 Okay, hello and welcome to Active Inference Lab. This is the third and a ongoing series of discussions between I guess Active Inference Lab and John Boyk. I'm Daniel, I'm a postdoc in California and I'm gonna be kind of working through the Jamboard as we've done in these other sessions while John goes through his paper, some of the implications. We can also raise up some questions from the live chat so people are totally welcome to leave questions in there. Wait, actually, it's already sort of messed up or not sure. I hope that it will work. It looks like it's okay. We're just sort of dealing with this dreaming tech and I'll play with it once you start talking, John. But yep, I hope this is a fun discussion because the other ones have been awesome as well. So I'll flip over to slide 14 from our second slide show and let you pick up with some points from last time and let you know if anything is awry. Okay, sounds good. Thank you. My name is John Boyk. I'm a courtesy faculty at Oregon State University Environmental Science Graduate Program and the series is about, our series of discussions is about a series of three papers that I recently published in the Journal of Sustainability and the topic is societal transformation. I think somewhere in the middle of this talk we'll show some links to those papers and to my website where all this information can be found. So do you have the slides up? Do you wanna put up slide 14? We're just gonna jump in. We're gonna jump in to halfway through the last talk. We had a bunch of slides that we didn't get to. So we're gonna try to get to them now. Yep, I'm there. So I'll be following along but you just won't see it but we're good to go. Oh, I won't see it. Okay, all right, okay. All right, so slide 14 is a list of 14 desertata that I had proposed in the second paper, part two. I should probably say what the title of that series is. The title is science-driven societal transformation parts one, two, and three. So we were talking about part two last time which was on motivation and strategy. And we're gonna continue with that now. So in that part two paper, I had listed 14 desertata for what any, what would be, what maybe could or should be in any R&D program aimed at societal transformation. And throughout the series, I touch on all of these 14. I would just like to make clear that what I'm proposed, the series, the three-part series is itself a proposal for an R&D program. It's kind of an informal proposal to the world about a program that could be started. And it is possible that other people have, other groups have other ideas of what kind of R&D program could be constructed. And it could be that in the future there's numerous R&D programs that are in operation that are all looking at this topic of societal transformation. And I wanted to offer a list of points or topics that each of them might address. And that way, at the very least, we could compare programs by comparing these 14 points. I just, we went through them last time so I'm not gonna go through them again but I just wanted to say a couple of things. Daniel, I remember from our last conversation you had said something about, you know, getting a bunch of scientists together and maybe they won't agree on different aspects of this and there could be some arguments for one point over another point or one approach over another approach. And of course that is true and that's fine and that's good. There should be a variety of voices heard and considered in any R&D program. But I did wanna make the point that the R&D program that I'm proposing, the concept is an evidence-based science-driven R&D program. So suppose that, well, let me say that that an evidence-based program is different than a popularity-based program. So it's not just that everyone gets to vote for what they want and the R&D program would go off in whatever direction everyone votes for. I mean, in a sense that's true because how else would you govern things? But in another sense, it's not true because there has to be some kind of science process. I mean, we have to be addressing with a design program, we have to be addressing the science-based aspects of it. And that is largely driven by a purpose and fitness. So the, so... All right, let's see. Yep, just continue going ahead, thank you. Okay, we're good? Yep. Okay, all right. I was saying that there could be multiple R&D programs proposed by different groups, driven by different groups, but I would expect that they would all be based on some kind of science-based evidence approach. So while each R&D program might have a slightly different way of measuring fitness, those discussions would be based on scientific evidence and reasonable evidence as opposed to just a popularity. We're good? Yep, working with tech. Thank you everyone for standing by when it kind of blips out for a second. We'll try to restart it like as soon as possible, but that's why we use the asynchronous modes of work because we're really lucky when we get to have synchronous time with the authors or with anyone, but the asynchronous and the updates on the comments are what really matter for this kind of a project, not someone's internal understanding or just a dialogue. So thanks for bearing with us, everyone. Go ahead, John. Okay, just to make that last point, is the R&D program that I've proposed is really, it's really a meta framework for learning new systems, new system designs. So the idea is that the R&D program, we would look at a wide variety of designs, people, it would provide a way of means for evaluation of different designs that different groups might come up with. So that's all I wanted to cover on that topic and then on to new information. Yep, okay, we'll be able to do it, right? Okay, here is this. This is gonna work, give me one second. All right, thanks everyone for holding by. I hope that you enjoyed this section, the silence to think about these ideas. Okay, so now. It's called making lemonade out of. Yeah, okay, so now I will spend some of the time when you're talking, John, to just continue like trimming and stuff, but we actually are back in the game. So please continue on side 15 in two and go from there and then continue on. So sorry about that, first 15 minutes of fun, now go for it. Okay, just one little thing. My screen, my picture on my computer is quite dark, but I don't know if I'm coming through on everyone else's computer. Looks good, and I'm gonna fix it on the other computer. Okay, all right. All right, so slide 15. I was saying before the interruption that this R&D, just a few things on the R&D program. It has a design vision of a 50 year horizon. So this is a careful movement towards societal transformation. It's gonna take a long time, but studies will be, if it were funded, if it move forward, studies would start right away. Field trials would follow in the near term. And those communities that choose to participate in field trials would see benefits far faster than other communities that are waiting. So this is a 50 year horizon towards a global scale or near global scale transformation. So it's gonna take a while to achieve change at that huge scale. I also wanna mention that the program involves the public in a variety of ways. This is, I envision this as a partnership between the public, especially between local communities and the science, the global science community. So there would be field trials, there would be a variety of studies and the public can get involved in a variety of ways. There would be lab studies, there would be surveys, there would be the ability for the public to input their ideas and their wishes and desires into the program. So hopefully this would be a conversation, a multi-way conversation between the stakeholders that are involved. And on the science side, I mostly want to make the point that the program could involve nearly every academic field that's out there, literally an A to Z. And at the bottom of this slide, I just list a few of the many fields that would be appropriate here. Certainly some of the science fields we've been talking about, complex science and other fields like that, but also education, media, literature, linguistics, like you name it. And there would be a place and need for participation in this project. Okay, next slide, 16. Okay. We're good? Yep. All right. So the idea that the science world would be actively participating in a program for societal transformation, this is in a sense, this is rather bold and new novel. There obviously are programs aimed at the study of transitions and developing the knowledge base for societal, some kind of a societal transformation because many in the science world and the public policy world understand that bold change is necessary if we're going to, you know, avert the problems that we're facing due to climate change, biodiversity loss and a number of other very serious social and environmental problems. So big change, deep change, is on many, many people's minds, but science, the role that science has, that many people envision for science is more of a knowledge, is more than the traditional science world, the knowledge gathering, knowledge production that would support the public and politicians and choosing new policies and various things. So the idea that science would be involved in the actual R&D of making new systems, designing new systems, testing new systems, that's more of what is being called second order science and that is fairly new in the science world. So this is breaking new ground for the science world. So I have a bullet list here of various reasons why the science community might want to be involved in this. So I'll just go through a few of these real quick. First, the idea that society can be understood as a organism, a super organism if you will, and a cognitive organism at that is timely and important and offers numerous opportunities to advance basic science knowledge. So while in participating in a program of societal transformation, a great amount of science knowledge would be generated. Next is societal systems are sophisticated. So the skills that are offered by the science community are extremely important. They would have to be computer simulations for example and a variety of other very technical problems that would have to be solved in order for new systems to really work and for the design process to work. So for that, the science community is uniquely qualified to contribute to those effects. The goals of transformation are ambitious. This is the third bullet and yet the time available for action is short. So this has to be not tinkering but moving in a direct and focused way towards societal transformation and the development of new systems and the science community can help with that to keep it all on track and moving forward. I've already mentioned the second order science threshold, the threshold to a new way of doing science and that can be useful and involve science more in the public sphere in the future. And the next is that the science community is trusted by the global public. And then is a position to help focus public attention on pressing problems and solutions. Next to the last one is the science community can offer a global perspective and global coordination. So what we would not want are a whole bunch of experiments to occur for the R&D program and then to be forgotten and never picked up again or never integrated into the body of knowledge after that. So the science community can offer a central, in a sense, a central focal point or repository for data and for assessing data and for assessing the many experiments that might occur in such a R&D program. And lastly, the science community can act as a powerful ally to the public in promoting and ensuring that transformation moves forward. So I hope that the science community will find interest in this kind of R&D project. That's really the purpose of the whole series is to offer the idea to the science community that we could do this together. This could be a very important and timely service that the science community offers to the world. All right, on to 17. All right, so, and I've said this before, but if I were the smartest person on the planet and I came up with a fantastic new design for societal systems and for viewers that are new by societal system to mean economic systems, financial systems, governance systems, educational