Do you have free will? - Part 2

Welcome to part two of our discussion, Kevin Mitchell on Free Will. Let's get straight into it. This is simplifying complexity. A podcast where we explore the underlying principles of complex systems. Systems that seem to defy our rational view of the world. Like economies, ecologies, or even you or me. I'm forensic engineer, Sean Brady, and I'll be your host. So where do we go from multisallot our nervous system, brain in some cases. And I presume evolution is weeding out the multisallot our organisms that failed to get coordination. Right. It is right. And yet there are of course lots of very successful multisallot organisms that have taken a different route. Plants, for example, do very well for themselves, but without a nervous system, because they have a different ecology. They've made different lifestyle choices. They just like to sit around and let sunlight come along and be their energy source. So in the book, I follow the trajectory that goes towards humans. So we get multisallot life, you know, eventually we get vertebrates and then mammals and then primates and then humans. And that's the lineage that I was interested in. But of course, there's all kinds of diversity that I ignored because you can go down the lineage to insects or mollusks like, you know, octopuses or invertebrates. You can go to birds or bats or whatever it is. You can get different styles of living in the world and different kinds of agency and different kinds of control systems in the brain. A lot of it depends on the life strategies that the organism is pursuing. But for us, there was along the our lineage, we can see this progression of increasing cognitive sophistication. The brain's got bigger and bigger. And as that happened, that opened up some new kinds of ecologies to us. It opened up new environments, new ways of living, which then made it even better for the brain to get bigger. And so these are expensive. They're like metabolically energetically really expensive to run. They're expensive to make. They're expensive to wire up and they're expensive to run. So if there wasn't a good payoff, we wouldn't have such big brains. And evolutionarily what happened is this kind of a snowball effect where the bigger the brain got along our lineage, the better it became to make an even slightly bigger brain. So we're not sure that we multi-cell you know brain, but we're still in invertebrate. Yeah, exactly. So if you think about a little nematode worm. So there's nematode called sea elegans, which is a favorite model species for geneticists and neuroscientists. And it's really simple. It has a thousand some cells, 302 neurons exactly. And we know what they all are. And we know how they all connect to each other. And we kind of wriggles around in the soil, hunting for bacteria to eat, trying to avoid getting eaten itself, looking occasionally for mates and so on. Pretty simple kind of life that it leads, but it needs to detect things out in the world. So it has a sense of smell. It has chemical receptors for lots of things in the world. It has very sensitive sense of touch on its nose or its tail. And it has kind of a simple intervening layer or a few layers of neurons between these sensory inputs. And the motor outputs. So basically it takes in a bunch of information, does some kind of calculations in there and then says, OK, I should do action type A, which is move forward or I should do action type B, which is move backward or whatever. So it has these command neurons that just execute an action once it's been decided on. Now that's doing some decision making it's doing some integrative holistic kind of processing, but it's very simple. It's not very deep. If they're thinking about anything, they're not thinking about much because first of all, they there aren't many layers to the internal processing. But also the things that they're detecting are just right next to them. So they're detecting chemicals that they're binding literally touching the chemicals or they're literally touching something in the world. So in a sense they inhabit the here and now, right, they're only responding to things right now, right here, right next to it. They're not thinking ahead. Living in the moment. They're living in the moment. Exactly. Now what happened, you know, in the evolution of vertebrates is you got distance senses like vision and hearing like your visual system is detecting photons. That's what's literally happening, but you're not interested in photons. You don't care what the photons are. What you care about is what's out in the world that the photons bounced off of before they hit your retina. And that's a really challenging thing to figure out because you need all these levels of processing internally in your visual system in order to make that inference about the objects in the world that are the causes of your sensory data that you're getting right now. And that you know, the same is true for hearing. And of course, what that means is, first of all, it means you need to have a lot more central processing in order to make those inferences. And then secondly, it also means that because you literally can see things a mile off in some cases, it pays to think about things that are a mile away in time. So now your cognitive time horizon expands because now it's worth it, right? You have something to think about. Things that are far away in space are a bit further away in time that could affect you. So you get this sort of expansion of cognitive abilities that goes hand in hand with those sensory capabilities. And does that come before we get vertebrate animals? Is that one of the reasons why we get vertebrate animals because now that you've decided something's forward in space and time. Anything that can get you there quicker and easier presumably is good. Yeah, so it comes at least much earlier than that because we can go, for example, invertebrates like mollus and insects and so on have vision and hearing and distance senses as well. So the common ancestor of those things, which is some unknown creature simply called or by Lataria, which just means this sort of primitive bilaterally symmetric thing that had eyes and ears and some kind of a brain that must have had some of that lifestyle. So probably during what's called the Cambrian explosion, there was this proliferation of animal forms and animal lifestyles where predation became probably a major evolutionary force at play because things could move around. They could see things coming so they could hunt. And if you have that, you have a kind of an arms race because the