Agency - Part Two

By Stephen Downes
Jan 05, 2022
Transcript of Agency - Part Two

Unedited audio transcript

Hello, and welcome to ethics analytics and the duty of care. We're in module eight, ethical practices and learning analytics. And we're working our way toward the end. Still got a little ways to go in the second part of our discussion of agency and the role agency plays in our creation of ethical practices in learning analytics.

And the previous video, I talked quite a bit about different conceptions of agency and how these change, how we think about ethics in general and ethics in relation to AI and analytics, and I gave for basic concepts of agency. This is neither exhaust ignore inclusive agency might include these four mighty code.

Other things he's might not be the best ways of talking about the conceptions of agency. But the mean intent here is to get across the idea that we can think of agency in different ways we can think of it in terms of power or the yoke comes either in power.

And since I'm pushing, or making things happen or power in a sense of creating limitations and what we can do, we can think of it in terms of competences, are capacities to do things green, think of it. In terms of how we feel about ourselves, what we can do, our place in the world etc, or we can think of agency as ownership, taking ownership over our actions, taking credit where it's due etc.

These conceptions and other similar to, it can be found in a lot of the discussions of agency. Just as an example of that, I've put up here a diagram from Gail Jenkins from 2019, who looking, at teacher agency the effects of passive and active responses, to curriculum change. Things like self-reflectiveness intentionality for thought reactiveness.

We can see how these conceptions of agency can be seen to play different roles in our understanding of teacher agency. I also talked about how our understanding of what learning is has an influence over our understanding of the role that agency plays. And I mentioned Benward word's article from well, I guess it's yesterday.

Now, 16 hours ago on the performative demonstration of education, which is a, what a lot of the content based ideal-based models of learning look like, or our based on their more about the educational provider, whether it's a company, or a school or an institution, proving that some content has been learned.

Then it is about whether the learner actually learns anything. The contrast here was with what I call free learning or personal learning where the individual themselves. The learners themselves are in control of the learning and this is significant because when we think of the relation between ethics and agency, the more agency you have over your own learning, the more the ethics of that learning is in your own control, but if all of the agency is in the hands of the institution, or the teacher, or the artificial intelligence, providing the learning, then the ethics of these becomes, much more important.

The ethics of the institution, the teacher or the AI, providing the learning becomes crucial. So that's where we left off. And this part of the presentation, I want to look at a bit about where learning or where agency comes from. But before that, just to underline the point. Let's begin looking ever.

So briefly about the subject of science literacy and this underlines the point about the ethics in the agency. So this is ethan segal writing back in 2021. In other words a few days ago and he could be talking about the chart of I just showed he writes for decades.

We've been mistakenly measuring science literacy by measuring which specific facts people of memorized instead, developing an awareness of an appreciation for science. And what it does for society is a far more impactful approach. Science is our best tool for separating what is true from what is not and we abandon it at our own parallel as a society.

No, he's talking about science and science literacy. But I interpret this in terms of agency and who is actually able to be responsible in the ethically responsible for determining, what is true and what is not true. If it's all the artificial intelligence, or the teacher or the institution, then we're placing on them.

The oneness of determining what is true? And what is not true and that makes the ethics of that determination significant. If one, the other hand, it's up to individuals to develop for themselves, the capacity through science and I'm not sorry about making stuff up here. But through science through an appreciation of science through the application of scientific method for themselves, whatever that may be, then we're removing the need to be so ethically, stringent on the part of educations and information providers.

I know it sounds like kind of a slippery argument here, but part of the argument here is along the lines of where would you rather have that ethical responsibility base? Would you rather put it in the hands of the institutions and the artificial intelligences? Or would you rather put it in the hands of individuals and indeed forgive us science for a second.

Just think about literacy generally, right? If we remove all the choice from people and tell them what to memorize, they're not ethically responsible for what they know at all. And so they're not ethically responsible for what they believe. They're not ethically responsible for how they vote. They don't really have any agency and that's why they're not ethically responsible, all of the responsibility.

Now falls into the hands of these agencies and it's arguable that these agencies aren't actually capable of having ethics argue. I'm not I'm not saying I'm making that argument here but it's an argument worth considering Certainly if you believe and only humans can have ethical standing and then you're probably not happy about institutions and machines telling you what's true?

And what's not true, possibly, you're not even happy about the institution of science. Telling you what's true and what's not true? That can also lead to bad consequences. So, you know, it's not going to be a nice simple picture that we're talking about here. But certainly the ethics and the agency of one's capacity, to deal with facts and information about the world are closely tied together.