systems, legal systems and the like. If I were the smartest person, I came up with perfectly designed societal systems. If there was no way to implement them practically, then my effort would have been wasted. So any R&D program aimed at societal transformation has to have some kind of viable strategy for implementing any new designs that might be created. And in the proposal in the series, I offer a theory of change and a strategy for change that I call local global viral. And the idea there is the local global is the idea that much of the work would occur in local communities and by local communities, I mean volunteer communities, volunteer individuals who would participate in the testing of new systems and even the design of new systems in cooperation with and partnership with the global science community. So we have the local global connection there. And the viral connection is such that the concept of the, I use what I call a club model. And the concept of the clubs is that clubs would be implemented locally on a volunteer basis, as I mentioned, and they would be designed such to produce substantial benefits for the participants. And if that is, if they're well-designed and indeed the participants obtain substantial benefits, then the concept of the technology and the concept for the clubs and the club model and the systems that they embody would quite naturally spread from one community to the next. If a club starts in a, say in a city, just not as a city, but just even a small fraction of a city that could be a club, like a club could be any, maybe perhaps a thousand people or more. So just a subset of a city could start a club and be part of this R&D, the whole R&D program. A field trial could happen via a club. And if the club really does produce benefits, then the next city over, the next community over is going to have an interest in starting it also because the benefits are substantial. And at the same time, a club can grow in the participation rate and local area could grow. So clubs can grow in size and they can replicate into new locations. And that sets the stage then for viral growth and that it would be how the societal transformation and the use of new systems could spread to a near global levels over a period of a 50 year horizon. So clubs are designed to basically out-compete native systems. And by native systems, I mean just the legal, economic, governance systems that we are using today. These new systems will obviously be designed for a purpose and that purpose is to benefit and improve the quality of life and the security of participants. And if they achieve that, they will out-compete native systems for the public's attention, engagement, trust and respect. So that is what they should do. And if they do, then their use would quite naturally spread in a viral fashion. Now the clubs, I just want to say about the club level. So we're not talking about say a family making changes. That's like this isn't really family focused, the whole project. It's not really focused at the business level like a particular business, a new business model, for example. It's not focused there. And it's not focused at the city level. So a new city charter or a new way the city can operate to improve somehow. It's situated in between the city level and the kind of family slash business level. And I think of that as the Goldilocks level. If a club was smaller, I would say this level is in between level, the middle ground, sometimes I call it, is the smallest level of organization that would allow a club to be viable, testable, self replicating and effective. So anything smaller than the club level, you could imagine say if we're just talking about a business, well then that business has to compete in an environment against other businesses to survive. And if those other businesses are not doing the same thing on the same program, they might have advantages that the new business model might not have. At the same time, if we went to a higher model, say a city level or trying to change things at the city level, well that's suddenly really expensive and difficult, especially for a large city. So there would be a lot of political blowback for trying to change things dramatically at the city level. And the club level is such that you could implement, you could implement new systems without having to deal with the need to win elections or get approval from any public body. So 1,000 like-minded individuals can just start a club and no need for anyone else in the community to join in. They can do it all on their own just as is. So that's the idea is to reduce the barriers to participating in a club and in new systems and have it be such that new systems could virally spread over time if they show their benefits. Okay, so just to make the point then, when I'm talking about societal transformation, the new designs that we're coming up with, I'm talking about the designs for a club, not the designs for a city, not the designs for a county, not the designs for a state or for a nation, but I'm talking about societal systems, designs for a club. And the club will then demonstrate to the world in field trials whether these new systems work well as advertised or not. Nice, good point. Because sometimes when we're talking about these complex multi-scale systems, it's good to make it clear that we're talking about policies, affordances, capacities that are at a certain level of analysis. We're not invoking a nation state or something like that. We're doing something, we're working at a workable level. This is focused at a workable level. This can happen. There's really no reason it couldn't happen. And obviously getting the initial funding for the project is a big deal. But once the initial funding happens, I don't see why this could not move forward exactly as or close to what is planned here. Yeah, next slide. Is this the end game vision? 18, is that where we're at? Yep, yeah, okay. So where is this all leading? I've already said there's a 50 year horizon aimed at societal transformation on a near global scale. But the real goal, obviously the real goal in this is for society's humanity to use systems that are beneficial to them that are most beneficial or widely beneficial to them. I believe, I should say there's good reason to expect that a well-designed R&D program that engages local communities and the science world could come up with designs that are substantially better than our current ones. I mean, if the goal is to increase quality of life and reduce uncertainty, if that was how you were judging a design of societal systems, then I think you would have to say that the current systems are failing. They're terrible in the sense that we now face enormous stress and strain due to climate change, due to biodiversity loss, due to a host of interrelated problems that all of which were preventable, all of which have been talked about in the science world and in other communities for the last several decades, and all of which had we taken action, we would not be under the risk that we are right now. So if native systems have given us a world of extreme risk, then I think it would be fair to say that native systems are failing humanity. And it would be hard to imagine that a sophisticated science-based effort with some of the best minds in the world focused on this topic could not do far, far better than current systems. Just to connect it to one note from Game Theory, it's that the strategy has to be resilient along the way. So it needs to have that kind of like introductory dynamics and getting off the ground and bootstrapping or not relying on lateral connections amongst clubs. It has to be sort of an end in and of itself. And then it also has to be a resilient mixed strategy at 50%. It has to be a resilient dominant strategy. And so flexibility to play different strategies at different phases of the uptake is critical. And that's why you've laid out a timeline as well. Yeah, yeah. And that's also kind of related to your point is that there could literally be hundreds or thousands of experiments going on around the world in different communities of different system designs and different aspects of a particular design. So you would want to be very flexible, very adaptable. We're learning as we go as a matter of fact and would change the strategy slightly as we go and change the focus slightly as we go, adjusting to circumstances. And starting small and growing organically helps you to do that as opposed to starting big and perhaps failing big. So in the end game vision, the idea is that in the future, humanity is using systems that really serve humanity. As we've talked about many times now, the view here is societal systems as a cognitive architecture is a way of facilitating societal cognition. And obviously we would all like our societies to be good at thinking, good at learning, good at adapting, good at making decisions so that we can all enjoy the benefits of such. So the end game vision is perhaps, I mean, here's a couple of scenarios for the end game vision. The R&D program moves forward. We demonstrate really useful designs where that gives substantial benefits to communities and the existing native systems slowly uptake those ideas and blend those ideas into their systems. And then in the end, the native systems become much better. So that would be great. That would be fantastic. And another scenario is that native systems fail to adjust and change and uptake these new ideas. And if that were the case, since the new systems are designed to earn the public's trust and respect and attention, they would seem reasonable that they would out-compete native systems for the public's trust. And if that's the case, then people would be using these new systems and relying on them more so than native systems. So those are just two scenarios, but you could imagine a number of other scenarios that might occur from such an R&D program. But either way, or at least with these two scenarios, either way the end goal is reached of a large majority of the human population using systems that actually serve them and don't serve them well. Slide 19. Yep. Okay, so a little bit more on clubs. So by the time the first field trial occurs in the first system, and this could potentially the field trial, the first field trial could occur maybe anywhere in the world, practically. But by the time the community participates in a first field trial, it should be really clear what the expected outcome is, right? The community does drive its systems. The community makes decisions day-to-day of what kind of, you know, about services, about waste control, about energy generation, about all kinds of things. The community makes its own decisions. But the general trajectory towards improved quality of life, improved well-being, improved security, improved economic security, reduced income inequality, like all of those things, there should be very few surprises as to what occurs in a field trial. It should be, there would have already been numerous simulations that would have occurred before a first field trial already, before it even begins. There would have been all kinds of user testing that would have occurred. There would be all kinds of game playing with new systems that would have occurred. So by the time the first field trial starts, this should be a trajectory of exactly what is predicted. And so there should be few surprises for the field trials. Now, obviously that doesn't mean that something, you know, there might not be hiccups or whatever, but by and large, there should be very few surprises. The community should know really what they're getting into, how the system operates, and what the expected benefits are in each year of participation. Second item is that while the initial field trials might be really focused at the individual club level, as the RMD program progresses, there would be an effort to connect clubs via networks so that clubs themselves are participating in some kind of cognitive operation as a community of clubs that also needs to make decisions to learn, to adapt, to make decisions together. So the idea is that some of the same kind of approaches and technologies that are implemented at the club level could also be implemented at the network level so that the entire range of clubs and the individuals within clubs could all be participating in some kind of