Transcript (more or less) from The evolution of democracy: Explaining Trump, Brexit and the Colombia peace deal, a keynote at Inteligencia Colectiva para la Democracia’ in Madrid, 2016.
My name is Heather Marsh. I am a writer and a programmer and I have been studying and experimenting with both local activism and methods of mass communication and collaboration for many years now. From 2010 to 2012 I was the administrator and editor in chief of the Wikileaks news site Wikileaks Central where I experimented with creating knowledge repositories, tying that information to things that were happening in the news and creating action based on that information. News without action is just voyeurism and action without information creates a very easily manipulated public, so I was trying to bring the two together in one place. In 2012 I concentrated more social media collaboration and I wrote a first book called Binding Chaos about all the problems I had seen while working with Wikileaks, Occupy, Anonymous, M15 and many other mass movements in recent years. We all seemed to be coming up against the same issues with hierarchy, direct democracy, consensus and collaboration. As I kept working on various projects it became apparent that we as societies had been butting heads for millenia on these same issues which really come down to trying to create a balance between autonomy, diversity and society, which is the title of my next book. And along the way I have been thinking of what tools we would require to help us achieve this balance, and the primary one I have been working on is a universal database and trust network called Getgee.
So today what I would like to talk about are some ideas from Binding Chaos and a little bit from Autonomy, Diversity, Society which will hopefully help when we are thinking about creating products for mass communication and mass collaboration. The focus is on creating a balance between personal autonomy of those doing the work, diversity of ideas and solutions and allowing the participation of the whole society.
How many people here have heard of a technological singularity? A technological singularity is something IT people and science fiction writers have liked to talk about for years. The idea is that we will reach a point, or have already reached a point, where technology is beyond the scope of human understanding and artificial intelligence will be programming itself, in a Skynet sort of world. It’s funny, people have been talking about this for years, but not many have noticed or acknowledged that we have instead reached a completely different type of singularity which is a societal singularity. We have reached a point where no one can understand every aspect of society which affects them. If you go back in history, crafts people could know everything there is to know about their jobs and people could know everything that went on in their villages but this is just not true any more. And even our villages and neighbourhoods are not autonomous, they are all connected now at some level with all the other communities in the whole world, even uncontacted tribes.
We need to collaborate with others not only to develop tools but even just to understand the news. We have to put our faith in other people and believe in what they tell us or trust in their skill to create their components if we are building a product. As programmers we have been used to working as an ecosystem like this for years, we always have to incorporate other people’s work and their bugs into our own, but this has spread to almost every aspect of public life. And this is one of the biggest challenges in creating tools for democracy. We can’t have real direct democracy or self governance any more because none of us can understand every aspect of everything. We need to rely on collaboration instead and this is going to require a completely different set of rules than we have used in the past. We need more than simple referendums and voting to govern ourselves this way. We need to somehow create nuanced and detailed information we can trust and we need to coordinate goals with people we will never speak to.
There are two main areas to talk about which are idea based collaboration and action based collaboration. The challenge in a societal singularity is how to allow all people to participate and communicate but still be able to filter signal from noise and how to allow people democratic choice but still retain worker autonomy.
First of all, let’s look at mass action based collaboration.
For action based tasks, the model that has become almost ubiquitous is the competitive hierarchical model. Most of us are all too familiar with this model. The typical response to a situation which requires an action is to create a noun, in the form of an organization, government body, or an official person and the focus is always on the organization and the personalities involved instead of the action. The hierarchy creates what I call personality based systems, as opposed to idea or action based systems. A new idea in a personality based system remains completely bound to the owner until it is legally transferred to another owner. All contributors work for the owner, not the idea, and you have to wait on one specific person for approval or direction at each level so there are bottle necks everywhere.
Most workers do not enjoy hierarchical systems as they lose autonomy, mastery and creative control over their own work, they just become an instrument under somebody else’s direction. The orders come from the top down and so there is very little diversity of ideas and we lose all the talent and ideas downstream. Because it is a closed system, collaboration between people does not happen unless they are hired by the same project. Competition is the opposite of mass collaboration. It’s really people working against each other, not together. So there is no autonomy, no diversity and no society.
The alternative to competition has traditionally been cooperation. Cooperative groups try to replace the top down hierarchy with a group consensus driven system which allows diversity of opinion at the top.
This is most effective only in groups of two to eight people. For groups larger than 25, cooperation is extremely slow. It is still a personality based system. An idea in a cooperative must be approved by the entire group, both on initiation and at each stage of development. The majority of energy and resources are spent on communication, persuasion, and personality management, and a power struggle can derail the whole project.
