if enough individuals make the necessary pledge, a significant action will be performed that will target the cause with which they are concerned. This situation is a relatively simple one, in which two constituencies are trading influence. It is not even necessary for the two constituencies to know of one another, or even to agree with the cause that the other supports. All that is necessary is a willingness to commit to the pledge proposed by the system. Of course, the obvious question here is: what action can the system propose? This is addressed by the next major key concept of Platformer.

3.5. Pledges A typical, traditional campaign asks respondents or members to make pledges—pledge to devote a certain amount of time to a cause, pledge to provide financial support, pledge to march in a rally, pledge to write to a legislator, etc. The action part of a pledge usually targets, more or less directly, the audience or entity whose attention or action is desired. The people who are asked to pledge are generally people who have self-identified as having a stake in the issues under discussion. With our system, however, we have the potential for far greater “leverage” of the interests, needs, abilities, positions and desires of the individuals in a dynamically-identified coalition. Since the set of concerns within a coalition may be quite diverse in nature, it is conceivable to ask members of a coalition to lend support to the various causes encompassed by the coalition, by requesting pledges that are within the abilities of members to fulfill, and yet might never have been imagined by the individual members. What is more, this request can be accompanied by a description of the expected outcome, with a greater degree of certainty than is generally available in traditional campaigns. Let us construct an example. We’ll consider two unrelated causes. The first, let us say, is police brutality in Russia. The second, for contrast, concerns an endangered bird in a forest in Oregon. People concerned with the first cause want to compel the Russian government to crack down on a rising wave of police brutality against non-conformist youth in Russia. People concerned with the second cause want to balance out powerful paper lobbies in Oregon that are seeking to destroy the natural habitat of the endangered bird. Superficially, there would seem to be no way for these two constituencies to assist one another—or, more bluntly, no reason for them to imagine trying to do so. But Platformer has identified a substantial number of individuals who belong to these constituencies as also being members of a coalition formed by several similar platforms. As a result, our system can propose what amounts to a trade between these two constituencies. The group concerned with Russian violence is asked to commit to one pledge. The group concerned with the Oregon bird is asked to make a different pledge. As it happens, the pledge made by the Oregon-interested group will have an effect on the Russian situation, and the pledge made by the Russian-focused group will affect the Oregon situation. The system does not need to explain the mechanics of the whole arrangement, although this information is Platformer: Introductory White Paper (Draft)

3.6. Deep Profiles In order to propose relevant actions, and indeed in order to identify the most potent coalitions, it is necessary to understand the political, economic and human power associated with each individual. A fatal weakness of traditional political organizing is that the need to collect and centralize this information for analysis poses an intolerable risk to privacy and personal security. Not only are there severe moral hazards in attempting to build comprehensive profiles of individuals for a political database, but in many jurisdictions this is expressly forbidden by law. Laws rightly protect individuals from being profiled by political organizations seeking to recruit them, or by corporations seeking to market goods to them. With Platformer, however, our model has a different center of gravity. Instead of locating the power to identify, categorize, and organize in an entity like a political party, we center everything on the individual. In the same way, so-called “personal data” must remain the inviolable property of the individual. Yet, we want the individual to be able to allow the system to locate him or her within constituencies using this data—without revealing the data or the user’s identity to other users. The precise means of accomplishing this are one of the areas of research, but an outline of the technical concepts already exists and builds on proven approaches to security and privacy. The key with Platformer is that the system provides verifiable guarantees of the safety of data, and allows an individual to safely build up a very deep profile (what would amount to a corporate marketer’s dream) containing a comprehensive set of information about an individual encompassing everything from income to reading habits to type of abode to usual diet and so on. In today’s world, the prospect of aggregating so much data may seem frightening, but Platformer will provide such a guarantee of security, and will allow individuals to put this data to such spectacular uses, that concerns will be allayed. So to extend our fictional example above, let us say that the system has identified a significant 7

http://platformer.org

Illustration 7: Platforms compared with one another. Similarity between two platforms is indicated by line thickness.