optimal cognitive system, right? Making good decisions and opportunities for learning. Next slide. So why the club model? I've already mentioned a few things, but it's in the Goldilocks level that allows for replication and viral spread, but there's a few other advantages that I would like to mention. First, when you think about bold change, violent revolution is one kind of bold change. So we obviously are not talking about revolution, we're not talking about any kind of violence whatsoever. We're talking about improving the well-being of communities at every step along the process. So this is nonviolent and that's one part. Another part is that success is measured success and progress are measured throughout the process. So not only might you be doing better, but that would be validated through a variety of metrics. This is an experiment. The science world is engaged in this ongoing experiment to develop new systems. So data would be gathered and everyone would know that you're doing better if you're doing better, right? We would know if they're working or not. Maybe that's a better way to say it. Fitness, the evaluation of system fitness and the quality of life for community are central to this whole project. Again, because the clubs are operating in that middle ground, that minimizes political opposition. There's, as I said, any group can join a, can form a club so that you don't need city's approval for this. You don't need a nation's approval for this. This is just a civic club. So a civic club can start nearly anywhere in most countries or in most democratic countries. And so you don't need a vote. And this is really important because I am speaking to the global science community here and I'm inviting the global science community to participate, but I don't expect that everyone, all scientists will be interested in this. Nor do I expect that all individuals in the general population would be interested in participating in a field trial or anything like that. As a matter of fact, I would expect that most people, most people in any given city will have no interest in this at all, but that doesn't matter. Because again, all it takes is 1,000 people in your city or your area to start a club. That club shows benefits. It quickly becomes clear to others that this is a very useful, this is a very useful program. Incomes are rising. Incomes are stabilizing. Quality of life is improving. The club feels more empowered to make decisions that affect its future. There's a stronger sense of community. There's a better ability to deal with problems as they arise and find long-lasting solutions. So on numerous levels, quality of life and environmental quality should increase. Information should increase. Transparency should increase. Democratic participation should increase. And those are all really attractive characteristics if indeed the club model provides that. So if that's the benefits, then again, the idea will spread from one community to the next and grow within different communities. And it's not necessary that a large percent of the population jumps on board on day one. It's just not necessary. There must be, I would argue that even today if we looked around, there would be a community of 1,000 people who would be interested in participating in a field trial. And the first field trial, of course, is still some years away. The program isn't even funded yet. So by the time we get to the point where we really need a community to run a field trial, I imagine that will be quite easy. As a matter of fact, I would expect numerous communities to offer themselves as participants for the first field trials. Just a few more points here, but in whatever city the first field trials occur, they're likely to have beneficial spillover effects. So tax revenues might rise, for example. Other spillover effects from the club might help the city at large. So maybe even cities would welcome such a thing if it's going to improve the quality of civic life for those who participate and for those who don't participate. This club model is opposed to, say, going big, to changing the, for example, the U.S. economic system or the global economic system on day one or in a short amount of time. This local global viral strategy is far less costly, far less risky and far more flexible. We're talking about starting small, bootstrapping or growing organically and that all can be done on a practical budget, as we'll talk about shortly. And as I mentioned, it allows for parallel testing. So numerous field trials in numerous countries or cities could be happening all the same time. And as far as cost goes, what would be the cost? Well, it's a little hard to say because the program isn't really developed yet. But this is just a kind of a guess on my part. But I would think that in the first few years, the cost might be less than $10 million a year to get this started. And then in subsequent years, still in the first decade, the cost might be on the order of tens of millions of dollars. Now, that is, in a certain sense, that's a lot of money but in a certain world, that's a lot of money, but in another world, that's peanuts. So this is about the same cost as a modest size or a good size startup. So this level of funding is available in the world. This is not a crazy cost. And I argue that with this modest cost, we could really change the direct trajectory of humanity here. Okay. That's it for part two. Oh, that's it for part two. Okay, all right. So we will continue this talk about part two in our next Jamboard. So I'll add the new link and we're now on slide two of Jamboard three. Okay, all right, good. So the slide two is just where you can go for more information. So if you'd like to know more about the proposed program, if you would like links to all the articles and links to all the previous videos that we've made and that I've been involved with, you can find them all on the PrincipledSocietiesProject.org. All right, slide three. So there's a few remaining topics from part two that I'll talk about now. Some opportunities outside of clubs, a little bit about fitness metrics and then some final thoughts on part two. Yep. So we've already talked about opportunities in clubs which could include for example, increased income and higher quality of life and a bunch of other material benefits, material and social benefits. But there's also some opportunities that could occur for those outside clubs. And as I mentioned, clubs sit at the midway level between the kind of family slash business level and the city level. And both of those levels are already well developed and there's an enormous amount of revenue attached to both of those levels. But middle ground is not yet organized, but as it organizes, I would imagine that revenue opportunities are similar to those at the city level and those at the family slash business level. So the R&D program would help to organize this middle ground to the club level. And some of those, there's a little bit of evidence that such organization is already happening. For example, is it called, what is the app called? The Neighborhood app? It's the name escapes me right now. But there's an app that is widely used and is used by millions of people in the US at least about communicating within your neighborhood and sharing information within your neighborhood. So that's already a big business. That app alone is a big business. Cities are also developing community energy supplies, community solar, and that's just another way that this middle ground is organizing. And if the R&D programs goes forward, there would be a tremendous amount of organization happening within this middle ground. So that opens up opportunities for consultants, for example, most of the work in developing the R&D program, the tools and technologies needed for field trials for new systems. That would mostly be open source, publicly available knowledge and information. But it could be that businesses, consulting businesses or other businesses could provide services to new clubs, could provide, these could be training services, they could be like internet services or any other kinds of services that would help clubs get started. And I imagine that could be a big business as this moves forward. Maybe a good model for that is Red Hat Incorporated, which provides services related to the Linux operating system, provides services to businesses related to Linux. Linux itself is an open source product or it's open source technology. And Red Hat, which is a enormous company. I don't know if it's, I see a billion dollars in tax revenues recently. So this is a very, very large company operating on top of an open source stack in a sense. And the same could be, the same similar opportunities could occur for businesses who want to work in this consulting or service providing space. Another opportunity is for cities and other governments because of tax revenues. Later in, I don't know if we'll get to it in the next session, but maybe in the next next session, a section of these discussions we'll get to the simulation model that I've published on what I call the LIDA framework, Local Economic Direct Democracy Association framework. That simulation is just illustrative of a potential kind of a prototype new societal design focused on the economic system. But in the simulation that occurred, the simulation was for Eugene, Oregon, using census data for Eugene, Oregon. And the simulation occurred over a 28 year period, but by the end of that 28 year period, let's see, I've added down there just a few words about what that provided, but billion in tax revenue, right, right. Which was more than double the initial tax revenue. So potentially this is a, could be in a sense, kind of a windfall for local governments, city governments and community governments for tax revenues. And then finally, there's an opportunity for large investors. We are looking at a future world that is gonna be very stressful because of the implications of climate change, the effects of climate change, biodiversity loss and other social environmental problems. So if it happens that there's a shortage of good investments to protect your money, the club model and the club process in the R&D program might become, they might be seen as safe havens for parking money until things improve. And if that's the case, I could imagine that some institutional investors who desire to make social investments or who see the wisdom of making social investments could invest in this whole process. And of course that would provide funds for developing the club model and helping the club model to spread and grow. Yeah, who knows what is gonna happen in the stock market as the ravages of climate change approaches. And maybe this club model idea could really actually help large investors by providing a safe haven for their investments. And I'm talking social investments, so I'm not talking highly profitable investments that some might be used to, but I'm talking about social investments where it's invested in a low expectation of profit. Okay, slide five. Yes, cool. Okay, so just a little bit on fitness metrics. We've been mentioning them all through these talks. I just wanted to say that they serve several purposes. For example, this fitness metrics, the information gathered about fitness metrics. So for example, just to name a few fitness metrics, could be family income, could be education levels, could be water quality, air quality. I mean, you name it, there would be potentially hundreds or thousands of topics to collect information on. So this information could help a club assess its own status. How am I doing, how are we doing? You need information to answer that question. How are we doing? Well, we can say that we are doing blank. We're doing so well or so poorly or whatever the case may be on education, on environmental quality, on economic security, on all these different issues. Second is this information helps clubs to solve problems. We have, for example, how do we, how do we, maybe the problem is focused on economic security. How do we solve this economic security problem? Well, we need information. We need to know how we're doing right now. How were we doing last year so that we can know, we can make good decisions for solving these problems that we're facing right now. Metrics are also useful for assessing club and system design. Again, clubs