prey is trying to get away. The predators are trying to get better at hunting. Of course, the prey may be hunting its own prey somewhere else, right. So you get the idea at least is that there are all these sorts of pressures which led to the elaboration of these systems and then evolution is really being a harsh mistress there because it's straight up killer be killed. Literally a doggy dog world there and so that's the kind of thing that could have led to the pressures to elaborate the nervous system and make it more complex. So that organisms that can take in more information and integrate that better. Think about things in a longer term way execute plans. So whether not just doing an action right now, they have a goal and they say, OK, well, actually in order to achieve that goal, I need to move over there. Or I need to do this first or I need to do that right. So you can start to get a goal directed behavioral control where it's not just binary decisions in the moment. It's these organisms have an agenda. They're doing something through time. They're engaged in activities, not just actions. And that changes things as well. And that's the kind of distinction that you see obviously manifested to its highest degree in humans because we can plan for things over our whole lifetime or even into centuries beyond our own lifetimes. And the knowledge gets built in there as well, presumably you're planning over time and space. And you're using what happened to you before or what you've seen happen to others in that decision making process as well. So information that it's not happening right now, what happened in your past. And that's the critical role in that process. Yeah. And so that's not just current information that stored knowledge. And that comes from our lived experience, not just as human beings, but you know, even simple organisms that have the nervous system organized like that will learn from experience. And so they'll be able to associate things in the world and basically what they're doing, you know, they're using their perceptual system to make a model of what's out in the world. And they have a model of themselves, right, what's their internal state. And that determines or influences what their motivations and goals are at any moment, right. So, you know, if you're hungry or not, that influences your motivations towards one kind of action versus another. But as well as that, they have this knowledge that they build up of the causal relations in the world, which is when I see this thing in the world, I know that it can do that. This thing can eat that thing and maybe it can eat me. And also they have a knowledge of events, right. So when a happens, it's often followed by B and then C and then D, right. So they learn up this mapping of the causal relations in the world, which is exactly what an organism needs. And so that's the information that it needs to extract and store as knowledge, because that's the useful information. And in fact, you know, so many things. happened to an organism during the course of its day, and most of it is not stored, because it's not useful. It's only the really regular stuff that happens repeatedly that gets reinforced enough to form that part of that map of knowledge that we have, that model of the world and the causal relations within it. That's the model that we then internally use in order to simulate our actions. So when we encounter some scenario, occasionally there may be very familiar scenarios, and we'll have been through that hundreds and hundreds of times, and we know exactly what to do. And that's true for us, it could be true for a mouse, true for a bird, whatever, right? So they could have very habitual behavior in some scenarios, because they know that that's the adaptive thing, they don't have to waste any time thinking about it. But in other scenarios that are a bit more novel, we'll have to try and think, well, okay, what should I do here? So based on some past experience, not exactly in that scenario, because I've never been in that exact scenario before, but I've been in this similar scenario. So maybe I can use what I learned from here and transfer it to this new place, right? So then what happens is there's a system that can kind of generate some options for what you should do. And then there's another system that evaluates those options, basically by predicting the expected utility of each of the outcomes, right? So you have an action, you say, okay, I could do a, b or c, if I do a, what's likely to happen? And how good would that be for me? And if I do b, what's likely to happen? How good would that be for me? And then there's a kind of a competition between those things, and the ones that you're predicting would turn out best, get upvoted, and the other ones get downvoted, and eventually you do that result, and you do one of those things, right? Now the way I'm talking about that, it sounds like very overt deliberation, like in a human brain where we're talking it out, all of that can be subconscious in other organisms. It doesn't have to be a mental activity. It's just a cognitive activity. It doesn't have to be experienced mentally, but basically that's where the knowledge pays off. And the reason is, you don't have to try it out in the world where you might be killed. You don't have to try everything. You can think ahead and say, "Okay, let me think how this would work." Oh, no, that's probably not going to be a good thing. I shouldn't do that. So that's where paying attention to what has happened in the past, and investing in that store of knowledge becomes really useful and adaptive for an organism. And how much of that is the sophisticated version of learning truated by Query? You are taking in stimulus. You're attending what happened when you tried to rate it the last time and you fell off, and then you're using that to make a better decision when you try and write it this time. And then you just accrue all that information, very subconsciously with writing a bike, isn't it? And then you have that information for your life. So that kind of procedural learning, a learning of motor skill, is super, super important for organisms like us and many others, but especially mammals, which have very flexible behaviors. We can do lots of behaviors with our body much more than, say, birds can't, or insects can't. Their actions are pretty circumscribed, whereas we can learn new actions and new activities through our lifetimes. And so part of what the value of the motor control system, also being a learning system, is that we can improvise through our lives and develop new