Internet-based, AI-supported, Large-Scale Consensus-Building
To carry out grassroots governance, employing the CSD project approach for larger-scale projects at the regional, state or national levels, the numbers of stakeholders will be too large to contact and engage by traditional means. In this situation an Internet-based AI-supported mass engagement tool is envisioned that enables consensus to emerge out of the concerns and creative ideas of the large mass of stakeholders. For purposes of this discussion, that tool is called a Consensus Engine. Its purpose is to enable the creative inputs from individual members of the public, no matter how numerous, to be accepted and processed in a meaningful manner so their influence is felt in the final product.
Supplying the Missing Internet Function
The Consensus Engine performs a function, necessary for support of grass-roots governance, which is currently missing from the world of internet social media. The internet currently acts like a reverse funnel. A message from a single source goes in the small end of the funnel and comes out the large end. The large end may be as small as a single receiver, as with an email from one individual to another. Or the large end of the funnel may be as large as the whole world, as in a Tweet from Donald Trump. This is good enough to organize a mass movement originating from some elite source, but is not effective for tapping into the collective knowledge base and creativity of the grass roots, as demonstrated by the failure of Occupy Wall Street to produce any tangible results. We need a forward-funnel function in the social media world that can take in a large mass of varied input and digest it into a conceptually concise summary.
A Challenge to Silicon Valley
Creating the forward-funnel internet function, taking in a wide range of inputs at the large end of the funnel and processing them into a single or small number of consensus statements, can be done with current Artificial Intelligence technology, perhaps using techniques of deep learning, ontologies, and natural language processing. Watson as created by IBM is one example. Watson demonstrated superiority over humans in the games of chess and Jeopardy. It is now being developed as an assistant to professionals in medicine and other disciplines. It not only delivers an answer to a complex query but supplies the supporting rationale. The smart speakers we have in our homes that respond in human-like manner to our commands rely on Artificial Intelligence software in the cloud to interpret our intentions and select a response. Big data processing methods are used in commerce and government to extract behavior patterns for marketing and other aims. Similar capability exists in the national intelligence world for making sense of intercepted communications that bear on national security. In Silicon Valley the capability to create the needed Consensus Engine is at hand. Who will accept the challenge to make this tool for grassroots governance a reality?
Managing Complexity at the GrassRoots
In other branches of this site the four-question CSD Solution Search Path is described. How does this process roadmap deal with complexity in grassroots governance? In general, the process is open and thorough, so nothing of importance is overlooked or hidden. It accommodates powerful complexity-management processes and tools. Each step is taken with full stakeholder concurrence.
More specifically, the first process question in the Solution Search Path ensures that every stakeholder concern is recognized and honored, and the second question converts that input to an exhaustive set of criteria for judging candidate solutions. Thus, with the stage set to address the issue in its fullness, the third question invites expert determination of the methods to analyze the problem-system and synthesize a solution-system that covers all considerations. This includes proven methodologies to manage the myriad of complex interactions among the various parts of the problem and its environment. The fourth question invites expert skill in setting up evaluation that ensures, to the best of human ability, ensures that the selected solution will actually do the required job in all its complexity, and avoid adverse unintended consequences.
In general, the complexity-management process starts with breaking down a complex situation into simpler and simpler parts until understanding of the various parts is possible. At each step in the breakdown from whole system to ultimate detail, the functions necessary at that level to do the job, and the form elements that execute the functions, are identified. Then a solution is developed by assembling parts into higher and higher levels of complexity. At each level in the build-up process, the ability to perform all the necessary functions and avoid unnecessary and undesired functions is verified, either by analysis and modeling, or by actual physical testing. Rigorous information management techniques are applied to ensure that all the parts match up and necessary connections are established.
Stakeholder input informs experts as they create and evaluate possible solutions. As intermediate products of solution-discovery emerge from this expert team, those products are validated for stakeholder concurrence. This continues until the final solution is selected, again with stakeholder concurrence.