are built for purpose and are supposed to function successfully for purpose and achieve the aims that they were designed to achieve. And all of that, to assess club fitness and to assess how a club design is doing requires information. Another use of information is just informing members. So everyone has questions about just day-to-day questions in their life. What career should I choose? Or where can I buy fresh strawberries today? Or whatever information that individuals need in living their day-to-day lives and running their businesses day-to-day, the information gathered by a club could help them to make better decisions just in their own personal life. And then lastly, the data assessed would just be a windfall for basic and applied science. In a sense, we're talking about a large repository of longitudinal information that follows individuals and families and businesses through their life and how they're doing their quality of life as decisions are made and as the club makes decisions as they make individual decisions. That information would just be gold for understanding societies better, understanding clubs better, understanding health better. Then that information, there's only a few small repositories like that today. So we're talking about just a really a goldmine for basic science and applied science if this moves forward. Obviously there would be personal, you'd have to protect personal data and all of this, but that could be done. And just a few words about how this, these fitness metrics and data gathering system would be different from what's already done, well, already for example, the census collects information and the sustainable development goals, SDGs are proposing a collection of data at the national level. That would be, and all of that data for the census and for SDGs, it touches on some of the same topics like quality of life and education and economics and so some of the large topics are similar, but there would be many, many differences between this kind of data collection and the existing programs. So just a few differences, this would be a larger and more detailed set of information. It'd be collected and managed at the community level, not at the national level. It would be about local conditions, so specific for local conditions rather than say national conditions. Forecasting other kinds of models are integrated into this repository. So this would be an intelligent, the aim is here an intelligent data gathering system. It would be part of forecasting metrics into the future and things like that. And obviously the club is designed to serve a different set of purposes and in particular purposes related to community cognition, societal cognition. Lastly, clubs would collect information on a far more frequent basis. So some survey questions are asked every few years, sometimes every year, sometimes every few years, sometimes every five years and the SDGs I think are collected mostly on an annual basis. So this is on a perhaps, some data would be gathered minute to minute or hour to hour or day to day. So there would be a far richer and higher quality data set that would be that communities would generate. And the basic idea is a community needs to know what's gonna happen to me. Like a community needs to know what's going on, where did we come from? What's going on? Where are we headed? What should we do? And to answer those kinds of questions requires rich data and that's exactly what would be proposed here. Community could collect rich data about itself. So it would know how it's doing and what's gonna happen. If I could also add one comment there with active inference type models we can have a rich or deep generative model even if we have sparse data collection initially. And then we can ask what kinds of collection of what type of data or what type of modeling or what perturbation would make sense given our generative model or how can we differentiate between two generative models or two hypotheses that play. So the richness can be specified upfront. We don't need to just grab what we can see in terms of data. We can actually be principled in terms of what information is pursued. Yeah, yeah, yeah. As we've already said, part of the concept behind active inference is decision making under uncertainty. And obviously we're talking about a community that doesn't know its future and isn't sure what's gonna happen next because no community understands its future very well. And especially in today's world is very uncertain about what's gonna happen next. Will there be enough food for us next year? Will there be enough clean water for us next year? I mean, these are some big questions but there's a million little questions. And so somehow a community has to make decisions under uncertainty. And that's really what the point of active inference is is how that can occur. And then touching on the point that you just raised Daniel is active inference is a way to make steps in cognition, right? It's a flow of information processing, decision-making action, information flow, decision-making action. So it naturally lends itself to an ongoing process of learning. And also a multi-scale one. So there's some- And multi-scale learning, yeah. Just like you said, some are one-year questions, one day, one hour. There's multiple ticks to these models that we've seen nested models at different timescales. Cool. Right, right, right. I also wanna mention too is when we talk about models, sometimes we might be thinking of say computer models and that's fine. Computer models can extend the extension of the human capacity for processing information. But you can imagine too that individuals do process information. They learn themselves about their world and what might happen next and what they expect if certain decisions are made. So part of the flow of information and the processing of information at the club level is to ask people what they think will happen if we do this. What do you think will happen if we enacted this rule or put our funding priorities over on this topic? What is gonna happen if we do that? What's gonna happen tomorrow? What's gonna happen next year? What's gonna happen to my grandchildren if we make this decision today? I mean, these are questions that humans can take a stab at, can answer to the best of their abilities. And we might also imagine that certain humans are quite good at it just because certain that we all have talents in different areas. So, this is not just about logic and computer models. This is also about engaging the human spirit and the human heart in directing a path towards a better future and making decisions that lead to a better future, both in the near term and in the long term. Nice. All right. So some final thoughts to part two. I've already mentioned that it seems already that there's enough public support for this because it doesn't require a lot of public support to get off the ground. Again, just as far as the science community goes, I imagine this could move forward if just dozens or 100 or something like that scientists were really interested in this topic. I imagine it could get started. And if just one community somewhere in the world is interested in perhaps doing a field trial of any new system once it's developed, I mean, that's really enough. So we're talking about a tiny fraction of the world population that would have to be involved in this to get it off the ground. And just like any new technology, the iPhone or like it doesn't, once you build it and you demonstrate that it works and it reduces people's uncertainty about what's gonna happen next and it empowers them to make decisions that improve their quality of life and improve their security going forward, then the rest is history, I think. That's how I imagine this would work out as you just demonstrate a better way and it has such benefits that it just spreads. So to get it off the ground, just a fraction of the world would need to be involved, a fraction of the population or a fraction of the global science community. But once it gets rolling, then I think people will actually jump out of the woodwork to be a part of this because it's like who doesn't want a better world that is demonstratively better than what we have now. All right, so there seems to be enough support. There's various groups working on similar kinds of topics about say for example, transformation and other topics like that. So it's hard for me to imagine that there wouldn't be enough support to move this forward if there were funding. Funding obviously is the big thing to get over right away, but if there were funding, I think this would go fine. Another question mark I have at the bottom is, it's a good question. Would transformation actually be easier than reform? I mean, in free form, you're asking corporations, for example, large corporations to change their behavior, even to change how they measure success of their endeavors. For example, by including more social metrics in addition to just revenue metrics. You're asking power, as we'll talk about soon enough, you're also asking people, those with concentrated power to distribute that power to the rest of the world. And this is a big ask. Reform requires some big actions by big players. And that's great. I mean, reform is absolutely necessary and every effort should be made to move reform forward. It's, we might not survive without reform. It's especially good for the short term to get things, improve conditions and improve our trajectory in the short term. But in the long term, transformation is necessary. Can I ask a question? Yeah. You mentioned funding being a rate limiting step. So what would you spend money on? Oh, well, as a couple of slides ago, I mentioned that it might cost about whatever, maybe less than 10 million or around 10 million a year for the first few years. But there has to be a variety of things would have to happen just to get the program going. You know, there would have to be conferences. There would have to be some kind of staff to handle the program. There would have to be some early studies done, like the first studies would have to be funded. A lot of the funding would actually just go towards studies. But studies about what the public's opinion of transformation is, what the, you know, how the idea of transformation is seen, for example. Simulation studies of potential, showing potential benefits. Like there would just, there would be a slew of work that would have to occur. And we're talking about work by professionals, scientists and otherwise. And, you know, that just costs money. 