motor skill. But then there's also the aspect of the control system, which is the decision making aspect, which is, okay, in this scenario, which of these motor skills or activities should I do, right? There's a being able to do it, and then knowing when you should do it, and those are separate kinds of things. So the learning that we have, which is like a learning of events, sequences of events, that you could narrow it episodes, that's called episodic memory, and learning like associations between things, famous example would be Pavlov's dogs, where he trained his dogs to know that the ringing of a bell meant that food was coming, and eventually when he rang the bell, the dogs would start salivating because they had paired those two things, right? So a lot of the learning that we do in the world is like that. We're associating one thing with another thing. That's how we learn about the categories of, the properties of objects. It's how we learn about categories of objects. It's how we learn about sequences of events, and especially causal relations. So where do we go? We've got to the point we've got the brain. We can process information. We can retain historical information. Not only can we come up with options, but then we can evaluate them and pick them. Where do we go in your story from there? Yeah. So the next step is the most amazing step in humans, where you know what I've said is with inner nervous system. We have a model of ourselves, both of our internal states and of these motor control things of like how I can control my body. We have a model of the world. We understand, I have some knowledge about causal relations and so on, and we're going to work through all those things and decide on some action to do, as we've just been discussing. But we have, maybe no other organism really has an extra level. So we've developed some extra levels of the hierarchy that actually have a model of the workings of the lower levels. So we have what's called metacognition. We have a model not just of the world at our bodies. We have a model of the workings of our own brains, our own minds. I mean, this manifests in things like, for example, not just having a belief that such and such might happen or that such and such is the case, but having a degree of certainty in that belief and knowing what that degree of certainty is. So say we're out of the public quiz, right? And the question comes out. How many number ones did the Beatles have? And there you are. And you say 20. And I say, "Shaw, shaw, are you sure?" And you say, "I'm pretty sure, but I like I'm not 100% confident." And that's useful. Now that degree of certainty becomes a useful piece of information for us to think about because that says how much confidence should we put in Sean's wrong decision that the Beatles had 20 number ones? They had 17, I think. So the metacognition gets us that kind of level of certainty. It also lets us think, you know, I might have a belief and I might have another belief and they could be in conflict with each other. And if there wasn't a system above that that was aware that both those beliefs were being held and could kind of ask, well, what's going on here? Those can't both be true at the same time. Then, you know, we would just be operating at this lower level, right, of isolated kind of things. So this metacognition gives us an extra level to think about what we know, to think about the reasons that we come to and whether they're good reasons to have. And we can think again, we can reason again, we can change our reasons and so on. So we have this extra level of internal metacognition that gives us really some top down control over our own thought processes, right? We have introspection into our own thought processes. It's not just a black box for us ourselves as the user. And you're saying that we the only species that have this. So like a dog literally will do the computation. They'll rely on the current stimulus to put that together with old information. They'll come up with outcomes and then they'll pick one and they'll do it. But there's no system in them evaluating how sensible that is. But I presume that if it goes bad for them and they survive, then they learn. But that's just a new input at that level. It's not an input at a higher level. It's not a arrival of a higher level. That's the theory, yeah, exactly. And so now I may have been doing some of those other animals wrong by saying they have no metacognition. That's probably not an accurate way to put it, right? But we have the most elaborated form of that. So other animals will have certainty signals and confidence signals and so on that feed into that decision-making process. But probably within the same level, if you know what I mean, not this extra level that evolved. And it actually like physically is an extra level in our brains, the prefrontal cortex, which in some sense you can think of as a top of the hierarchy of decision-making. I don't want to give the impression that it's like the CEO sitting up there that's in charge of everything, right? It's very much a bidirectional, it's a whole system kind of thing. It's just that the top level of the hierarchy is concerned with this model of the mind itself. And it's also concerned with planning and goal-directness over the longest timeframe. So when you choose a goal, could be a long-term goal, like going to college, for example, right? Is a goal that somebody can choose. And when they do that, then they have to maintain all kinds of other things in order to achieve that goal four years later. And so having that goal acts as a constraint on their decision-making on a yearly and a monthly and a weekly and a daily basis and an hourly basis, right? Which lectures they need to go to and so on. So you get this hierarchy of planning and these execution of activities and behaviors over long timeframes, which is really how most of our behavior is controlled. Even though within the free will literature and within a lot of neuroscience, people are much more concerned with these instantaneous A or B binary decisions and generally ignore all of the context of I'm a living thing trying to do things through time. And that's constraining and informing what I do. So is that what brings us to free will as opposed to agency, even though I presume free will is a form of agency that you want to talk about in human. So is that what gets us to this concept of going, oh, we're no longer. this machine that fires up good ideas and worn winds, like it may happen in a dog, we have this ability to decide, oh, there's three good ideas here and I'm going