number of people concerned with the Oregon bird situation who also buy gas regularly from Lukoil, a Russian oil company. Similarly, a large number of the individuals focused on the Russian police issue have indicated that they have office jobs involving appreciable quantities of paper usage. So the system proposes to the first group to pledge a monthlong boycott of Lukoil stations unless that company lobbies the Russian government for reform of the police system, and to the second group it proposes a pledge to push their employers to boycott the paper company that is threatening the endangered bird unless the paper company stops obstructing legislation to protect the bird. While this example may seem strained, hopefully the concept is clear. In fact, the very difficulty that any of us may have in constructing an example such as this is, paradoxically, an illustration of the true power of this approach. Platformer is designed to discover correspondences such as these that could never be reasoned out by mere speculation. Building on decades of research in areas like collaborative filtering and data mining, Platformer: Introductory White Paper (Draft)

Platformer is able to leap ahead and produce truly stunning results because of the power of deep profiles. 3.7. Campaign Architects One remaining key concept in the system is that of a special kind of user role. While many people will use applications that tie in to Platformer without necessarily focusing on the fact that Platformer is in action, much like most of us use the web without thinking of its mechanics, there will be some people who take a more active role in building resources using Platformer. Among these people will be campaign architects. In a scenario such as the bird/police example, the work of identifying coalitions and all the steps leading up to it will be done by Platformer. But there will be certain steps that need human intervention. In general, these steps will involve composing some communication to a targeted entity, such as the gas company or the paper company from the previous example. A campaign architect will need to write a letter addressed to, say, the paper company, explaining the request that the 8

http://platformer.org

coalition is putting to the company. Platformer will provide the statistical data to back up the claims, including the number of individuals who have made pledges, certification that these individuals are actually valid customers of the company, and so on. The campaign architect will be required by Platformer to submit any communications to the coalition for review before the coalition is prompted to initiate action on a pledge.

4.4. Graduated certification of identity A pledge will carry more weight the greater the certainty is that the individuals making the pledge are real people, and are not “sock puppets”. Even though Platformer must hide the identities of constituents, it must also be reliable in its characterizations of the individuals who form platforms and make pledges. Rather than hard-coding a single model of identity verification, Platformer should have an open model that allows different forms of identity verification to be ranked against one another. Since some forms of identity verification are not available to all people, and since new models occasionally emerge, we should be able to consider the relative certainty that any given individual is a unique human being, compared with all others.

4. Technical Overview of System

What follows is a sketch of the necessary technical characteristics of the system that will be necessary to provide the functionality set out above. 4.1. Fully distributed Platformer must be a fully distributed system. This means that there is no “central server” that houses data or applications. All data and processing is distributed across an open network to which anyone may add a node. The main reasons for this are dual: first, it avoids any concern that a single entity might use data for nefarious purposes; and second, it eliminates reliance on any organization staying “in business” to keep the system operating. Just as the Internet and the web do not depend on any single company to continue operating, so will the “political Internet” enabled by Platformer exist independently of any organization.

4.5.

An open platform for application development Platformer should provide protocols and tools that allow an open range of applications to be created. Like the core technologies that power the web, such as HTML and HTTP, Platformer should avoid limiting the creative potential of developers. Just as the initial creators of web technologies could not anticipate the types of content and applications that have been built on the web, neither can we anticipate the sorts of dynamic systems that may be built using Platformer.

4.2. Open source (free software) All of the protocols and software that make up Platformer will be made available under licenses that permit and preserve free distribution and modification, and which encourage collaborative development. Not only is this a proven model for building solid, reliable software, but it also offers any skeptics the possibility to audit code and be satisfied with the robustness of the security and privacy models, the filtering and clustering algorithms in use, and so forth.