10 million is actually not much. That's a small team to get things off the ground and it would cost about that. There could also be funding for arts projects, funding for literature, funding for stories about what a transformed world would look like. So, you know, works to bring the idea of the, you know, like the potential of Trent to tell the public that we actually can, there's a path. We can transform. We could, we could live under different systems. And what would that be like? What would that look like? What would that feel like? What would we, you know, how would that be? You know, that's like a, it would be lots of art things and music things and literature things and media things that could support that, in addition to all the science based efforts. So, I don't think, you know, there's, there would be a few things you could do without funding. There could be maybe a few studies that you could kind of squeak by and spare funds or something, but without substantial funding, I don't think this would really go too far too fast. All right. All right, well, that brings us to our real topic, part three. Nice. So, part three is focused on design. And when I say design, I mean high level design concepts, not nitty gritty, how would you do this? How would you do that? But high level design, there's a few things that I talk about in this paper. The paper itself is a narrative on the topic of de novo design, you know, design from scratch of new systems. Obviously, as we've said many times now, the society is views it as a super organism and its societal systems as a cognitive architecture. So, in that sense, the goal of design is to facilitate cognitive, societal cognition through a better architecture. I talk, I spend time on a few topics in this paper, on attention, communication, self-identity, power, and influence are all discussed. And I also talk about prototypical societal system design. This is the LIDA framework that I mentioned just a minute ago. I don't think we're gonna get to the LIDA frame, I'm sure we won't get to the LIDA framework today. And there's a lot to say about that because that is, as far as I understand, is still the first simulation in the science literature about how an integrated system might function. So there's much to say about that. And I don't mean to say that the whole R&D program is aimed at the LIDA infrastructure, it's not, or LIDA framework, it's not, but those concepts could be useful in developing other ideas for societal systems. And at the very end of the paper, I highlight some design considerations. So we'll get to those maybe in at least two sessions, two or more sessions here. Slide eight, okay, so I just wanted to say a few more words about reform versus transformation. So as I mentioned, there are loud voices at the UN and even in governments, in national governments, state governments, local governments, and certainly in the general public and certainly in the science community, calling for bold, really radical change to avoid this cliff that were hopefully not gonna fall over here related to climate change, related to biodiversity loss and a host of other social and environmental problems. And just to kind of emphasize how dangerous our situation is, it is not clear if humans will go extinct with all the other extinctions that are happening. And I don't mean extinct in a thousand years, I mean, like, will there be, will there, at least will there be a civilization, any sensible kind of, any civilization that we're used to, will that exist in who knows, 10 years, 20 years, 50 years, 100 years, you know, like our future is highly uncertain because we have not addressed deep problems that have been with us for many decades. We've allowed them to grow worse. And now we are all at risk, at risk of just lower quality of life or at risk of mass die-off of humans and at risk, ultimately, of human extinction. So there's lots of voices calling for bold change and that bold change can be seen in a couple of ways. More, the typical way to see it is reform of isolated societal systems. Like there's a lot of focus, if you just read the news, there's a lot of focus on green energy, for example, using solar powered energy and things like that, better trans, you know, electric cars rather than gas, gasoline engines. That's all great. Again, all of this stuff is wonderful, but I would offer that, that there's something really missing from this kind of assessment and that is the larger picture, the meta level. There are sometimes called in the MLP of the multi-level perspective of GILS, it's called the landscape level or the regime level. And I would say that we have to address that level if we're gonna really solve our problems, if we're really gonna develop, if our systems are really gonna act well as cognitive systems, then we have to address the systems level parameters and characteristics. So right now, the focus of all the work that's being done in sustainability and the idea of bold transformation, almost all of it is focused on this kind of meso in between level of the, say the energy infrastructure or the transportation infrastructure. And this paper in this whole series instead is really focused at the landscape level, the large level, the whole system's level. I just wanted to make that point clear. And okay, so now we're already in part three of the series and I've listed six systems that are six meta systems or landscape level systems that I'm interested in, economic, political, legal, health, science and education. But in the first few pages of this paper, I finally say what that actually means. So I'm basing my definition of a societal system on what's called a function system. And the more there are citations there by Roth and Schultz and others about what a function system is, it comes out of social systems theory. But they list three characteristics of a function system. It has to be observable and not a subsystem of other societal systems. So the economic system is an entity in and of itself, not a subsystem of other systems. It involves a communicative system whose operations refer to a society and manifests the functional perspective at work. So Roth and Schultz identify 10 function systems. And I would say that six of these, I'm calling the overarching systems, the economic, political, legal and so on. In parentheses there, I have their functions. So the function, the one word function of the economic system is distribution. The one word function of the legal system is standardization. The health system is restoration and so on. The education system is formation. There's four other function systems that Roth and Schultz identify, art, religion, sport and mass media. And I do talk about these, but I talk about them more indirectly. So that's what I mean by a function system. And I also use the word, that's what I mean by a societal system. And they also use the word institutions throughout the series. And it's worth spending just a little bit of a, spending a little time describing what I mean here. I use that in the term in two senses. One is just concrete organizations, for example, a tax institution inside of a governance system, for example. But the second meaning is more broad. It's more along the lines of a social institution. And this is kind of interesting because Grabner and Gobani define social institutions. They give a little loop there. Codophile systems of social structures and particular norms and rules that incline individuals to act in specific ways. And interesting, the idea, this is, you know, they've developed these ideas apart from active inference, but the language is not so dissimilar because they talk about institutions as constraining and enabling people to act and coordinate and complex and uncertain environments. They stabilize the expectation of individuals about the behavior of others. And thus the institutions allow individuals to offload some of the cognitive burden that they would otherwise be required to compute. If there's some kind of a social norm and I expect you to behave in a certain way under a certain circumstance, I don't have to spend a lot of thought figuring out what you're gonna do next as I go about my life. I can just assume that you're gonna act according to the norm and then that's a whole lot of work that I don't have to do to forecast what I think you're going to do. So institutions allow individuals to offload some of the cognitive burden and anticipate the state of play and the beliefs of others. So obviously that's related to active inference, right? Active inference is when we talk about forming, structuring a society or structuring a society to reduce uncertainty, one way to do that is to develop a set of norms that everyone agrees to where everyone follows or mostly follows so that the uncertainty about what's gonna happen next is reduced. So developing institutions, if they're quality institutions actually helps them to reduce free energy and expected uncertainty. I think there's a few dimensions to this and it would be fun to unpack with what does it look like to have the active inference institution? What does that mean as far as the operating procedures of the internals of the business versus how the stakeholders interpret it, how it's communicated? So how it actually plays out will be something we're just exploring. Yeah, yeah. And when I say institutions, I'm talking widely, so financial institutions, social institutions, cultural institutions and so on. And I'm suggesting that the institutions can be viewed through the lens of active inference and may be improved by understanding how human cognition works and how group cognition works. We might be able to improve our institutions. Just some reminders from our previous talks about the goals and purpose of the R&D program and of a purpose of an organism itself and thus the purpose of a societal system that serves a group of organisms, the society. So the goal of transformation from this view is to develop and implement systems that can best facilitate societal cognition. So that's the goal, we have metrics to assess whether a new design is better or worse than some other design. And importantly, if we weren't thinking about cognition, if we were just thinking about conditions, we might say the goal of transformation is to improve education levels or improve diets or improve quality of life in some way or improve economic security or any number of things like that, all of which are fantastic or good. But once we bring this new idea of societal system is as cognitive architecture into play, then we're talking about the conditions such as clean air, clean water being an outcome of a decision-making process. So the focus starts to shift to the primary, the primal goal of quality cognition versus the secondary goal of what quality cognition brings. Quality cognition at the societal level brings clean air and clean water and low uncertainty and higher good education and freedom and any number of other characteristics, conditions that you might wanna list. So it is important, by looking at societal systems as a cognitive architecture, suddenly we have a much richer understanding of not only how a society works and how decision-making in a society works, but also we have new metrics by which we can measure the success of a societal system. And so those would include cognitive metrics, data metrics, information metrics, processing metrics, all kinds of metrics. So that's a little bit on the goals of transformation and the goals of the program and a little bit on the purpose, the purpose of the program and the purpose of a new design, right? I just wanna say that the purpose of a design is both aspirational and functional. We've already talked about design as the last step in the process of worldview leads to purpose, leads to fitness metrics, and then finally to some kind of system design. So in that process, obviously, purpose is functional, like practically everything else depends on purpose, or many things depend on purpose. But beyond that, having an explicit purpose for a club say it, like for a particular community, a clear statement of useful purpose can actually help in and of itself to a shared purpose can help to unite individuals and focus attention on common causes and encourage cooperation. If we all know why we're doing this together, if we all know what we're doing and why we're doing it, and that purpose is to improve our collective quality of life and improve the environment, well, then it's gonna be a lot easier to get people to come together and talk and unite behind a shared purpose. So in this sense, a purpose itself could aid societal cognition, right? Societal cognition depends on cooperation, depends on trust. So just having a purpose can aid in societal cognition. And as a counterpoint to that, what is the purpose of our current society? We have these native systems of governance and economic systems. So what is their purpose, right? What is the purpose of an economic system? I mean, we could debate that, right? But I would say that if you just take things that face value, one purpose would seem to be to promote revenue and sales and consumerism. And if that's the purpose, if that's part of the purpose of an economic system, then why would, it's hard to get behind that. It's hard for a whole society to say, oh, that's wonderful, let's get behind consumerism because it's so destructive. Okay, so a little bit on goals and purpose. And on to the next slide, this is just a quick review of active interest. Maybe I don't even need to say this because we've gone over it a few times, but essentially, the short view is that we can view action cognition as an uncertainty reducing process. So that's what this is all about in a nutshell. And via the active inference mechanism, we take in observations, we update our models, we learn, we take actions that can reduce our uncertainty, either explorative actions that bring in new information, epistemic actions or goal directed actions that actually immediately or in the short term act to improve our situation. So it's all about reducing uncertainty. And especially it's about reducing the uncertainty of