to go up shini. Is that what's happening in the brain? Yeah, no, I think that's a good way to put it. And yeah, I mean, you could think about what would satisfy a definition of free will or what would meet the criteria for the phenomena that we observe, which is that in the world, we seem to be able to control our actions. We seem to be able to do so for our own reasons. We can judge those reasons, we can think about them. That is, we're capable of reasoning as a verb, right? That's an activity that we can do. We can do reasoning to come to our reasons, which is, you know, maybe different from some other, certainly different from a bacterium, which just has the reasons that evolution gave it. So if we can do all those things, we have a system for controlling, we have multiple options open to us. We literally have systems for picking one action versus another based on our reasons, based on our knowledge and our own individual history. Well, to me, that satisfies what, you know, reasonable sort of criteria for free will. That sounds like free will, to me. Of course, there's a whole other set of questions then that come back to this issue that we mentioned at the start, which is the issue of moral responsibility. And then this question is, okay, well, what do we think about that? To me, that's a separate thing. And in fact, what's interesting is that one of the pieces of evidence that we really do have free will and control our actions as agents is that it clearly, that capacity clearly differs between people. Some people have more free will than others. Some people have more control. Some people are more impulsive. Some people are more compulsive. There's all kinds of psychiatric conditions that impair decision making, like attention deficit disorder, obsessive compulsive disorder, PTSD, schizophrenia, dementia, you know, and they can sort of dissociate different aspects of those systems. So we can see it's not magic. This is an evolved biological system, a suite of capacities that involves lots of different elements, perceptive cognitive memory, decision making, evaluation, impulse control, planning, you know, all of those things are capacities that humans have in general, but some people have more or less of individual capacities. And of course, you know, babies have less control than adults do. So I would say babies have less agency. They're not able to control their, even the control their bodies to begin with, but, you know, toddlers are not able to plan delay gratification for a long time, right? And that's why in our legal systems, and just generally, we don't hold small children as responsible for their actions as we hold adults. And we don't hold people who are mentally ill or insane or have dementia responsible to the same degree as healthy people. So I would say free will clearly is a thing that exists because for people who deny it, who just say no one has any free will whatsoever, then I always, you know, want to ask what do we noticing in the difference between a healthy person and a drug addict? What are we talking about? I will call that their capacity for free will. And if you're not talking about that, then I don't know what we're talking about. Can we take a small segue there before we go to the next bit, which is when people mount arguments that we don't have free will? What's the basis of that argument? So there's a few different ones. And there's a really basic one, which is just from physics, which just says, you're just made a physical stuff. We know that the equations of physics are fully deterministic. And therefore, whatever is going to happen to the atoms in you is just determined by the forces of physics and the basic laws of physics. You can take the state of a system, even a complicated one like you or me in our environments. You could say, here's the position and the momentum of all the atoms in it at time t. And we just play out the equations. And here's going to be the state of time t plus one and t plus two and t plus three and so. And it turns out that's just wrong. The equations of physics just aren't deterministic like that or the equations might be, but their idealizations. And so, you know, there was an idea that classical physics, like Newtonian physics, just really is like that. That's why we can predict the orbits of the planets with such accuracy, thousands of years in advance, right? Turns out the orbits of the planets are just very obliging. They're really, really simple systems. And most systems are not like that. So quantum mechanics came along. Quantum mechanics is sort of inherently indeterministic, although people have different views about what that means. And there's arguments that that indeterminacy at the really, really lowest level doesn't bubble up to what's called the classical level, the level of objects of visible size. I call those the Las Vegas arguments because it's like what goes on in the quantum realm stays in the quantum realm. But again, there's actually no good evidence for that either. Those are just idealizations or assumptions as well. And in fact, most, you know, most systems are actually not these nice simple linear systems where the future trajectory really is plotted out. It's predetermined as of this state, the future trajectory just exists. And it's one line. Most systems are not like that. So you can predict the orbits of the planets, but you can't predict the weather on the planets. And that's not just a statement about our ability to predict. It's a statement about the system. The future of the system, the weather on this planet doesn't exist. There just isn't a fixed line that it goes to. It's a branching potential future. And so we're like that, right? We're chaotic systems. We're complex systems. And the little bit of indeterminacy all along in every molecular process that's happening within us all the time. What that means is that the future is not closed, right? The future is open. Lots of things could happen. Given the state of the system right now, many, many things could happen. And the challenge for an organism is to make happen what it wants to happen. We got this big possibility space and the organism wants to constrain it to just this. It wants to make this happen. And you know, we talked at the beginning, this idea that the first thing that living organisms do, that they make happen is themselves. They make themselves happen into the future. And behavior in the world is just kind of an extension of that idea. It's doing things in the world to continue to make myself happen. That's the argument from physics. And I think it just doesn't work