David Snowden
17,090 views • 18:34

For all of our history as a species, we've contrasted two types of system: ordered systems and chaotic systems. Ordered systems have huge value for humanity. It gives us legal structure. At the prosaic level, in operating theatres, it means we can rely on the fact that they'll count out the surgical instruments as they counted them in, and you won't have to be operated on for a second time. As I get older, I find that sort of process of increasing importance. Order is hugely valuable to human beings. On the negative side, the fear of chaos is being used to impose order unnecessarily and destroy creativity and to destroy freedom. The reality is over the last 40 or 50 years, we've taken an engineering focus on society, an engineering metaphor, and we've actually compounded order with excessive outcome-based measurement. If you actually look at the whole history of the last 40 or 50 years, everything has to have a target. Everything has to have a defined outcome and it has to be a number, whether it's KPIs, number of published papers or whatever else. The reality is, all of the scientific evidence says that when human beings are pursuing explicit targets, it destroys intrinsic motivation. There is no evidence to contradict that. And where do we most need intrinsic motivation? In health and education. And where do we impose the worst targets? In health and education. So we need to start to think differently about this, and we need to move away from a primitive dichotomy in which we contrast one highly structured system with an absolutely chaotic system into something more sophisticated. Now, there is actually a third type of system which exists in nature. It's a complex adaptive system. It's a system defined not by its structure, but by its connectivity. In a complex system, everything is connected with everything else, but many of the connections cannot be known. We talk about dark constraints, a reference to cosmology and dark matter. We can see the impact of a constraint, but we don't know what the constraint is. We see there are boundary conditions which aren't defined. We see connections making things novel or different in different ways. The Internet is an example of a complex system. Humanity is a complex adaptive system. Now, understanding these and understanding how we manage them is critical. And it's not about control. It's about understanding the connections, about changing the linkages. And the best way I've ever learnt to define the difference between the three types of system - ordered systems, complex systems and chaotic systems - is to start to think about how we would use those approaches to manage a day to day problem. So imagine, if you can, that tomorrow you have to organize a party for a bunch of nine-year-old children. Everybody can imagine that, yeah? And you're making the mistake of holding it in your own house. You learn not to do this, as you actually have more than one child, right? The advantage of community centres is they have fire hoses. Fire hoses are very useful for cleaning up after a party, and they're occasionally necessary for crowd control during the party itself. (Laughter) So let's imagine how we would organize a party, based on what assumption we make about the type of system. If we assume that the system is chaotic, then it means the children are acting without constraints. Their behaviour is random. So they'll probably discover drugs and alcohol and go on a personal experience of self-discovery. Your house may burn down in the process, but it was socially constructed in the first place, so why are you worried about it? I have a friend in California who did try this once. Yeah, he's never going to do it again. The recovery cost was too high. (Laughter) The ordered-systems approach, on the other hand, is taught in all good management schools and all good business consultancy outfits. Under this, it's of critical importance to agree learning objectives to the party in advance to the party itself. The learning objectives should be aligned with the mission statement for education in the society to which you belong and should be clearly articulated and printed off on motivational posters with pictures of eagles soaring over valleys and water dropping into ponds, and placed around the walls of the room where you're going to hold the party. As the children come into the party, they should be given Disney playing cards with the party value statement clearly printed on the back. You will, of course, have produced a project plan for the party. The project plan will have clear milestones throughout the party, against which you can measure progress against ideal party outcome, and the senior adult will start the party with a motivational DVD - you don't want the children wasting time in play which isn't aligned with the learning objectives - and then they'll use PowerPoint to demonstrate their personal commitment to the party objectives and show their children how their allowances are linked to the achievement of the milestone targets. (Laughter) Following the highly successful completion of the party, you conduct the after-action review, update your best practice database on party management and mandate future process improvement. If at this stage, for any remote reason, the children aren't happy, you hire one of the new happiness consultants who will train them to be really very, very happy, and God help them if they're not the next time they come to a party. (Laughter) The complex-systems approach, on the other hand, is much simpler. We start off by drawing a line in the sand known as a boundary constraint in complexity theory, and we look the children squarely in the eye and say, "Cross that and you die." (Laughter) And one of the things you learn pretty fast as an adult is the value of flexible, negotiable boundaries, because rigid