4.6. Re-implementable The reference implementations of Platformer should serve as examples, but not necessarily as the sole implementations of the technology. Eventually, a full specification of all core components, not just the protocols, should be available and should assist anyone who wants to rebuild some or all of the components using different programming languages or other technologies. 4.7. Verifiable All parts of Platformer must be able to verify one another. Without going into excessive detail, suffice it to say that this is a key aspect of the technological design and is especially critical due to the fully distributed nature of the system. Any Platformer node must be able to query another node and verify that that node is producing true and faithful calculations that conform to the Platformer protocols. Again, this system builds on established research while introducing some novel approaches as well.

4.3. Secure and private As has already been emphasized, Platformer must provide an ironclad guarantee of security for private data. Many people in the world live in regimes that mete out harsh punishments for advocating unsanctioned views. Platformer should make it possible to express opinions and commit to changes without fear of identification or reprisal. It will be necessary to guarantee that no personal data can be retrieved without the active, willing participation of the owner of that data. Despite the fact that data will be stored in a distributed, dynamic fashion across the network of Platformer nodes, no one should ever be able to reassemble private data that does not belong to them. What is more, an individual should be able to wipe his or her personal data from the system at any time and leave no residue.

Platformer: Introductory White Paper (Draft)

4.8. International, multicultural Constituencies exist across all national, cultural, socioeconomic and linguistic lines. Platformer must, from its inception, target full availability and usability in any language, and relevance for people of any means, including individuals who do not have personal, private access to computers.

9

http://platformer.org

describing the ideas of the project and the state of its implementation, and to discuss open questions and issues with thinkers in political science, activist organizing, and other relevant fields.

5. Technology Implementation Plan

The technology implementation plan has three components. Most likely, all three will be engaged in parallel, and released together, initially, simultaneously. Work on all components will proceed in public and will be available for inspection on open source code and document repositories.

7. Expected Impact

As may be gathered from the preceding discussion, the expected impact of Platformer is rather large. If the ideas described here can be implemented, we should expect to see something of a sudden revolution in how progressive aims can be addressed, especially by groups of individuals— constituencies—who previously went under- or un-represented. We should expect to see surprising actions producing surprising results—people in one country who boycott a company resulting in dramatic changes in government in another country. And we should expect to see a new birth of engagement by numerous people who never thought they could personally make a difference in the world. By putting the means of control of their own political, economic and human power back in the hands of individuals, we may hope to empower humanity to demand and achieve justice, equality, liberty and a higher standard of living across the world.

5.1. Protocol Specification A set of documents will be produced describing how different components of the Platformer system must communicate with each other. This will include the structure of messages to be used and the requirements for the transport mechanism by which they will be delivered, as well as the “contracts” that different components must fulfill with one another. This document will be drafted, reviewed and maintained following a process similar 5 to that employed by the W3C. 5.2. Node Reference Implementation A “reference implementation” of a Platformer node will be coded in a language or languages chosen both for suitability in producing a working implementation, as well as for their ability to provide a clear example for study for those wishing to extend the system or reimplement it on other platforms. The reference implementation of the Platformer node software should be robust enough to serve as a real-life program that can be deployed across the Internet and enable the processes described earlier in this document. 5.3. Application Reference Implementations Two or more reference implementations of Platformer-integrated applications will also be coded and released. Examples of such applications might be: Facebook applications, modules for content management systems such as Drupal, Ruby on Rails applications, mobile phone apps, etc. The focus of these reference implementations will be to demonstrate to developers how to use Platformer and how to develop end user applications for it.

6. Community Development Plan

A vital part of this project, as with any open source project, is building a healthy and active community. From the beginning, documents such as this will be published, and all in-process documents and code will be maintained on public resources that encourage collaboration. Using popular discussion tools and “gathering places” on the web, we will grow as a worldwide community of developers and users who seek to realize and further develop the vision of the Platformer project. Community development will also include personal visits to locations around the world where there is interest in use and/or development of the technology. It will be crucial to visit conferences and conventions, to give talks and presentations 5

World Wide Web Consortium (http://w3.org)

Platformer: Introductory White Paper (Draft)

10

http://platformer.org