from physics. Yeah, even if you stay within the discipline, it doesn't work just within the discipline, right? Determinism is just not a result of physics. It's not something physics has proven. It's an assumption and a mathematical idealization that's used to allow people to do equations and so on. Then there's a problem that arises from that because you could say, okay, well, look, find not all that low-level stuff is deterministic. Okay, the future may be open. But how does that help me if there's some random stuff going on down there? And that determines my behavior. Then I didn't do anything either, right? So it's either way you're screwed argument, like either it's deterministic or it's indeterministic, which is it? It doesn't matter because either way that argument is basically a causal reductive argument. It's saying that all the causes are happening at the lowest levels. And they just manifest as larger behavior of the large system because the system is made up of all those little bits doing things. So it doesn't really help if occasional quantum fluctuations push the system one way or the other, right? So I think that's the wrong way to think about it. That's a kind of a weird way of thinking, okay, most things are deterministic, but occasionally you get these sprinklings of randomness, right? These little burps of events happening that can push things one way or another. I prefer to think of it as just an under-determination. The current state of the system under-determines the future. So there's some slack in the system. It could go many ways. And what that does is create an opportunity for the organization of the system to have causal power itself, to act as a constraint, a top-down constraint on the way that its components behave. And it doesn't change the physics, it just imposes boundary conditions on the way the system evolves. So you can get macroscopic causation and even top-down causation in a hierarchically organized system that's perfectly unproblematic from a physical point of view or a philosophical or metaphysical point of view. There's no magic or mysticism going on here. It's just to say that the way a system is organized can affect how it evolves through time. And that just is what it means to be a living system, is to be organized in such a way as to affect how the system evolves through time by staying organized that way. I think you can get around the arguments from physics the either way your screwed argument by realizing that the indeterminacy or the under-determination is exactly what gives scope for macroscopic organization to have some causal power in the system. Not just some epiphenomenon, not just emergence with scare quotes around it in some mystical kind of thing. It's a very understandable process and it's actually lots of people these days are even measuring the extent of macroscopic causal sensitivity or causal efficacy. So neuroscience is the other one where you can say okay fine look I'll grant you that the atoms are not completely controlled but you can still think of an organism as a kind of a big complicated stimulus response machine where you have something comes in, you have the state of your brain right now and you can allow that that encompasses all of your beliefs and knowledge. and desires and motivations and so on. So you can say the state of your brain basically entails your mental state right now and everything about you. So we grant that, but you could argue, once you have some scenario or some stimulus, then it will be input to the big machine that does all those integrative calculations about what to do and so on. And then it will spit out an answer. So your brain is deciding, you're not deciding, your brain is deciding. And that's weird in multiple ways. So first of all, philosophically speaking, it's a very dualist position because it inherently says that you are somehow different from your brain. It sort of says if there's not a spirit or a ghost in the machine that's doing the controlling then it's not you. And that's just a weird view to take. It's a very anti materialist view to take, which is odd because many of the people making that argument are really, really materialists and reductive materialists. So they're kind of setting a bar that could never be reached without magic. So if you don't realize that what you are is entailed by the activity of your brain, then you're just on a different playing field. It's just a different argument you're having. But the other claim that some people make, so Robert Sapolsky for example has made this claim that in any given scenario, these computations that happen will always spit out one answer deterministically. There's no evidence for that. It doesn't fit with our common experience of being potentially conflicted in different scenarios. And it doesn't really make sense because we know that the brain is a noisy system where there's lots of indeterminacy and lots of randomness and stuff. And what the organism is trying to do is sort of buffer all of that to make the best decision. But there's background noise. So it won't always make the same decision exactly in every given scenario. There's no sense of thinking it's pre-stated in some way. There's just like a lookup table or in a sense, it's already pre-determined what you would do in any given scenario. There's no good reason to think that. So instead, it's much more a holistic exercise of the organism trying to get along in the world with incomplete information, with varying levels of uncertainty, with conflicting goals over short and long term, trying to do this massive optimization problem to maximize its utility through time with what economists call bounded rationality. Nothing about that is deterministic. That's just a big, ongoing problem of a process that the organism has to go through to accommodate to the world as best it can, not in some physically determined way. So when we bring all this together and we've got this meta cognition coming out, what does this mean for us in terms of free will and agency and understanding why we make the choices we make as human beings? This cash is out in lots of ways. So it might sound really esoteric, sort of like a fun argument to have in the pub late at night, but it doesn't make much difference. Do we have free will or not? It's kind of semantic, philosophical argument, but it really makes huge difference actually in lots of ways. I think we need to have a science of agency and we need to have a biological science where that's more holistic and systems-based. And I don't mean holistic in the mystical roots sense. I