It can be dominated by extroverted personalities who make decisions to control the work of others and are very justly resented by those doing the actual work. Cooperatives frequently use consensus or votes to make decisions for the entire group. These methods may not produce the best results, particularly in large groups, as many people may not understand the work if they are not actually doing it and they may demand things they would never be willing to do themselves. The feeling of the workers at the bottom is no different whether there is a horizontal or a hierarchical structure making the decisions, the workers still have no personal autonomy.
Both competitive and cooperative projects will die if the group that runs the project leaves and both will attract or repel contributors based on the personalities of the existing group. Both are hierarchical systems where individuals need to seek permission to contribute. Both focus on the authority of personalities to approve a decision instead of focusing on the idea or action itself. So we have a society, at least within the group, but not without, but we still have no autonomy and because of the need to reach consensus there is also no diversity of product.
This isn’t in any way to say that cooperative and consensus driven systems are bad. They are actually the most comfortable way of working in small groups who know each other and have similar styles and share a goal, but they are very difficult to scale. As soon as you have a very large group of people with opposing viewpoints and personalities that don’t mesh, it is very difficult to get anything done.
I use stigmergy to describe a method of action based collaboration that is suitable for mass movements. I didn’t make up this awful word by the way, it is lifted from biology where it describes indirect communication and collaboration among ants and termites and various other creatures. In human movements, it allows diversity of methods and autonomy for workers while still putting the ultimate authority of choice with the whole society, to try to achieve that balance we talked of earlier. It is neither competitive nor cooperative. It is action based collaboration instead of personality based.
A system is stigmergic if
– it follows one goal
– it is completely transparent
– it is open to everyone to participate, at least within the user group
– the output is free for anyone to use and improve on
Stigmergy gives people autonomy over their own work. With stigmergy, an initial idea is freely given, and the project is driven by the idea, not by a personality or group of personalities. So no one needs permission, like they would in a competitive system, or consensus like they would in a cooperative, to initiate a project. There is no need to discuss or vote on the idea. If an idea is exciting or necessary it will attract interest and the interest attracted will be from people willing to contribute so those with more involvement in the idea will automatically have greater influence through their contributions.
There are no official authorities but the power of the user group still exists in the ability to accept or reject the work.
Workers are free to create regardless of acceptance or rejection. Drastically innovative ideas almost never receive instant mainstream acceptance so leaving control of work to group consensus only cripples innovation. When we allow anyone to contribute we also have a great diversity of talent and people can step up to further the goal in ways the originator never imagined.
So here we have full autonomy and diversity but the entire society still has the ultimate choice.
Where is stigmergy? We have always had stigmergy in our social lives and it has been behind most mass movements that have had any success. You can see it wherever groups of diverse people who do not belong to any formal organization or have any formal communication with each other are all working together to carry out a goal.
If you look at something like the civil rights movement in the United States, that is a multi-generational movement of so many people and so many different methods and everyone who has contributed, whether they are groups or individuals, has decided for themselves how they can be most effective. If this stigmergy chart was for that movement, that big group can be the Million Man March, the square is Malcolm X and his followers, the heart can be MLK, and between them all by themselves is Ruby Bridges or Rosa Parks, none of them had to communicate or come to consensus but they are all trying for the same goal and are more or less aware of each other’s activity. The US still hasn’t reached that goal so they go through periods of great upheaval followed by periods of more calm working, depending on whether an event sparks more action or something blocks progress for a while.
A stigmergic movement will continue as long as the goal is not reached and people still share it, even if it dies down or goes dormant for a bit. That is the advantage to an idea based system over a personality based one, you can’t kill an idea.
Or another stigmergic idea is freedom of information. This has everyone from the free software movement, creative commons and similar open copyright groups, Sci-Hub which liberates scientific papers, other filesharing sites, Wikipedia, even the Internet itself in its original inception might be considered a node in that stigmergic movement.
What keeps these movements from burning out, like so many do in the massive assemblies, is the fact that they are not spending all their energy communicating except in small groups and they are following one clear idea.
It is not often you find one organization or group that is purely stigmergic, but Anonymous is one. This is why they say they are not an organization or group. They usually say they are an idea, but they aren’t really one idea either, they are a method of mass collaboration and the method is stigmergy. That method allows everyone to follow whatever ideas they choose, in groups or individually with perfect autonomy. Anonymous never tries to reach consensus. Anonymous is not unanimous. And there is no organization you can order to do something, Anonymous is also not your personal army. You have to just put an idea out and see who follows it.