And you can't really separate them. Okay, pause for a fact. Where does this agency come from Really? I mean, without a friend, we Venus. Well, a lot of people think that it comes from biology and there's an argument for that, but let's suppose we agree that it comes from biology, you know.

Let's suppose that it is our nature, whatever that is that gives us the capacity of agency. However, we can screw that. What does that mean? Well, we could go back to the old belief, desire and tension model. And that's basically the approach that people like Alan Newell and and Herbert Simon.

And others took new only introduced the idea of what might be called the control system or these days. We think of it as the executive function and the executive function is what I actually controls agency for us and the various aspects of the executive function would include things like activation focus effort, emotion, memory and action.

But then there's another way of looking at this. Where, There isn't actually a mechanism of decision or control that exists and we'll draw the analogy here with evolution because that's a way we can make a similar sort of mistake. So evolution, we think of, as the survival of the fittest, right?

So we get stories, like the bird developed its wings, so it could escape predators. Now, I don't know if that's a real story or not. Okay, We just take that as a sample story, substitute an actual real story, the argument will still work. So we can certainly say the the bird having wings, right?

We can certainly say, having wings definitely gives you a survival advantage because it definitely makes it easier to escape predators. So yeah, it seems like a pretty good explanation, right? Survival of the fittest means that things that are able to escape predators more effectively or more likely to survive and more likely to pass on their genes.

So they're more likely to pass on wing growing genes. Make sense. But let's go back to that original formulation birds developed wings, so they could become more effective at avoiding predators. Now, there's this aspect of intentionality here that isn't actually present in evolution. There's the suggestion, it's only implicit but I think it's taken as literal by a lot of people and that could be wrong, but I think it is where it's some sort of action that birds took some sort of decision that birds made to develop wings in order to escape predators.

It was, it's almost like, you know, there, I wish I could escape predators. What can I do? Oh, I know I'll grow wings. That's not how it works. Is it have evolution? Is isn't gold directed. It isn't intentionally. Directed is noble rights to describe the state of affairs. There is no actual selection, carried out by natural selection.

We're not actually picking, you know. Yeah. Wings those would be useful maximum, which actually happening nature. In this case is simply a passive filter. If you have birds, they develop any number of different things. Wings horns. Tusks whatever. Right. Third eyes. You know, and the ones that give them an advantage, are the ones that stick around.

So, forget the wind the horns, forget the tusks but the wings, they provide you. The advantage they stick around. Evolution is the filter, not the selection mechanism, The selection is at random. So,

Was suggestion is maybe that pictures, not right either. And this is where Dennis Lowe comes in. He says, well maybe the there is a role of agency and evolution and there are different ways we can look at this. We'll explore some of these. But if you think about it entities that have agency might be like birds that have wings, they might be more capable of survival.

So, the origin of agency, might be evolution. Certainly something worth considering.

Okay. What do we mean by that? Well, let's go back to what we think of as the control system that Allen Newell talked of it because that's kind of what we're talking about except that maybe it isn't, right? So what does a universal theory of control have to entail?

Well there's an awful lot of stuffing in here. Newell lists the following behave. Flexibly is a function of the environment exhibit. Adaptive behavior operating in real time. Operating rich, complex detailed environments, use symbols and abstractions use language. Learn from the environment and from experience acquire capabilities through development and operate autonomously.

But within a social community, be self-aware and have a sense of self be realizeable, as a neural system, be constructable by an embryologist called growth process and arise through evolution. Now, the question here we ask is do we need all of this in order to have agency? Well no and and that's what a lot of the the non-representationalist non-cognitive approaches will say, and in particular, it's the prescriptive things like you symbols and abstractions and use language or even operate within a social community.

These seem to be optional. Others seem to be requirement insofar as we're talking about humans, like be realized about as a neural system or be constructile constructable by an embryological growth process, right? That's kind of contingent on being human but we could maybe imagine something realizable as a connectionist system constructed by human computer constructors.

Say some might be epiphenomenal, they might arise as a consequence of having agency, but not be essential to it, for example, being self-aware, and having a sense of self or operating in real time, you know, perhaps that might be the case. These might happen because a system has agency, but they might not actually be a part of agency itself and then others might be what we might call performative requirements.

Like behave flexibly, adaptive behavior, operating and rich, complex environments. These are performative requirements, right? These are based on a concept of agency, as Well, as a power, perhaps or as competence. So I think that the theory based on a theory of control, or a theory of executive function is bringing a lot more into the picture than we need in order to have agency and a person in an animal, even a thing.