mean, thinking of the whole entity, acting as a whole entity as we talked about earlier. That's the right way to think about what a living being is. And if we continue to just do reductive biology where we say, what's this part of the machine do? What's that part of the machine do? We just miss the whole point of the dynamics of the whole system. So that's one thing just in terms of the way we think about biological systems. Of course, there's also issues around in the legal world about responsibility and blame and punishment and so on where the science that we're learning from neuroscience, from genetics and other areas becomes relevant. But to me, it becomes relevant in the sense that it says we should take those things into account when ascribing responsibility to people who vary in these ways that we already talked about, rather than saying neuroscience says no one has any free will ever, that doesn't seem like a useful step to make because then what are we doing, right? Like everyone's at zero. How are you comparing someone with schizophrenia versus someone without? That doesn't make any sense to me. There's a few other places where it really does kind of cash out and one of them is actually in AI. So we're building these systems right now that are getting so complex that they're forcing us to think about our own complexity. And there's two ways you could look at that. One is you could look at chat GPT and say, wow, this is amazing. Look how it can have these things that look like a really good conversation. It even looks like it's understanding things. And yet we know it's just these artificial neurons that are trained and trained and so on. So there's no life in there. So maybe either it really is a conscious thing like us or we're just a machine like it. So it's either like elevates the artificial thing or it deflates our own powers on the degree to which we're impressed with ourselves. So it becomes interesting then to think, okay, well, like I don't think chat GPT is conscious. It couldn't be conscious of anything because it doesn't have any connection to the world in my view anyway. It would be interesting to ask, well, what would it take to make an artificial agent? You know, this is where we started talking about robots that can control themselves. Well, if you built a robot that has all of these systems that we talked about, right? These cognitive systems, systems for planning, systems for accumulating knowledge, systems for caring about what happens to it. So things are salient to it for its own survival, say, would that be an agent? I think it probably would. You know, would it be a living entity? If we get there, we'll really be forced to wrestle with that question. And I think actually we probably will get there first in virtual environments. I think there's a absolutely high likelihood that we will have some virtual agents within virtual environments like in Minecraft or something like that where, you know, at some point you would say, look, this thing has all of the characteristics that we use to define our own agency or at least the agency of an animal, say. Is that the step to that medical ignition? Is that the key piece? It's certainly that's one way to think of it. It's a little bit of an anthropocentric way of thinking about it to say, well, obviously the step that got us where we are is the most important step. Whereas if you're having that discussion with an octopus, they might say, well, hang on a sec, you know, that's a little human centered. What about octopus intelligence or octopus consciousness or octopus agency? And so your Minecraft agent might have a discussion with you the same way to say, well, hey, yeah, I may not be conscious in the same way you are, but I have some degree of consciousness or some degree of agency that at some point we might seem worthy of respect or worthy of even calling it personhood of some kind, you know, in the same way that we have arguments about whether dolphins or elephants or crows or octopuses are worthy of personhood. But do we know why we have that extra level? Why we're able to think about how we think? Do we know how we got there? The simplest explanation is that we have, you know, our brain is really as these sort of our cortex, at least, is organized as these different areas that are arranged in a roughly hierarchical sense. And the more levels of the hierarchy you build, the more each level can integrate information from below. And so for example, in our visual system, in our retina, the individual photo receptors in your eye are either being hit by photons or not, right? So they're either being activated or not. Now, if that's all you knew of the world, you wouldn't know anything because you need two bits of information, which photo receptors are active and where are they relative to each other? So they don't know where they are relative to each other. It needs another level to integrate that information. So you need some neurons that are looking at this pattern of the photo receptors and integrating this patch here or this patch or this patch, right? And then they can say, you know what? There's an edge in the world. There's a place where the illumination changes. So our visual system is built from a hierarchy of levels like that, each of them drawing higher order information by integrating the signals that are represented by the ones below. So we go from representing things like just edges to forms, shapes, objects, types of objects and then categories and so on. So we go from percepts up to concepts. We get more and more abstract and we can represent faces, for example, so the particular part of the brain that represents faces. So when it's active, it means it hasn't meaning to the organism that there's a face out in the world. And even if the face turns like this or like that, so the visual information is really different. At a conceptual level, we can still represent it as the face of the same person and we can link it to our knowledge of the person, except in some people who have face blindness who can't do that, right? But again, that just shows that's a real biological process with some underpinnings that we can understand. So if you continue that logic of a hierarchical system where the higher levels are abstracting higher order information, then ultimately you get metacognition as the hierarchy of decision-making where the metacognitive level is now abstracting the highest level, most abstract conceptual kind of level of decision-making. And so probably that just happened as our prefrontal cortex got bigger and bigger and as it got bigger, it split