Can Swarm Intelligence Save Democracy?

From Agile Feedback Loops to Real-Time Swarm Intelligence. Swarm intelligence is very different from group intelligence and it is logically the final evolutionary step of the Agile Mindset. In contrast to the delayed feedback loops of individual and group decision-making, swarm decision-making produces real-time (or near-real-time) feedback loops within a clearly defined “decision-space,” which can be comprised of any number of structured questions and answers. All these questions and answers can be individually sequenced and collectively designed to arrive at a definitive, consensus-based decision to produce an optimal final outcome.1

 

Group-Based Decisions Can Instigate Mob Tyranny. In group decision-making, a majority or plurality vote determines what the group thinks. This typically leads to better outcomes than individual decision-making by “tapping into the wisdom of the crowd,” but there is still a high probability of a Tragedy of the Commons occurring. Additionally, group decision-making processes can easily degenerate into “group-think” as more charismatic and dominant personalities often influence the emotions of other members of the group. This creates the risk of mob passions hijacking the decision-making process, which can lead to various forms of tyranny over minorities.

 

Optimal Outcomes Depend on Good Faith and Egalitarian Choice Architecture. Of course, any opinion poll, voting ballot, or sequence of questions can be engineered to create varying degrees of bias for or against a given candidate or agenda. (I’ve addressed the concept of systemic bias previously. No human-designed system is completely free of bias so the goal should be to eliminate conscious and deliberate bias.) The point here is to employ an egalitarian choice architecture and design a decision space of questions and answers in good faith so that the members of the group can all contribute equally to the final outcome. Some brief definitions are in order:

“Egalitarian choice architecture” in this case means the structure of the voting ballot, opinion poll, survey, sequence of questions, etc. is designed and presented fairly without any deliberate bias for or against any particular outcome or ideological agenda.

“Decision space” means that the answers to the questions are presented in a clear and structured format so that all members of the group can choose an answer(s) without unnecessary confusion. Bubble sheets, multiple-choice, fill-in-the-blank, are all examples of a “decision space” that each member of a group can interact within as they decide which answers to choose. However, a decision space can also be represented as a computer screen with the choices presented on opposing sides of the screen. In this case, a push-pull mechanism can be created to allow any number of group members to negotiate in real-time until a final, consensus-based choice or solution is agreed upon by a plurality of the group.

“Good faith” in this case means the architects of the questionnaire have societal legitimacy and a widely recognized reputation for fairness so that the members of the group will believe they are not being manipulated by the architects.

“Contribute equally” means no single member of the voting population can dominate the decision-making process due to their charisma, financial power, political power, military power, intimidating physical stature, elite class in society, gender, age, technological advantages, nor any other attribute that might give them unearned influence over the outcome of the decision-making and policy-making processes.

 

Group Pluralities Are Not the Same as Swarm Pluralities. The “plurality” of a swarm-based electoral voting system described above is derived from a very different physical and emotional process than the plurality of group-based electoral voting systems today. Existing voting systems are exposed to numerous forms of bias and propaganda all throughout a voter’s physical and emotional voting experience, which undermines the incentive for politicians and voters to seek and demand rational, good-faith consensus. This leads to pluralities that are based more on the most recent propaganda soundbites and attack ads than on actual truth and rigorous policy analysis, which favors the wealthiest special interest groups who have the largest media, marketing, and PR budgets. This creates a very low signal-to-noise ratio and constant voter confusion, which makes it difficult for many voters to know which politicians are actively thwarting Democracy and which ones are actively trying to implement meaningful legislative reforms.

 

Swarm-Based Voting Encourages Rational, Efficient Consensus. In contrast to a group-based voting system, a plurality derived from a swarm-based voting system creates several physical and emotional incentives for voters and politicians to seek rational consensus as efficiently as possible while significantly reducing the impact of unnecessary and biased influences on voter behavior. This results in a voting system that is presented as an authentic real-time intention-measuring interface, which produces a real-time feedback loop that could substantially reduce the unearned influence of financial power, political power, media power, intimidation, etc., by imposing an objectively designed structure on both the choice architecture and the decision space of the voting system.

 

Swarm Intelligence in Nature. Swarm-based decision-making results in a real-time exchange of information and a real-time negotiation process among members within a large group. One of the most famous and amusing examples of this process in nature is the Waggle Dance, which honeybees perform when transmitting knowledge to one another. Each bee by itself is relatively unintelligent, but when each bee transmits and distributes its discrete knowledge to the entire bee colony through the Waggle Dance, the colony can make very complex existential decisions. For example: Where should the bee colony build their hive each season? How might weather conditions hurt the hive? How close should the hive be to other animals and structures? How to protect the queen bee most effectively at all times? These are questions that most humans would find difficult to answer; yet, honeybees negotiate all these and other discrete decisions through the real-time Waggle Dance with over 80% accuracy.