Okay. Well, what do we actually need? One way of getting at this is offered pine masculos hierarchy.

And I think most people are familiar with masculine's hierarchy.

And which useful about it is that it gets at the idea that there are needs survival, needs that are met by means of agency. So what are these survival meets? It's tempting to talk, simply about basic physiological, needs air, water food shelter, etc, but it's nice little points out as organisms become more complex and especially, they become more.

Social, the needs expand as well. So there are physiological needs, there are safety needs, but then there are needs that are more characteristic of social beings like love and belonging. And then there are needs that are more characteristic of being that are self-aware. Needs like, respect, self-esteem, status etc.

And then finally self-actualization. We could ask whether this is actually a hierarchy. We can ask whether some of these are actually needs. I know a lot of people are really expressive of the need for self-actualization and, you know, even that seems to suggest some sort of directionality, that, that a lot of people don't necessarily have and some of these might be more desires than needs.

We might desire respect, but is it true that we need it in order to survive? Maybe for some people but for everybody, but the main point here we shouldn't be lost. Is that entities have needs. And these needs can vary from very simple to very complex. And these needs can vary in intensity.

They can vary in the way they are actualized etc.

Another way of talking about the same sort of thing, might be to try to get it to the most basic components possible. So here we have regio and terrazza and 2016. Beginning with a model of agency based on some very basic needs and they basically draw out two basic needs.

One of them is coherence for something to actually be an organism, it needs to have what they say is a substantially harmonious and atomic and functional structure new. They're different ways of thinking about you know, particularly if we think about entities that are not human like social organizations or computers or whatever.

So having a substantially harmonious anatomic and functional structure, might not mean, you know, have a cell wall or have a skin but it does mean to have some kind of structure such that we as observers might say yeah that's a thing. And you know, in the game I've kind of changed a little bit.

Their definition, their definition is based on what it is, my definition is based on what we would call it. But nonetheless, the core aspect of that is the same. The other thing is autonomy and autonomy is described by them as living organisms, create and maintain and internal environment, which follows dynamics of its own.

And I find this really interesting because autonomy here now is not talked about, with reference to the external world, in the sense of what it can do in the external world is, you know, it's not based on, it can go where at once it can do what it wants.

It's basically the idea that the entity Well, you know, create some maintains an internal environment. It is self-contained and it's internal behavior, is internally defined? That's, you know. Now, strictly speaking, there's no such a no such entity if there were then, perpetual motion would be a thing. Excuse me.

But all entities have inputs and outputs of some sort, you know, to have an internal environment that is not completely static like the internal state of Iraq. You require energy, which very often means matter, which is converted to energy in some way or another, my decomposition, or whatever. And I don't know of any such process that doesn't also produce waste.

So, input notebook, right? But was definitive of something that has agencies. Something that has autonomy is that their internal states are not one hundred percent defined by the input in the output. I'll put it that way. So for example, in the case of a cell in the internal structure of the cell is determined by the DNA, which is, you know, the we blueprints of how it is created and then how that DNA expresses is expressed in the world, what depends on input note, put, but the internal environment isn't based 100% on that.

I'm putting notebook. So with some caveats, then we can talk about agency as being minimally based on coherence. And autonomy, you know, I'm just floating a theory here, right? It doesn't matter, whether it's true. What matters is, is this a good way of thinking about it or? No, what matters is can we think of it this way?

And then we'll sweat the details later because the point now can be made coherence and autonomy are things that we could reasonably expect through. Say an evolutionary process, there are also things that we could expect reasonably in non-human entities. So there's allows us to have non-human and and, you know, even soft created or naturally created things that have agency that aren't us.

So what does it take to maintain coherence and autonomy, So Brazil and thoracic continue. First of all, we create and maintain a permeable separation between the internal environment. And the external mu we exploit features of the environment relevant to the organism and perhaps we exploit the biological behavior for example, features of other individuals.

And then there are some organizational archetypes. We can. Look at that. Describe how this sort of coherence and autonomy is maintained well. Again creating and maintaining a permeable, separation can be thought of in two ways, one way it can be thought of is, as you know it's in the central property of the organism itself.

It is a skin. It is an objective barrier or it can be thought of as something more conceptual where an external observer can say. Oh yeah I see that there is a boundary here between these things. And part of agency though is that there's boundary is maintained by the organism itself.

So like for example if you're if you're driving through Canada from south to north, you're driving through a temperate forest for a while and then there's a really clear line where you cross that line and now you're into arctic forest. So that's a boundary but it's not a boundary that's mean by either of these forests.