essential variables. And this is of course, this is a rich topic in and of itself. What are the essential variables of a society? I mean, you would imagine that in some way they would include clean water and food and sufficient food and clean air and shelter and all those kinds of things. But as we talked about earlier in the last discussion, we talked about maybe eight or so core human needs that include affection and freedom and a bunch of other core human needs. So we're complex organisms and we have complex needs and these essential variables, the variables that really matter are themselves a complex set of variables. And if we focus our attention on addressing this question of uncertainty about essential variables, that almost means by definition, we're doing what I like to call solving problems that matter. Problems that matter are the problems that really relate to the essential variables about how are we really doing? How are we gonna do tomorrow? How will we do in 10 years? Yeah, so we'll just read the last sentence of the last block of text I have there. This is by Badcock and they just they phrased they're speaking about active inference and the active inference has a plays out at the social institution level. And they say that the elegant idea that we can operate together to minimize collective uncertainty stands to cast a new light on classical social phenomena and that's about it in a nutshell, that's it. And to draw one point on that in contrast to for example, reinforcement learning based approaches or reward maximization based approaches, it's almost like you have the Gaussian distribution, it has a mean and a variance. And the reward learning is like, let's shift the mean up as much as possible. I mean, isn't the average income right now in US dollars higher than ever? Well, maybe, but then if you think about what the dollar means and how the dollar is distributed then the average doesn't really tell you too much about the distribution. So active inference is about reducing your uncertainty on a specified generative model. So somebody might think, I think the distribution, we might wanna see this type of distribution, this shape or this kind and that's a discussion to have and then we can reduce the variance around that and let the mean fall where it may. So it's very different to have a reward highest as possible because having high expectations, but an unstated and often unformalized variance can lead to some ridiculous outcomes like the St. Petersburg paradox and other investing fallacies, let's say. Yeah, good point. So your point was that reinforcement learning is aimed at not really necessarily a probabilistic approach. And you will have to really ought to include the distribution of whatever targets you're looking at in order to get a better idea of what's a good decision and what's not a good decision. Absolutely. Another difference between RL and active inference is that the programmer, whoever is designing the experiment or whoever is coding the robot or whatever has to decide in advance what the rewards are for certain actions. So you fall through a trap door, negative 10 points, you win the chess game, positive 10 points or whatever the game is. And active inference allows the agent to learn goals as it goes, learn the world as it goes and change its priorities as it goes, which is exactly what humans have to do. We have to be learning as we go and changing our perspective and changing what our goals are and what we're valuing. We have to change that on the fly in order to maneuver in a highly uncertain world. And if I could add one more point on that, and this was expanded upon by Noor Sajid and Philip Ball in one of our previous discussions, that relationship and the learning of the cues in the world can be basically semantic and arbitrary. So sugar isn't just always simply a reward for the human in the environment. Like right now I don't feel like having sugar so it's not part of my preference vector. So with active inference type models, we can separate out, okay, what's on the table in terms of our current affordances? How about future affordances? What are our current preferences and how sure are we about them? And all these parts of the actual action plan that will help us align on shared action rather than thinking that if we simply align on shared thought then all of a sudden we'll know what to do because that's not even true. Right, good, good. Okay, so we've been talking about societal systems and we've been talking about societal systems as a cognitive architecture all through this. But maybe it's time we actually gave a little thought to what cognition actually is, right? So Lyon lists 13 cognitive tasks that he says qualifies as universal tasks or capacities. And this is part of the worldview that we talked about in the first discussion in that all of life is intelligent. All of life is engaged in cognition. All living things are engaged in cognition. And these are some of the tasks and capacities that what we mean by that. So these hold true for a human certainly but they also hold true for a bacteria or a slime mold or a plant or any other living organism. I don't know if I need to go through all 13 one by one but I'll just kind of jump around a little bit. So just number one, there has to be some sort of repulsion or neutrality or some kind of judgment of what's good and what's bad, what's useful and what's not useful. Certainly to cognate you have to sense you have to bring in information. There has to be some kind of memory storage or some kind of recall. There has to be some kind of learning, some kind of problem solving behavior. There has to be communication with other agents of the same kind or different kinds. There has to be anticipation. Certainly if we're going to make a decision about our future, we have to anticipate what our future might be if we do or don't do some kind of act. There's attention. There was a million things to spend our attention on as a human and what we spend our attention on will govern in large part the quality of the decisions that we make. And the last one I'll talk about is self-identity. So if I say I'm going to the store in that sentence, I'm talking about this particular body and my self-identity is focused on this particular body but I can also talk about myself as a citizen of a community, as a citizen of a nation, as a part of an ecosystem, as a part of the biosphere. So we have the capacity to self-identify at numerous levels. I call that extended self-identity, not just ourselves, maybe our family, maybe our community, maybe our environment, maybe our local environment, maybe our global environment and maybe of human civilization, we're a part of human civilization. And as we'll see shortly, how we self-identity, how we self-identify governs to some degree the kinds of decisions and the kinds of behaviors that we display. So these are, we can think of these as the occurring in every organism. And the next slide brings the two ideas together, the idea of function systems and the idea of these universal cognitive tasks. How do they relate to each other? And I'm offering that in a sense, these 13 cognitive tasks are really, you could think of them as building blocks for the higher level functions. So for example, an economic system, say in the function view, an economic system might involve all 13 tasks in one capacity or another. A education system might involve all 13 tasks but might particularly involve the tasks about learning, for example. So we can think of a lines list as containing low level cognitive building blocks and function system as being the higher level functions built on top of these or from these. And then we can think of active inference as the multi-scale process that connects the two. And these function systems then provide a cognitive architecture that provides a society with a cognitive flexibility. That is a society needs to make decisions. It has these six or 10 different systems, different function systems that it can make decisions through that it can learn through. And there would be certain kinds of problems that are maybe better suited for, say, an economic system or a governance system or something. And other kinds of problems that might be better suited for a legal system or an education system. But maybe some problems really have to engage multiple function systems. So the fact that a society has these various systems of cognition gives it a certain cognitive flexibility and provides a certain kind of redundancy. And then obviously raises the question of what kind of balance are these eight or 10 function systems? How does the dynamics of that balance play out? Next slide. In this slide, I'm just listing the next four things we're gonna talk about. And that'll be the, if we get to it, that will be the end of our discussions today. Those are neuronal versus human societies. What are the similarities? What can we learn from neuronal societies that might help us when we're thinking about design? All of these four points are related in some way to the design of new societal systems. Influence, how do we influence each other? And particularly what is the role of coercion and influence? Self-identity and selfishness. We just were talking about that when I was talking about expanded self-identity. And we'll pick up that theme in a short discussion about cancer, what is cancer? And then lastly is power distributions. We have eight or 10 or six or 10 functional systems. If one of those is unbalanced with the other, has too much emphasis in a society than others, how does that affect a society? And then also within a system, if for example, in an economic system, if financial power is concentrated, how does that affect a society? So we'll talk about that, all that. And then with the idea that all of these pertain to the design of new systems in some way, we can learn something from these topics about how we might wanna design new systems. So slide 16, so it's kind of interesting. You know, the associations and community that happens inside of our brain for human to solve problems is not so different in some ways about how we engage with each other in potentially solving problems as a society. So I'll just kind of read a little bit from that first paragraph, but similar to neurons, individuals connect with each other in complex ways, just as neurons do. We exchange a variety of signals over a variety of distances, just like neurons do at a variety of strengths. Both neurons and persons can potentially connect to a very large number of others. Each individual, each neuron, each person can share multiple connections with others. And these connections can form metal level structures like institutions or tissues or whatever the case may be. Neurons and persons self-organize into modules, tissues, families, clubs, businesses, et cetera. And neuron, neuron and person-person connections are dynamic and context and need-dependent. The neuronal architecture is quite flexible just as the human social interactions are flexible. Associations come and go, even if there remains some kind of large scale, consistent structure. And each individual is cognitive, just as each neuron is cognitive, that's actually kind of a new result in the science literature, the capacity of a neuron to process information in and of itself. So, obviously we want our neurons to work together. We want them to associate with each other in a way that helps us to solve our particular problems. We want their signals to be strong and good so that information moves throughout the neuronal process as it should. We want our neuronal architecture to reorganize itself as we need to, as we need to learn new tasks, for example, or new topics. And all of these are, this is also what we want for societies. If we're going to be part of this cognitive architecture, if we're engaged in societal cognition, we need to associate freely with each other, join the collaborations and groups as we see fit. That's part of societal cognition. Societal cognition and particularly the designs of societal systems might impact any in this bolded list that I have, but the first is the strength and the quality of associations between individuals, the richness and extent of our communication. For example, a design, a societal system design might have as part of it a means of communication so that say a