It may seem difficult to figure out how stigmergy can be used in a corporate setting where everything is set up around organizations and official people, but it helps if we remember that each of those organizations, no matter how they are organized internally, can be a node under a stigmergic idea. I am often asked if Wikipedia is a good example of stigmergy and no, it is not. Wikipedia is a cooperative. You may contribute work without asking anyone but your work can be thrown out and you can be locked out of contributing, or the topic locked, and there is a definite personality based hierarchy and a need to reach consensus around one final product. There is no diversity of product tolerated and there isn’t any real autonomy either.
But Wikipedia is still one of many nodes under the idea of Freedom of information because what they produce is completely free for anyone else to use or modify. I said earlier that Anonymous was stigmergic but Anonymous very frequently works with other people like news or human rights organizations or other hacking collectives such as Redhack who are themselves internally communist. It doesn’t matter what the internal organization of each node is as long as they are all following the same idea and their work is available for everyone else to use.
So the same methods can be used for corporate work. The key is for corporate style organizations to recognize what stigmergic ecosystem they are a part of and follow the guidelines to make their work contribute smoothly to that idea. One place where stigmergic development has really taken off is in the IT industry because free software has meant that the output is available for everyone to use and improve or modify. If we look at one stigmergic idea: We need better web development tools. If we had left this to Google, and Google had been acting like their competitive corporate selves, we would have just had the Angular framework, and progress would be Angular 2.0. And we do have Angular 2.0, but we also have Facebook’s React, Ember and many others. As long as the user group has not reached consensus over what tools we want for full stack development we have many contributors creating different frameworks for us.
When we start to reach wide consensus in some area, like yes, we don’t want any more black and purple websites with green sparkles and the vast majority of us are going to create websites that are very uniform, we start to see more and more conformity around standards like Twitter’s bootstrap styles but as soon as someone has a very divergent idea that people find interesting again, like Google’s material design, many people will start hacking on it and trying to create different solutions again. The same periodic upheavals of innovation and change are apparent here as in the social movements driven by stigmergy. In this case HTML5 and ES6 stimulated a rush in web development tools in general.
Of course this example of corporate stigmergy has some major issues., first in who is getting paid and who is not. Google employees are and free software programmers frequently are not. And even with free software, when you have players like Facebook and Google and Twitter it is going to be a bit hard for anyone without their development team budget and user groups to compete so it is not a level playing field for all to participate, but as long as the code is commons property we can have consensus without monopoly which is a huge improvement. It is starting to approach stigmergic organization, just from the addition of this one change, of software that is free for anyone to use or modify. You can see this especially as you move away from the big corporations to the later development add ons, in all the diverse people writing packages and tools for React and Angular and the other frameworks.
This is better than academia and science manage. They are supposed to be stigmergic as well, the idea in science and academia is that everyone is supposed to publish and build off each others findings, but because they do not have open source and permissive copyright or even access, their work is frequently corporate IP property, and they don’t allow or reward outside contributions, they are very far from stigmergic and their progress is not nearly what it could be. If we look back at the principles of stigmergic organization, the last three of four points do not apply to science and academia so they both need to change if they are going to truly act as epistemic communities for us all. Which brings us to my next point which is about idea based collaboration.
We have stigmergy for action based collaboration which follows an idea, but what if we want to collaborate on ideas themselves, to build knowledge and find some most reliable facts? If we look at the 2010-2011 movements, like M15 and later Occupy and all the rest, they were fine with action based collaboration, especially when they used stigmergy, but they really struggled when it came to idea based collaboration, like setting goals. This is kind of important since without the ideas, you don’t have the action. Stigmergy follows ideas and information, so management of the ideas and information here is as important as management of personalities is in competitive or cooperative systems. If you think of this in a governance context, we won’t be electing personalities, we will be electing ideas.
To see what happens when an idea loses its clarity, or its idea has been co-opted, look at feminism. The civil rights movement in the US retained its clarity because it has set specific goals in each cycle whether that is to end slavery, end segregation, or end police violence. When feminism meant fighting for the vote and legal personhood it had a clear goal and was a stigmergic mass movement but second wave feminism allowed itself to become a noun instead of an action, its goals became very loose and because there wasn’t a clear goal it was used to advertise corporate product and promote prominent personalities, primarily from the United States, who felt they could speak for every woman in the world on every topic. A noun is not a stigmergic goal, a noun is an organization, so when feminism became a noun it stopped being a stigmergic movement and became a competitive, personality driven, organization which became completely divided, as is typical, these types of organizations do not scale.