boundaries have a habit of breaking catastrophically when you least expect them. We then introduce catalytic probes - a football, a barbecue, a computer game - any type of thing which will actually trigger a pattern of play, which is called an "attractor." If it's beneficial, we give it more resources. If it's not beneficial, well, that's where we deploy the fire hoses. What we manage is the emergence of beneficial coherence within attractors, within boundaries, and we manage the only three things that you can manage in a complex adaptive system: the boundary conditions, the probes and the amplification strategy. Management and governance is much simpler when you understand the nature of the system and you stop trying to treat an ecosystem as if it was an engineering problem when it's an ecological problem. So the work I've been doing with my research group and elsewhere is to find new methods or new ways to deal with this, but you have to understand what's going on. You can only understand a complex system by understanding the small particular parts of day-to-day interaction. For humans, those are the anecdotal data of the school gate, the street story, the beer after work. They're not the grand narratives of workshops. It's the day-to-day anecdotes of people's existence. And we need to understand them through the voice of the people who tell them, not through an AI machine interpreting the text or an expert making them fit their cultural expectations. People's own voice has to be subject to their own interpretation. And then we need to allow those in power at any level of society to have direct access to the raw stories of the people they govern, without multiple levels of interpretation which allow them to hide from reality behind the guise of policy reports and everything else. So that's kind of like the work we're doing. In order do that, we have to engage people. And people have had enough of surveys. They've had enough of focus groups. Focus groups are biased by the facilitator within 15 minutes. Opinion polls aren't trusted anymore. We've never, ever lived, by the way, in a fact-based society. It's just we used to trust experts and now we're trying demagogues for a change. The reality is we actually hand over a lot of our cognitive processes to structures; it's not based on ourselves as individuals. So in order to do that, we have to engage people. So, for example, in South Wales, we found that girls' rugby clubs are the main agent of change in some of the industrial communities. So we give them tools by which they can capture their experiences and the experiences of the people they're trying to recruit. Then they go into old people's homes to actually help those people become more healthy because they've become engaged. We're actually working now through schools because teachers need to teach kids about community engagement, about statistics. So we give them tools so that every month their children can go into their communities and gather stories from those communities so they become ethnographers to their own condition rather than relying on outside experts. This is a huge programme which we're starting off through small countries and city states - Wales is a small country. And the goal is to allow people to be ethnographers to their own condition, to report their own stories, but within a quantitative framework, not a qualitative framework. Over the last 30 or 40 years, I've seen many attempts at digital storytelling, at gathering people's stories, but nobody thinks about how to scale it, and nobody allows people the power to tell you what their story means, because that's something done by the people who gather them. That kind of like leads us on to using people as ethnographers to their own condition. This is a picture from work in Africa, where we got young girls trying to understand the abuse that they and their colleagues have suffered, but making them the ethnographers, not external experts. In order to get there, as I say, we have to empower people. And power comes from the power to interpret your narrative, not from the power to actually hand your narrative over to somebody else. In order to do that, we have to find a way to scale it. So there are several key elements on this. One is we need to work with abstraction. If you don't know it, art comes before language in human evolution. The ability to abstract allows for scientific invention. In fact, the overemphasis on STEM education at the moment around the world will destroy scientific creativity because without art there is no invention. That's basic scientific fact. So if we move up a level of abstraction, we avoid gaming. We need to increase cognitive load because we need to use the popular language to have people thinking slow, not thinking fast. We need reflection, not the immediate response you get on questionnaires. This is an example from work we've been doing in Northern Ireland, another small country, in which we've been empowering patients to keep the stories of their own journey through the health system and then self-interpret those stories so they become a subject in their journey for health, not an object. It doesn't challenge the role of the health professional, but it makes them a partner in the journey, not the dominant partner, but a co-partner. And here, as you can see, by telling the story and balancing it on a triad between three qualities, you don't know what the right answer is, you get better data. That gives us quant data at scale, typically five to six triangles together with other geometric shapes. It is more like a game, but it's quant backed up by qual, and it means we can capture narrative in written form, oral form or pictorial form. We're not restricting it to text for data analytics, because what you can write down is at best 10 percent of what you know. If you focus on information, you radically