large group of people can get together and collaborate and coming up with a decision on some topic, right? So the societal system itself could have communication and coordination aspects to it. Temporal and spatial locations and movements of individuals affect our societal cognition and designs can affect that. Quality of information that is gathered. And in particular, I'd like to point out the motivation to engage in say economic behavior, hopefully will be different than it is in native systems where the motivation is largely to increase personal income and wealth. We would want, go ahead. Oh, just one, I was gonna ask what the differences were in contrast, but finish whatever you're gonna go ahead and say. Oh, just on that topic, a new design would want to employ a set of motivations that is very useful for societal cognition and societal health. And certainly encouraging people to act in a kind of a selfish, self-centered, strict selfish way for their own gain would make little sense. That's like a design flaw in our current system. That we're not encouraged to cooperate and encouraged to solve common problems together. And the design of a system can deeply affect what our motivations for actions are. As a matter of fact, a design ought to make cooperation so easy that it becomes the de facto default behavior. It should, that should be the kind of, you would think that would be how a good design would work is it would just reduce the barriers to cooperation and communication and coordination because those are all key to societal decision-making and learning. Just a few others here. The design might affect, for example, how well a society's beliefs aligns with its reality. For example, if in a new design, there's a strong role for science and there's funding in clubs for scientific exploration. Well, then maybe it's more likely that new systems will be more aligned with the beliefs of people working in new systems or engaging in new systems might be more aligned with realities because science at its best is aiming towards a better understanding of our world. Design might affect how independent and cooperative individuals are, as I mentioned a minute ago, also might affect how well, how deeply the level of critical thinking is, right? How well does a society, as a community, encourage critical thinking through education and through norms and through other processes, but obviously critical thinking is important in problem-solving. You can't solve problems if you're not doing critical thinking. You can't solve problems well. And also the capacity to plan, the extent of the anticipatory rise and what a design might affect how much attention is paid to the various predictions made by organizations and groups. Maybe it would be important to pay attention to the predictions that people made so we can understand who is good at, who has a skill, who's good, how can we improve our capacity to anticipate what's gonna happen next. So in numerous ways, in the same sense that design of a brain system might affect cognition within the neural communities, some of those same ideas can play out in the design of societal systems because associations and et cetera, et cetera, would be important at the societal level. I'll just point out, as an evolutionary biologist, that selection has shaped a lot of the social conditions for today have benefited from developmental, basically, patterning that has been selected on for a long, long time, many, many, many generations. So real evolutionary pressures, the kind that create resilient neural systems at the level of societies, would require essentially population-like dynamics, competing and the birth and the death and the diversification of the neural systems. So that's what we're trying to do here. The birth and the death and the diversification of human societies. So systems mapping is really helpful and the toolkit transfers there for sure, but there's a key evolutionary bit that neuronal ensembles have, that ant colonies have, that simply saying distributed systems are like those two things is not there, but in the long range, that's where it would be. Yeah, yeah. I think I've said this before, but if we're talking about how can society do better at cognition, what kind of systems might facilitate cognition? Well, then all we have to do is look at evolution. We look at what is the world doing around us? How does a brain work? These are the information and the ideas and the patterns that we see throughout nature can be, we can learn from them as to how good cognitive systems work because nature is built. Nature is a good cognitive system, right? There's so much to learn from nature. Cool. All right, so this is the second of the four topics, influence, how do we influence each other? So we have a group, obviously the people will be, there's problems to solve, there's activities to undertake, there's challenges that arise and group members would try to influence one another in some ways through communication or some kind of communication. And Turner, the fellow I'm focusing on here, his ideas come from, he did work in social identity theory, self-categorization theory and some other related fields. And he identifies two forms of influence. Now, this is not coming from the active inference world but I'm going to make the point at the end that either the two are closely aligned with each other. One type of influence is persuasion and that's where you simply argue that your understanding of reality is in some way better than other understandings or your path of action is in some way more beneficial than other paths of action but already then we're talking about aspects of cognition in the human brain or active inference because we're talking about some kind of testing of a model of looking at potential paths forward in action and then evaluating them as regards to which is the most beneficial path to take. So already we're touching on the ground of active inference. Another form of influence is control and that comes in two flavors, legitimate authority and coercion. And legitimate authority, we've already talked about this because we've talked about institutions and offloading of cognition from the individual level to the group level through social institutions of one kind or another, norms or actual concrete institutions and things like that. By and large, legitimate authority, people are willing to, accept legitimate authority because they understand that it's in their best interest. One example is most of us in the world are okay if a traffic cop stops someone who is speeding at a very high speed because that's dangerous to us all and we are glad the traffic cop is gonna take some action and exercise authority over this fellow who's endangering the rest of us. So by and large, we are good with legitimate authority and we don't feel that it limits our freedom. And then there's coercion, which is more of a confrontational form of influence where you maybe force someone to behave in a certain way or to act in a certain way. And there's various forms of coercion that can occur including various forms of manipulation. I'll just list a few, but that could be disinformation, deceit, omissions, obstruction, the exercise of false authority, drowning out or silencing of competing voices or various things. So this is rampant in our societies today. These occur, you can turn on the TV and you'll see evidence of all of these. And Turner makes the point that coercion, all of these forms of coercion is easily abused and very dangerous to a society because it tends to reduce trust in societal institutions. It tends to reduce the long-term health of the group. It can be destructive in the sense that coercion breeds resentment to being forced to behave in a certain way and it can reduce trust and reduce cooperation and it increases an antagonistic climate. And that means that the society who's during the coercion or the group that's during the coercion has to use even more energy and more resources to buckle down and make the coercion even more severe so that they can get the individuals to behave in the way they want. So there's a role for coercion, especially say like a police arresting someone dangerous who has a gun and is about to shoot someone, you would hope that they would tackle that person to the ground or whatever and prevent them from hurting others. So there's a role for coercion, but it can be easily abused as I said and must be used very transparently and judiciously or else in the long run, the whole society is gonna suffer and potentially go down the tubes. So that's the way he sees it. And then we can see how this plays out in the active inference framework. Already we've talked about persuasion and the similarities between persuasion and learning and then legitimate authority as being similar to offloading of cognition to social institutions. And obviously when we're talking about influence, we're talking about a group associations. And we've already now mentioned that and with respect to neuronal associations and how neurons and people in a society have to have the freedom to associate at will to provide flexible cognition for the society. Right, so association is a part of influence to just the fact that we're associating and trying to influence one another is a part of cognition. So what are the impacts of this on design? Like how would we, how might we change a design or what kind of design elements might we incorporate into new societal systems if this was our concern? If we wanted to encourage persuasion and legitimate authority, how, you know, this is the question I'm asking, you know, the world in this paper is how could this impact design? And you can think of a few things off the top of your head. For example, transparency would help this process. Coercion is very difficult to achieve if there's high transparency. If there's quality information flowing, then coercion is more difficult to achieve. If there's quality education happening, maybe even especially education about what coercion is, what is manipulation? What are the many faces of manipulation? How do different groups in society use, how do they manipulate in order to get their will? How is, what is the role of media in coercion? You know, these are all good questions to ask if we're talking about what kind of new designs would we want in order to avoid some of these problems? Nice. The next of these four topics is self-identity and use the example of cancer here. So, Levin has published a couple of papers that I think are really interesting on this topic. And it's true that as cancer develops in a body, the communication tends to be changed. There can be less quality communication between a cancer, you know, between a tumor and the surrounding tissues. And at the same time, there can be more communication between cancer cells or within a tumor. So communication kind of goes haywire in a cancer or can go haywire. And the idea that Levin has is that what happens is we often portray a cancer as being a selfish, you know, a selfish growth that's occurring in the body. And he says, you know, it's not really quite like that. It's a little bit more like both normal tissue and cancer tissue is self-interested and acting in a self-preserving manner. The difference is what the extent, you know, the functional extent of self is for the two populations. So the functional extent of self for a normal tissue, normal cells in a normal tissue is the whole body. Like the cells are connected and are communicating with other cells in the body, both locally and distantly. So the body is a cognitive apparatus. You could say an extension of an individual cell outward to a larger cognitive apparatus, a larger environment, a larger milieu. And in a cancer cell, that communication is breaking down and the functional extent of self for a cancer cell might grow smaller and smaller and smaller. So the cancer cell, just like the normal cell is acting in a self-interested and self-preserving manner, only its functional extent of self is small. So it's taking care of us, it's including in its self a small group of cells, small tissue relative to the body. So its ability to think, in a sense, cognate shrinks because it's no longer attached to the rest of the cognitive apparatus of the body. It has to do any cognitive processing that it can do within a much smaller set of