The single biggest factor I’ve found for whether or not someone will participate in a stigmergic action is whether they are sure of the idea behind it. Not whether it affects them, or if its simple to grasp or easy or even safe to do. I have created many actions where the audience was completely removed from the people affected or where the action was dangerous or very difficult to understand or even initially believe. None of this mattered. All that mattered in whether the action was a success was whether people could be sure the goal is sound. And the easiest way for someone to prevent action is to sow doubt in the goal. That initial kernel that makes up the idea looks simple but it is everything. But finding the information we need for conviction in our goals is not easy.
If we think of a large population creating a knowledge repository stigmergicly, we have a picture of a bunch of ants sifting and sorting information and putting the best in a pile. And that’s probably how we thought we were going too do things on Twitter. But that’s not how expert knowledge, like the kind we have in a singularity, works and it’s why a bunch of people in a horizontal group can’t just do that. Especially a personality dominated horizontal group like Occupy or any direct democracy that starts from the premise of all voices and opinions being equal. This goes back to the idea behind a societal singularity, we can’t all be experts at everything and we don’t want to be either. We don’t have the time and we may not have the interest. We can’t keep berating voters for not spending all of their time studying everything that affects them, it’s impossible and it’s not fun either. We need to find a better solution.
I love this chart because it illustrates exactly why we need concentric circles in a democracy. This space between innovation and acceptance is where demagogues and gate keepers lie in wait to control information before it reaches the public. Like little trolls under the bridge. This is why we need knowledge bridges to replace the gatekeepers because most ideas can’t make it across this chasm on their own.
If you think of recent examples of elite working groups whose ideas were rejected by the wider society, like Brexit in the UK or the peace deal in Colombia, it was because of a failure of the working group to establish effective knowledge bridges between them and the public. The public did not see their viewpoints being heard and responded to and they did not see or understand or trust the decision making process, which gave demagogues on the outside of the process the ability to derail the acceptance of their recommendations. The people had information that was too difficult to audit themselves, and they had no faith in the people offering solutions. People in the UK said repeatedly they were sick of being lied to by the media and experts.
When people lose faith in those who are supposed to be their experts, like politicians, or those who are supposed to be their knowledge bridges like the media, they lose faith in any stigmergic goal these people present and they will block it, as I said earlier. The information from the opposition in both cases was certainly no better, nobody was offering a fully developed and audited plan for an alternative peace deal or a detailed plan to exit the EU, but even very poorly supported information and hyperbole is sufficient to overturn an idea that the public doesn’t trust. We can see this also in most elections, there are just demagogues and hyperbole on both sides, there is no process of reconciling any issues with the public or providing information people can rely on. In fact, the goal seems to be to deliberately confuse and immobilize the public and then just give them a binary vote in the hope they vote against the establishment which has lost so much trust. This is why both sides seem to only be interested in painting the other as the most corrupt establishment.
[On May 8 2017, former US Director of National intelligence James Clapper suggested the solution to the misinformation of the US 2016 election was to further fund USAID and spread more misinformation against US enemies. Those are exactly the actions that caused people to lose faith in information emanating from the US in the first place.]
You may have seen a lot of people blaming this current state of low information on social media and they are partially right but corporate media certainly doesn’t get a free pass either. Most of the tools we have to communicate simulate direct democracy and look for popular ideas, the most retweets and the most readers, but not only is popular not innovative or expert, the two are mutually exclusive. Popular ideas are riding the peak of the wave of socially acceptable opinion. They already appeal to the widest audience. They are not new ideas, by definition, and they aren’t at a level of elite expertise that is difficult for all but a few to understand also by definition. This is why we now elect politicians on the basis of their tweets and this is the secret for politicians like Donald Trump who speaks at a grade 3 level. The more easily understood and the less challenging your message, the wider your appeal will be so an age that amplifies the most popular information, as we do now, will be an age of demagogues.
It is counter intuitive to think popular ideas are what we need to give us the best information. If we need some specialized level of knowledge to explain something like Brexit or a peace deal or the issues in an election, or if we want those making the decisions to hear the voices that are seldom heard, that may expand their Overton windows and give us some fresh perspective, or represent a rare case that will cause their solution to break, amplifying the most popular ideas or people is the exact opposite of what we ought to be doing. And really this is what direct democracy, representative and liquid democracy do, so of course it is also what the tools for democracy have been doing.
We have tools that are very useful to find out what a population thinks and tools that are great for discussing things and coming to consensus, but we need to also go to where their opinions are formed. Opinions are based on information. We need to be able to find expertise and accurate and diverse information that we can trust before we form our opinions and long before we measure them.