into separate areas and they formed this kind of at least semi-hierarchical relationship with each other. And so we know that that's the case because damage to your prefrontal cortex in various parts of it will impair Medi-cognition, they'll impair executive function, they make it impossible to plan, impossible to control impulses, impossible to regulate your working memory, to switch from thinking about one task to thinking about another task, and so on. So there's a good neural understanding of executive function and the Medi-cognition that it depends on. So that's really interesting and I mean this was a mix since, doesn't it? At the end of the day you still just got another layer, I mean this is clearly not right, but it's like having another layer of neurons above the neuron, you already got to process that new information, which is a highly, as you say, highly conceptual, it's not a cell in your eye, basically do, say your mother's a photo on there or not, but it's the same sort of process that gets you there. So go into the free will, right? What is the difference between say you or I when that layer below throws up, hey John, there's three things you should talk about next and hey Kevin, here's three things you should talk about next. What is the difference between me deciding on one of those and you deciding on one of those and because that's what we call free will presumably is that that's example of free will to make that decision. What's the difference there? So I think you're touching on the individual differences between people and what I would say is that the way that we decide things, absolutely of course is idiosyncratic, that's why we know that different people behave differently, right? So that comes about from probably at least four different sources. I would say one is just evolution, right? So we all have inherited human nature generally and different organisms have their own different natures which informs how they behave. We have genetic differences which are sort of inevitable, right? We don't all have the same genome. So we don't all start with the same individual nature in terms of psychological predispositions and I would add into that that actually there's a lot of developmental variation that just happens even in identical twins or clones in other organisms. They can have the exact same genome, they won't develop the exact same, right? So their brains of identical twins by the time they're born are already different from each other. I've heard you speak about this before they're laying the neurons down differently. Oh, there's variation in how they're laying neurons down. Absolutely. So there's variation in the genome kind of encodes some rules for how that happens, but there's a lot of indeterminacy in how that goes. So again, it's sort of like we were talking about with a physics example of the future of the system not being predetermined. It's a process that gets it there, but it could go lots of ways. But the point is we start with some innate predispositions and then we have some experiences that happen to us that we learn about. But then we have the really important bit is the accumulation of our own choices through time because every time we're making a choice, we may be selecting an environment, we may be choosing an experience, we may be altering things in ways that will feedback to inform our memories and our policies and our habits and our commitments and our projects through time. Right? All of the things that collectively along with our basic personality traits make up what we call our character. And our character is just much more than whether we're shy or outgoing or the neurotic or those are very basal starting points, but our character is really much more sort of situation dependent adaptations to the world that we found ourselves in and that we have actively navigated through. It's not just things have passively happened to us. We've been making choices all along. You know, this gets to a point where something will say, okay, look, yes, you can do things for reasons. You can do what you want, but you can't want what you want. You can't decide what you want to do. This is sort of infinite regress that says, oh, you know, ultimately all your decisions are just because you were born this way and these things happen to you and you're configured a certain way you are, you are the way you are and you can't do anything about it. And is that saying that you have no free will? Yes. The next decision is what the program says you're going to have. Exactly. So this is a kind of determinism, which is like the highest level determinism, right? It's saying even if the lower levels are indeterministic physically, the highest levels are deterministic in the sense that every action that you do issues fully from your current character and your current character just is entailed by the physical configuration of your brain right now at this instant. Here's the important bit and you had nothing to do with the way your brain got configured. That's the bit that's important for moral responsibility. And actually, this is not a claim about free will. It's a claim about moral responsibility. In a sense, it's saying, yes, you can control your actions for your own reasons, which is what most people's definition of free will would be, but you can't be held morally responsible for that because those reasons, you came by them just by past experience. So it's a weird dissociation like it admits free will is right, but then says there's no moral responsibility even still. And actually, I think that that's just wrong. So in the first place, we choose goals. We're not just choosing actions. In fact, we're never really choosing actions. That's not the decision. The decision is say, I want to I'm choosing to take a drink of water. Well, then I'm going to tell my motor cortex to reach out and get it. I'm not going to tell it how to do that. It knows how to do it. I don't have to tell it how to do that. You know, when we choose a goal, we absolutely are choosing our reasons. But if I choose to play around a golf, I'm absolutely choosing what my reasons will be for actions over the next four hours or however long it takes. But the other thing is we are capable of thinking about our own character and we do. And we think about other people's character and we tell them about it all the time. So especially children, because you know, as children are growing up, we're giving them moral instruction. We're saying, no, you have to share. Don't be selfish. Take turns. Tidy up. Be responsible. We're basically inculcating pro-social