reduce human knowledge. It's far more than we can write down. Then, of course, we have to find a way to enact what comes out of that. We have to find a way to make real change. Then we come on to a whole new theory of change. This is an example from a project working on health and safety. You get very similar patterns, whether it's in aerospace manufacturing or the health sector. These are two patterns from two groups: the vertical dimension here is rule compliance; the horizontal dimension is job completion. This comes from thousands of micro observations, recorded and interpreted by patients or by health professionals. This is not based on surveys; it's a continuous mass capture. Now, you can see immediately on the left they've got a problem. You either follow the rules, or you get the job done. The two are mutually incompatible states. All of the questionnaires, all of the focus groups, all of the interviews say they're doing both because people know that's the right answer, so that's what they tell you. The day-to-day micro observations tell you something completely different. On the right, it looks better because you've got a pattern at the top in which everybody is actually following the rules and getting the job done. But that turns out to be a crisis. Yeah, it's accident and emergency, it's not the day-to-day operations. On a day-to-day basis, people have to break the rules to provide empathetic care to patients, but increasingly - bottom left - more people are just giving up than doing what they need to do to survive, a pattern we're seeing in universities. So now, how do we change that? A traditional change mechanism, and this is the engineering culture, would define a desired future state and try and close the gap. That's bad complexity science. In complexity, what matters is we describe the present, and we make small changes in the present to nudge the system in the right direction. But I don't mean the conventional approach of behavioral economics and nudge economics because to be quite honest, that's more yanking than it is nudging. It decides where it wants people to be, and it tries to tug them towards it. What we want to work out is where people are and see when it's ready to change. So in this case, I look for what in complexity is called an "adjacent possible," a cluster of narratives, a cluster of stories, near to where I am, but going in the right direction - and this, by the way, is called a "vector measure." I don't measure outcomes, I measure vectors: direction and speed of travel for intensity of effort. I click on the material and I say, What can I do tomorrow to create more stories like these and fewer stories like that? That is an instruction which can be understood by anybody at any educational level. What can I do within the compass of my power to create more stories like this and fewer stories like that? If this is a hospital administrator, they ask the questions of themselves. They don't say, What can I get my staff to do? Because basically, as we go underneath that, every ward, every nurse group, every health group, at a government level, every district has their own map from the same source data, so they can all nudge their systems in a direction appropriate to their context rather than being subject to the tyranny of the average approach, the global campaign. We need to start doing small things in the present rather than promising massive things in the future because that just leads to perpetual disappointment. And we can go beyond that. This is from work on peace and reconciliation, where we've taken ideas as to how we can resolve a conflict, and we've presented it to hundreds and thousands of people for them to interpret it, and we've drawn a pattern how different groups in society represent or see the same data. You see here that the blue and green groups see the world in such a different way that there's no point in getting them to sit down and talk because you'll just increase conflict. The only thing they're both agreed on is that the red guys are really stupid. So we actually have to try and find a way to channel that: How can we create more stories like this, fewer stories like that? How can we change the day-to-day narratives of the present so a new future is possible? If you don't change the way that people describe the past and the present, you haven't got a new future anyway. So that is called fractal engagement, where the new key concepts on citizen engagement will abide. It goes beyond that. In the South Wales valleys at the moment, as we gather this data in, we're then starting to pair people. Instead of young people coming together and having a great event and be terribly idealistic and then nothing happens, we're saying to young people, we'll train you how to use the software, we'll train you in data analytics, but you have to bring somebody from your grandparents' generation with you so you work together as a trans-generation player, because actually, grandparents will tell things to their grandchildren they won't tell to their children, and vice versa. And if you start to create interventions at your local level which work, we'll make you a trio with somebody from the government who can make your ideas happen. So at the ministerial level, What can I as a minister do to create more like this, fewer like that? At a council level, What can I do to create more like this, fewer like that? At a local community centre, an arts centre, What can we do to create more like this, fewer like that in our community? Multiple mass fractal engagement to achieve genuine sustainable change in society, and based on a scientific approach, not an engineering approach, based on managing a complex ecosystem rather than trying to maintain a machine. It's an open programme for governments, for countries, for engagement worldwide. Thank you very much for your time. (Applause)