Ideas need to be audited and promoted by those users qualified to understand them to allow diversity of ideas and prevent the process from being dominated by celebrities without the expertise required. But if we have an elite discussion group with only elite experts or ideas in it, we are at great risk of having an elite oligarchy based on control of information. This is what we have now. We don’t maintain the necessary balance between autonomy, diversity and society unless this quiet place to talk remains a fully associated part of the wider group. So to avoid a hierarchy and leave control with the entire user group, I use a structure I call concentric circles.
Concentric circles relate to sound amplification. In a concentric circle, people or ideas promoted to the center by their peer group receive greatest amplification and their findings will be audited, amplified and explained to the general public by outer circles. They are not hierarchical as they have no direct control over the actions of anyone. An epistemic community is a knowledge resource only, authority remains with the entire user group which provides a good incentive for the epistemic community to ensure transparency and knowledge bridges so their ideas are accepted. As in stigmergy, votes in a concentric group are frequently replaced by actions. If this little drop receives no amplification, it is just an idea that goes nowhere.
With knowledge bridges, you don’t have to have personal expertise on every aspect of society. All you have to do is have a transparent concentric circle that you can look at, you can see the activity, you can get feedback if necessary, and you can say yes, there are a lot of people auditing, there is a lot of discussion, I trust some of the people in these circles, I trust that they know what they are doing. Everyone can review the work of the experts both directly and through the review by their peers. Experts can also be created by the system itself as users develop knowledge and reputation and move towards the centre and you will find this happens increasingly if users lose trust, they will realize they need to start auditing this circle.
Communication should not be the full responsibility of the experts in the centre, which is where government initiatives like Brexit and the Colombia peace deal have failed. Ideas should be carried over expertise bridges by full transparency and user participation. The epistemic community in the centre should not need to protect themselves from demands or attacks from completely uninformed users or demagogues. The circles of expertise which promoted them to the centre should also verify and explain their findings to the outer circles. And, concerns and arguments from the user group should be carried back to the epistemic community if the user group finds the points valid. So the epistemic community can work without noise but still receive ongoing feedback from the users and acceptance is a process, not just a binary vote after the fact.
If this all sounds familiar, it’s because this is exactly what happens in open source communities.
In open source software, the code for each project is available for all to see. Even if the end user cannot understand the code, they can go to discussion groups or listen to programmers who have read and audited the code, they can read the bug reports. Any urgent bugs will be broadcast to the general population and amplified by media as we have seen many times. The people with the greater knowledge of the system will provide knowledge bridges for people at a more novice level and increasingly, that’s how people are learning to code. Good ideas from forum discussions can be read, possibly implemented by the developers as well. Transparency goes both ways.
Open source software with forums open to all are a perfect working example of fully transparent and audited systems of elite knowledge. While the decisions are made by the developers, review and acceptance or rejection of the software is the right of the user group. If the developers refuse to listen to the user group and another development team is willing to work on the project, the original code can be forked and modified to meet the user requirements. Which means you can only be attacked by another fully developed, open and transparent epistemic community which also must be audited by knowledge bridges. You can’t be attacked just by a demagogue and rhetoric, you can only be opposed by another working solution so the user group has a choice between two or more working solutions instead of simply rejection or acceptance. Which means we need the final most important point for concentric circles – the information is free for anyone to use or modify.
Intellectual property in a stigmergic system is like an ant that finds food but doesn’t leave any pheromones to tell the other ants. Or worse, actually blocks the other ants and that idea is so ridiculous I can’t even think of a stigmergic example of it. Ownership of ideas is in complete opposition to stigmergy which is to say it is in complete opposition to rapid progress, finding the best solutions and democracy.
So what we have been looking for here are methods of collaboration that bring us a balance between autonomy, diversity and society. We want to allow the maximum amount of autonomy to those doing the work so we can include all of their ideas and abilities. We want to allow as many diverse solutions as people are inspired to try for each problem. And then we want to allow the entire user group to easily make an informed choice of which solution is best for them as is their right in a democracy. So our methods are stigmergy, which we use for mass action and concentric circles which we use to audit, teach and amplify information.
The best part of stigmergic work and transparent concentric circles for knowledge is our work doesn’t get wasted. When you come to an event like this with a specific project, it is easy to feel as though you are in a competition where you are only associated with one project and your success or failure is tied to that one project and the group around it. But even if you organize your own team in a completely different form, if you follow these principles you will still be contributing to progress as a node in a stigmergic idea. For this two weeks the idea is: Let’s develop better tools for democracy. If you follow these principles, if other people are free to contribute to your project and you to theirs, if you add what you learn to the epistemic community of ideas and act as a knowledge bridge to those learning and most of all if your code is open and free, you will still be part of the community around this idea contributing to the goal we are all working towards. You will be part of the ecosystem.
Q and A not translated.