This solution-discovery process advances in a style that builds consensus incrementally as the four questions of the Solution Search Path are addressed in turn in an iterative manner. Currently, stakeholder engagement can be accomplished through traditional means: public meetings, door-to-door canvassing, questionnaires and polls. However, with the advent of powerful Artificial Intelligence techniques, it becomes possible for stakeholder engagement to extend to the entire population.
How It Works
These two elements, process roadmap and Consensus Engine, work together under the third element, transformative leadership. To address an issue for resolution at the grass roots, a CSD project is set up with the four-step Solution Search Path at its core. Travelling along the Solution Search Path, the project leaders pose the questions to the stakeholder base that are needed to execute each step in the path The Consensus Engine accepts answers submitted intentionally by stakeholders and also scans the media for relevant input. The questions are posed in such a way that the response is not just the direct answer, but also includes the stakeholder’s rationale supporting that answer. The rationale is important because it supports a process by which consensus emerges, not as the result of sheer numbers of similar answers, but on the quality of the reasoning behind the answer.
From all that input the AI tool produces one or more provisional statements of the apparent consensus along with a summary of the rationale leading to that conclusion. The result is broadcast to the stakeholders with the question “Is this what you really mean?” This broadcast statement of provisional consensus with rationale stimulates another wave of responses from the stakeholders, which is again processed for consensus content. Presumably and hopefully, this iterative process will converge on a consensus acceptable to most stakeholders, perhaps with other diverse viewpoints attached.
For any issue of concern, the four basic questions of the Solution Search Path are asked and processed by the Consensus Engine in an iterative manner until stakeholder consensus is reached. The result is to set up a Solution-Discovery Project for the issue. The answers to the first two questions of the four-step Solution Search Path define the stakeholders’ needs and concerns, and establish the stakeholders’ vision of a successful resolution of the issue. The answers to the third and fourth questions of the Solution Search Path define how solution-discovery will be done by a suitable team of experts, again to the satisfaction of the stakeholders. The Consensus Engine is also used to set up, or at least ratify, an organizational structure and roster of accepted experts, constituting a core team to carry out the Solution-Discovery Project. The Consensus Engine then serves as the communication link that keeps the stakeholder base engaged with the solution-discovery process being carried out by the core team.
Flaws in the Process
The MuddleBuster can see some (there may be more) shortcomings in this process.
What is to prevent the process from converging on the most banal form of group-think?
This consensus process depends not on the volume of support for a particular idea, but on the quality of that idea. A good idea is a good idea, whether supported by a million people or just one. Its value depends on the rationale that backs it up. Furthermore, there may be a way to give weight to knowledgeable expert review of the ideas submitted in order to screen out the losers..
How would the unique and highly creative solution from the fringes of the concept world be recognized and adopted?
The AI tool would include the ability to identify and single out promising fringe ideas for subsequent review and amplification. This feature may not be infallible, but helpful.
What would prevent some organized interest group from hijacking the process by flooding the system with input supporting their own position and crowding out other viewpoints?
Again, this consensus process depends, not on the number of supporters of an idea, but on its quality. Further, if bots or hoards of human repeaters of a fixed idea were making identical or nearly identical submissions, the consensus engine could pick that up and reject it.
Clearly, in the spirit of grassroots governance, the development of this concept can flourish as a crowd-sourced project. Constructive critique and suggested solutions are encouraged.
An extension of Constitutional process
As a tool for grass-roots governance, this process is superimposed on the existing formal governance that has been set up under our Constitution without displacing that setup. Its function is advisory, but its power lies in the breadth of consensus support and the logical strength of the answers it delivers. As such it would be difficult for elected leaders to ignore. Furthermore, it is a return to the democratic principles of our founding fathers, as contrasted with the extreme domination of current politics by the elites of wealth and political power.
Fortunately, launching this new paradigm does not require the participation, or even the approval, of any element of the existing power structure. In that sense it is truly revolutionary. All that is needed is a small independent group with entrepreneurial spirit, visionary thinking, and access to the necessary AI technology. This group creates the first version of a solution-discovery project template based on the collaborative model with accompanying Internet/AI support platform, sets it free, and watches it go viral.