 

A Real-Time Informal Negotiation. The real-time “negotiation” within a swarm-based voting system is not a formal negotiation like two people facing each other at a proverbial negotiating table. The negotiation can be structured simply as a series of questions presented to a group of any size to illicit individual answers in the form of a real-time push-pull debate within a structured decision space. The most interesting example of a swarm-based voting interface today is the deceptively simple UNU A.I. online game. When an UNU A.I. player engages the game, they experience the real-time feedback loop as they negotiate with dozens, hundreds, thousands or potentially millions of other players simultaneously as each player attempts to “push the puck” to their chosen answer for each question.2

 

Real-Time, Consensus-Based Voting Would Deradicalize Voter Behavior. Given the consensus-forcing physics of the swarm-based voting process, a plurality coalesces around less divisive and more rational choices. This is a natural byproduct of the real-time pushing and pulling negotiation process between all the members of the group, which leads to outcomes that more closely reflect the true aggregate desire and intent of the group rather than the desire and intent of special interest groups with various forms of unearned influence (money, media influence, intimidation, political connections, etc.) If swarm intelligence could be applied to democratic elections in a secure, tamper-resistant, user-friendly interface, it would substantially reduce the power that special interest groups currently have over the political system, which would reduce their ability to drive their constituents to ever-more extreme positions along the ideological continuum.

 

System Security and Integrity: Attack Vectors. Of course, any computerized voting system must be protected from various forms of manipulation, back-doors, trojans, rootkits, corporate espionage, social engineering, hacking, denial-of-service attacks, etc. A computerized swarm-based voting system is no different in this regard. The most important question would be: How do we prevent well-funded, politically powerful, and technologically sophisticated special interest groups from manipulating the swarm-based voting process? Theoretically, it could be possible to infiltrate a computerized swarm-based voting system with automated agents that push the puck toward choices that support their special interest agenda like an army of robot voters. (That might describe some hyper-partisan human voters, too!)

 

Identity Verification and Vote Validation. Like any voting system, a swarm-based voting system would need a robust identity verification mechanism to limit the voting to “one citizen, one vote.” Technically, this could even be achieved by using blockchain-based voter verification and vote validation processes to simplify the infrastructure management and significantly increase the incorruptibility of the election results. All these technologies already exist and have been proven in high-stakes environments. That means the real obstacle to implementing any kind of new national election voting system will always be political, not legal or technological.

 

Can Swarm-Based Intelligence Save Democracy? Swarm-based intelligence by itself may not completely rescue American Democracy from the iron grip of self-serving politicians and their special interest group patrons, but it could incrementally improve the U.S. national electoral voting system. Unfortunately, the political obstacles to implementing a swarm-based voting system will inevitably block its adoption until the American people start sending enough authentic leaders to Washington who have the political and financial independence to resist the temptation to sell their votes for cash. That’s what The Platform is intended to accomplish.

 


Notes:
1. I define the “optimal outcome” as follows: “a domestic or foreign policy that satisfies the largest population of U.S. Citizens, within the shortest amount of time, at the least cost to American taxpayers, and which is sustainable over the longest period of time.”

2. The UNU A.I. game is currently in the beta phase and it’s not clear if the UNU developers ever intend to adapt the game to any kind of serious voting system. However, after playing the game myself a few times and thinking about how the principles of the game could be adapted to more serious voting systems, it’s clear to me that some of UNU A.I.’s underlying principles could substantially reduce many of the problems that plague the voting system in the U.S. today.

3. The UNU A.I. game has nothing to do with The Platform, although I might consider discussing with the UNU team a potential collaboration sometime in the future. Additionally, the UNU game could theoretically reduce only some of the weaknesses in the existing group-based national election voting system. However, it would do nothing to address the distorted incentives that congressional politicians have to use their legislative and regulatory power to benefit special interest groups in exchange for campaign financing that promotes their personal political careers and secures their post-government lobbyist jobs. Those are the problems that The Platform is designed to resolve.

4. If true Democracy scares you, see Are You Afraid of True Democracy?



About Ferris Eanfar

Ferris has over 20 years of experience in the field of International Political Economy, including leadership positions within technical, business, financial, media and government intelligence environments. If you want to learn more about Ferris, please visit the About Ferris page.