cells. So its capacity to act, say over multiple timescales to behave in such a way that it will exist in two years or five years is reduced because it's, he has a nice term for it. He calls it the cognitive light cone of self is reduced in a cancer. And this is really interesting. When we talk about how this might play out at a societal level, there's maybe some injury or stress in a certain population, that population is retracting itself from maybe from the rest of the society, the rest of the environment, because it's maybe feels threatened or it feels unheard or it feels unappreciated or it just is under stress. And it's the local group is forming that is more interested in the local cognitive light cone of self, a smaller light cone of self. And that can be dangerous. If that happens too much in a society, that can be dangerous. It can also be useful because people are getting together in small groups to think about how to solve a problem. That can also be useful. But if there's too much of that or if it happens under the wrong circumstances, then you can get the situation where the society is really kind of starting to disintegrate because there are elements in the society that are acting, preserving the small self and acting to protect the small self as opposed to the whole. And you can just imagine what that leads to. And so the task of design then is to create a cognitive framework that in which communication is rich, in which everyone's voice is heard, in which problems are actually addressed and not just glossed over or ignored where the needs of the whole are taken into account, where everyone's story matters, where everyone has an opportunity to impact the larger workings of the society, the decisions that the society makes. And building such a design, building such a society and a cognitive system then would reduce the risk of, you could say, cancerous growths occurring. Right, so cancer is informative in this sense. It's an amazing, my background is in cancer biology and it's a never-ending source of information and to look at how the complexity of the situation surrounding cancer. So that's it. So we can design systems so that cognitive light cone of self is expanded rather than contracted. And that's gonna go a long ways towards, towards kind of encouraging cooperation, encouraging coordination, encouraging group problem solving on group problems. Nice, thank you. The last topic is power distributions. So we've been talking about functional systems. Each functional system has a function. I've listed some of those. And each functional system also operates via a particular medium. So for example, you know, a good example is the economic system operating under, through the medium of money. So some authors that I cite suggest that each functional system if left unchecked tends to have a growth imperative. And we certainly see that in the economic system. It's like everything is now suddenly my couch in my living room is now worth money because I can sell it to someone who wants to visit my hometown for a weekend, right? So everything is commoditized now. Or many things are commoditized that used to be more freely shared without just in the course of life with people who had needs. So I sell my time, I sell my resources, I sell my ideas and the financial system seems to have overtaken many aspects of our life. So in that sense, in our current society, you might say that the financial, the economic system is overemphasized relative to the other functional system, say the education system and other systems. So obviously then when we're talking about design, one of the goals of design is to balance the influence of each function system. And that is a topic that we already mentioned very briefly about integrating systems together. So integrated economic system, legal system, governance system, all serving the same purpose and all acting as a flexible cognitive architecture for the greater good of the whole. If I can give a comment on that, you mentioned checks and balances and that in the American context referred to several separations within the governance mechanisms and the development that happened subsequent. Here we're talking about checks and balances in a little bit more of a general cybernetic goal-oriented system way. And we're talking about checks and balances between these different sectors it almost sounds like. So it's not just that there's gonna be market mechanisms with all of the manufacturers of a product competing on that one dimension. And then on a separate sector, there's competition of the academics on this one question in this one way. How can we link those competitions in a way that actually is aligned with the overall goals of society? Right, exactly. How can we align all of the systems to the same purpose? Which is the reduced uncertainty and expected uncertainty of the society in relation to the essential variables and things that matter. Right, so that's one aspect of design on this is how can we maintain a balance among systems? But another aspect is how can we maintain a balance within systems? So the economic system again provides a really good example here. And I have a sentence there that today the wealthiest 1% of individuals globally own about 40% of the world's wealth. Now, that is like insane. I mean, that is crazy, especially if we think of money as a voting tool. I mean, money is a voting tool. It's not a recognized formal voting tool, right? But it's an undemocratic one. If I'm a billionaire, I have lots of votes in this world. I can get people to do any number of things that I wish them to do. I can direct nonprofits to focus on particular issues. I can make a new company that provides a particular service or I have a billion times more influence than just the normal person, right? So money is a voting tool and it's just used as an undemocratic one. But it could be used as a democratic voting tool. The money system, the economic system could be highly transparent. It could achieve a reasonable degree or even a strong degree of income and wealth equality so that the money as a voting tool becomes a fair voting tool, right? Like in a democracy, we would not stand for I get one vote and the guy down the street gets a billion votes. I mean, that would just, we would recognize that as being deeply unfair but it's the same thing in an economic system. If one person has $1 and the next guy has $1 billion, that person has enormous control over how society makes decisions. So it's not just economic power. I mean, I'm focusing on the economic system but there's six or 10 systems that we could be focusing on here and each of them has some kind of medium and we could talk about the distribution of power in any of them. For example, we could talk about the distribution of power in a governance system. Currently in the US, there are what, 400s? I don't know how many legislators and senators do we have in total? I don't know what that amounts to but compared to the population, this is a tight, we have empowered, we've put enormous power in the hands of a tiny sliver of people. And is that a good idea in today's world? Could we do better? I imagine that a different kind of governance system would do a far better job of distributing power so that everyone can feel like their voice is heard and that they can be part of the decision-making process. You know, as it stands right now in America, I get to vote every two years about like a yes, I get like a thumbs up or thumbs down on certain candidates and that's kind of it. Formally, that's it. I mean, I can do some informal things. I can write to my congressman or I can complain in the newspaper or write a letter to the editor or something. But as far as formal things that I can do to have my very nuanced understanding of the world heard, all I really get is like a yes or no every few years on a group of candidates. And that's hardly taken advantage of my understanding of the world, right? I have much more insight to offer than just a simple yes or no every few years. So I think it would be possible then to design systems that balance this medium of communication in each of the function systems, achieve some much greater degree of incoming wealth equality, achieve some much greater degree of power distribution within governance systems, power distribution within legal systems. There's a great room for improvement on all of these. Yeah, so I think, is there one more? Oh, there's one more slide. Yep. And that is really focused on economic power. So the first one I already talked about using money as a voting tool, we already do that. And we would be better if we just did that above board and in a democratic and transparent fashion. And new systems could do that. And when I say economic systems, again, I'm including in that monetary systems, financial systems in the bigger picture. A large set of studies have shown that incoming wealth inequality is bad for society. It affects health, cooperation, trust, anxiety, happiness, social cohesion and mortality and social stability and numerous other factors. So it's quite clear that high amounts of incoming wealth inequality are detrimental to society. And again, today we are extremely unequal, extremely unequal. And then the last, just to end this whole talk today, I would like to point out that income equality is really not a, it's kind of, it's not unpopular, the concept is not unpopular. There was a large study done a few years ago where they showed participants graphs of the division of economic power or income and things like that to, they had three different choices. An image that was really very equal, like 50-50 equal, one that was fairly equal but not quite perfectly equal. And then a really extreme one. Well, it turned out that the people didn't, they asked people like, which of these do you prefer? And it turned out that they didn't like the extreme one at all. And that happened to be the one that was the picture of the US, inequality in the US. And the one that was more mild, I think was the inequality maybe in Sweden or something like that. So they preferred the distributions of economic power that were less extreme. And when asked, which of these do you think is the most fair for the society? The more than half of those queried, and this was many thousands of people who were queried, chose actual income equality, like completely equal incomes as the most fair way to run a society. And I don't suppose that people really thought that was possible. I know it's not possible, but that would be the more fair way. That's what I would imagine people were thinking. And when we next come back to talk, we'll talk about the leader framework and some potential designs that actually would achieve income equality, income and wealth equality over time. So that's yet to come. Wow, awesome. That really flew by after some little tech bumps it was awesome. And that is a little bit like an analogy for maybe this program. There's some technical details up front, few live streams to start, few teams, few memes, but maybe once there's momentum and once there's proof in the world and a little bit more of a contact between the theory and the practice, we would see some new kinds of momentum. So maybe give a last comment, what is gonna happen in the fourth session other than our two group discussions and... Yeah, yeah, we're gonna work it. We have one more, I'm glad you said that because I did want to say to folks that we have one more discussion in this series coming up next week. And in that discussion, we'll talk about the leader framework as a prototype and some of the things that might achieve how it might work. And that is a big part of our discussion next time. And then there's two regularly scheduled, regular scheduled discussions on part three of the paper that will happen in May. I think the first one is in two weeks. It's the first two Tuesdays of May at 7 a.m. Pacific and that will be like the group Active Inference live stream, so we'll have a bunch of other colleagues and that will be more of a group discussion. But the fourth conversation between just us will be April 27th at 3 p.m. Pacific. So I'm really looking forward to that and thanks for coming on for so many times and giving us so many good insights. And we appreciate your time and all the participants who stuck with us also for the tech. Thanks for having me.