norms and conventions, which eventually get sort of internalized such that they become guides to the individual's behavior without anyone having to be around to tell them what to do at every moment. So we internalize those kinds of moral policies and eventually because we have metacognition and because we have metavolition, we can ask, what are my motivations here, like an actor, right? What's my motivation? And is it a good one? Right? We can ask that higher order question because we can look down on those moral policies that we have and say, you know what? Maybe I would like to do this. I find a wallet on the ground and I'd like to pick it up and take it. But maybe I should turn it in somewhere. So we have a higher level constraint on your behavior. But then you might have a higher level above that and say, well, do I want to be the type of person who is honest like that even when nobody's looking? And so I think we do that all the time. Some people do it maybe more than others, but I don't think it's right to say that we have had no conscious hand in the way that our own character emerged through time. Wow. And I mean, what strikes me about what you're saying there is it's the level of complexity above the basic decision that that too is not a solo direction. It's a suite of directions. And that has got so many inputs to wish your history, your experience, what you were told, your moral code of even use that term. And that we would call the decision that comes out of that process, the free will. That is free will in and of itself. Yeah, I think that that's right. And you know what one way to see that is that your free will is not this just capacity that is exercised in an instant and then exercised again in the next instant and the next and the next. It's this continuous thing. You know, we talked about the continuous behavioral control of selecting goals and plans and activities and behaviors through time, but it's also this continuous sort of reflexive reflective thing to say while you're making those decisions, you're thinking about are these good decisions that I'm making, but also am I making them for good reasons or not? So it's an extra level on top of that. It's not just are things turning out well for me, but have I calibrated my goals correctly, my incentives correctly? You know, so that extra level of being able to like I said kind of look down the top levels of the system can look down at the rest of the system and see its operation and then make some judgments on it. That to me is the ultimate expression of free will, but it's one that's expressed continuously, not just instantaneously. And that's presumably then your view differs with the deterministic mechanical view that when we come to the law and deciding whether the person had free will, that process in and of itself for it seems to me you're saying that for you that is proof there is free will and that is good theory for why there's free will. You've got free will and you should be held accountable for those processes that you're going through at that level. Yeah, I think you can be held accountable for your actions, but also held accountable for the reasons for your actions. And in fact, that's like a lot of legal argument is about that, right? But also, you know, to some degree, you can be held accountable for the actions that you took that led you up to that point or that led you to being a certain type of person with certain types of habits. on. Now, I think it's really important to emphasize that I'm talking about these things in the abstract as capacities that we all could exercise, but it's really important to emphasize that some people may be able to exercise them more than others. And so, you know, some people may be with very high impulsivity, for example, may just not be able to do that as well as somebody else. They can't take that reflexive control because they're driven much more by very immediate kinds of motivations. And some people in some social circumstances just don't have the wherewithal of the opportunities to develop those capacities in the same way as somebody else. So, I think it's important just because we can say that free will exists in that sense as a capacity doesn't mean that we should judge everybody in the same way because they may not all have the same opportunities to exercise that capacity either due to constitutional differences in the way they're made up or due to social circumstances that they find themselves in. So, just to finish, Kevin, what would you say to the listener in terms of what this means for their free will and how they should think about free will in themselves? You know what, I've tried to avoid giving any kind of self-help advice from this stuff, but I mean, generally speaking, I'm in favor of the idea that greater self-knowledge is just a good thing and how people choose to use that or think about it is up to them. I can say for myself that the exercise of going through writing this book and doing the research for it has been really, really interesting. I don't know that there's any tricks I've learned to be able to exercise better free will, but I definitely think about my own decisions in different ways, sometimes in quite abstract ways where I'm asking, why did I do that? Was that really this habitual thing? I think that's interesting and useful to keep in mind. And I guess maybe the important thing would be to realize that different people, those differences between people, maybe more pervasive than we think. And that can be a useful thing to keep in mind. When you see somebody who's done something and you say, "God, why did you do that?" You shouldn't have done that. It was the wrong thing to do or it was just a stupid thing to do or whatever, occasionally thinking, "Well, okay, well, there's some machinery here, even though I don't like to reduce it to that, but there can be differences in how the systems work that can contribute to that and maybe occasionally cutting people some slack is maybe a good advice." On that note, Kevin, thank you very, very much for being on the show. It's been fantastic. Thanks, Sean. My pleasure. Thanks. Thanks for listening to Simplifying Complexity, where we look at the key concepts of complexity science with expert minds from across the world. Concepts like emergence, self-organization, adaptation, networks, scaling, tipping points, and much more. This podcast was produced by Brady Aeword and Wevelyn Creative. To make sure you don't miss an episode, be sure to subscribe to or follow the show in your podcast app. I'm Sean Brady and I'll see you in our next episode.