You have quite a diverse background – law, mathematics, philosophy. What motivated you to found a democracy technology company?
Our current system is terribly broken. It was invented three centuries ago, before there were even telegrams, not to mention digital technology, which has radically changed the complexity of the problems we face.
Many of the world’s problems are actually there because our democratic systems are unable to deal with them. In fact, they may be a side effect of our democratic systems. The polarisation. The inability to grapple with climate change. Our political systems can’t keep pace with the strength of the capitalist system. They’re outgunned and we need to really upgrade them if we’re going to take democratically grounded action to solve these problems. There’s no shortage of solutions. We just lack the political machinery to implement them.
I’ve always been interested in social activism and in politics particularly as a way of doing social change. I also kind of have a maths brain. So in law school I was initially looking at mediation theory and just wanted to know if we can measure the fairness of a contract. Is there a way to measure the fairness of an agreement mathematically? And that kind of turned into an independent project. I built a web platform to do two party arbitrations and then just kind of set it aside for many years. I was engaged in a lot of legal activism through Pivot Legal, my first organisation, but eventually realised systems of government were where the rules got made and you need to go to the root.
With Ethelo, you’ve chosen to focus on building consensus. How would you describe the challenge of consensus building in contemporary democracy and what differs in your approach?
In a normal or ‘traditional’ democratic decision making process, you generally start with a problem and someone proposes a solution. There might be processes for amending the solution and then it goes to a vote. Maybe there’s an opposing proposal or even two or three. The options are very different. Different people are aligned around different ones and voting is a kind of ‘winner take all’ process.
But when we’re talking about solving complex problems, it doesn’t necessarily have to be like that. There are many different components of a decision and various ways you can break it down into different options. What are the criteria by which the options are going to be evaluated? What are some of the constraints? How should different issues be weighted? Complex decisions have all these different pieces and there are not just one or two or three possible outcomes, but millions of scenarios with slight variations from each other.
So it’s about understanding the scope of options available to people? How does Ethelo present ‘millions of scenarios’?
There’s a way of characterising decision making processes that talks about them as a spectrum – a space of scenarios that is built out of these decision components as the core parameters. So the way Ethelo works is it takes a decision, perhaps a budgeting process, and breaks it down into a few basic components so that people can give feedback on the different pieces of the decision.
Each individual has a personal experience of playing with feedback tools to find their favourite scenario. But as a group, because we’re gathering rich data from each participant, we’re able to estimate not only what their favourite scenario was going to be, but how much they might like or dislike the other scenarios as well. We’re able to look at all of the scenarios and find optimal scenarios with a high average support, and also where the variation in support is low. In other words, outcomes where people are roughly equally happy.
And what about processing? Are there technical challenges associated with aggregating all that data?
When you look at all the combinations or permutations, just ten options, or variables, will get you a million scenarios, because it just explodes as a power function. In the past that would take us a day and a half to process. That’s a crazy amount of time, even though we built the computation engine pretty efficiently. But over the past 10 years we’ve made a huge amount of improvements. We’ve had a research partnership with the University of Waterloo Computer Science Department since 2019 and we just posted the most recent set of updates to the platform. Now for almost all of our decisions, and we’re talking about tens of millions, even hundreds of millions of scenarios, it can solve them in under 5 seconds, even when you have hundreds of simultaneous users.
How do you measure success? What constitutes a positive outcome?
If we can achieve outcomes where people are moderately satisfied but equally satisfied, those are strong decisions. Those are good democratic decisions. Because it’s not really about a winner and loser anymore, it’s about finding an outcome that we can all live with or we can all get behind, and we have solved very contentious problems with this kind of methodology. We do exit surveys with our participants on Ethelo and we’re regularly hitting close to 80% in terms of those who would support the outcome that is proposed.
That’s our contribution – breaking things down and taking a step back to think about what we are voting on in the first place and how we construct those proposals.
How does that process work from a customer perspective? Can they set this up themselves?
Up to now our process has been to consult with the client. We workshop with them, we take the information and then we build the workflow using the platform tools, going back a few times until they’re ready. Then we launch it to a community. We’ve taken that approach because our admin panel is kind of complicated and it’s not the kind of thing you can just pick up and deploy like SurveyMonkey. There are so many different features and functions that we were worried that if we did, people would have bad experiences, people would get confused and create bad processes, which would cause problems.
But we’ve changed our view and so this fall we’re going to be releasing Ethelo and the admin panel as a freemium product. So anyone is going to be able to set up their own process. We created a very comprehensive knowledge base. We’re going to have some templates that people can use to get themselves started and we’re going to have an optional support service that people can use to ask for help if they want it. Recognising now that there are engagement professionals out there that will take the time. It’s not for everybody, but some people will really want the full feature set we’ve developed. It’s kind of a big pivot actually. We’ve been refining this tool as a consulting business for 11 years and now we’re going to turn around and just give it away.
And finally, what distinguishes Ethelo from other platforms on the market? How do you view your work in the context of the wider democracy technology sector?
I think what we’re doing around the construction of what proposal gets voted on is unique. Figuring out how we fit things together into an outcome that’s coherent and consistent, and that will have broad support. We’re working on the aggregation of crowd intelligence for complex problems – the “convergence” stage. But there are other pieces of the problem. There’s different people working on ideation, or how to do dialog on platforms, aiming people towards more deliberative consensus, etc. There’s all these people working on different parts of the eDemocracy problem.
What I think is going to be really interesting and what I’m hoping to work on in the coming years is a protocol that allows for the integration of all these different technologies into a unified tech stack that is available for governments to use. Something that can cover the broad range of use cases.
Obviously, there’s lots of work that needs to be done to achieve this kind of integration. There’s lots of great work being done all over the place on different facets of the problem, and we need to figure out how we bring it together into something that can become the replacement for our current democratic system.