It's a boundary that's created by the soil conditions, the difference between Canadian shield and sedimentary deposits, and by climate, but by contrast, the skin of a person or the cell wall of the cell, these are actually maintained by the organisms themselves. So the maintenance of the boundaries and important aspect of agency to have agency, you need to be a distinct entity at where this distinctness is created by a maintained by yourself.

Okay? But we, we can still say that of all kinds of things, right? You know, I used the phrase, distinct entity, very deliberately because in Canada, we have this idea of Quebec or at least Frankfung Quebec as a distinct society and it is a distinct society. Not because I say it is but because that society itself means that it is a distinct society, so it's generating, it's own agency, it's generating, it's own coherence.

And autonomy by maintaining this separation, we are a distinct society. Okay, so, take that idea. The idea of agency as something that develops through evolution and with it, the idea of agency as intimately, connected with ethics immorality, and you got the idea that morality could be thought of as a type of evolution.

Now, the headline there says social evolution because for our current discussion, we're still talking about culture and cultural practices and in our current discussion where therefore thinking of ethics as something that is produced by a culture. And in particular through the concept that agency in that culture. So let's think about that.

A Robin Dunbar? Proposes. Something called the social brain hypothesis where, and there's two sides to it. First of all, there's the idea that because we live in a society there is an evolutionary, there's an evolutionary advantage to as it's put here, acquiring social intelligence. On the other hand we could say that morality is something that evolves as our sense of agency evolves.

As we become more effective agents. Whoops, we become more effective, ethical beings. Both of those are interesting approaches. And interestingly, they kind of map to the the ethical theories that we've already talked about. So, for example, let's take three types of behaviors here. And they're listed here. Moral acts, which corresponds to the moral philosophy of consequentialism, which might be based on instinct and behavioral genetics moral motives or intent in the United States a bit different, right?

That's a lot different than the moral philosophy is very, different is stay ontology. The philosophy of manual content others, the biological science behind that is psychology and maybe neurophysiology then, we have more systems, the artifact of ethics, the moral philosophy that that corresponds to is the social contract and the biology that that may be related to is sociology and communication.

And I think that's interesting. There's now, we have remember, we had the different types of agency. We have the different types of morality that almost kind of, pretty much map to the different kinds of agencies. Right. Agency is power is consequentialism agency as feeling or maybe capacities corresponds with day ontology agency as ownership corresponds with social contract.

Arguably each of those three things, has an origin in the biological sciences and arguably as a result, they develop through a process of evolution. So, the argument here is that it's an evolutionary process that leads to different aspects of agency and leads to different aspects of ethics, and we can think of this either in terms of ourselves as physical objects.

We can think of there's ourselves as cognitive entities, or we can think of this as ourselves as collectives like societies or governments or agencies. There's a evolutionary theory for everyone. So there's been a lot of work done on this. And I found this chart on an article in Biomed Central that looks at the evolution of morality.

And this one, look let's look at some of the, the elements of ethics that that this chart picks out. And I find this sort of odd, but it may learn distinction that is ethics as an innate versus a learn distinction relevance of learning to morality. Only one study picks up on that kin.

So action seems to be very important to be able. Reciprocal altruism is important to some people cultural evolution the idea of a culture being subjected to similarly similar evolutionary processes rewards sanctions, nobody's interested in now, which is interesting nature, nurture, caveat mentioned by a bunch of people. Emergent properties may be mentioned by one and then discussion of a reductionist bias mentioned by well about half of them.

So all I want to say here is even if we say morality or something that's evolved through something like the well-known, the revolution, our understanding of it is still all over the map.

But it's interesting the relation between biology and knowledge, especially once we start saying, well, you know, there's a relation between, how we came to be what our, what our nature is, what our biology is. And what we can come to know what we can come to believe, what we can come to feel is right or wrong, and there's a bit of a danger here.

And then Williams been Williamson points us to them points to a couple of areas where, you know, the ethical risks seems greater than the actual benefits. So let's look at them, briefly on one hand a data intensive bioinformatics driven approach to education remains in development. Basically using a person's genome to identify their learning needs and potential.

Now even if we agree that agency and knowledge etc, or even, you know our capacities are biologically determined, this sort of approach is the sort of approach that I defined earlier as control learning, right. And there's a difference, I think between determining what a person's biological state is and then defining, what their education should be based on that versus having a biological state and being that biological entity determining for yourself, what your education will be to very different things.

And there are substantial. Ethical risks, in trying to derive a person's effort, or a person's educational needs from their biology. I think that should be heavenly by any examination of those two models, because really, it takes again, the agency out of the equation person can't do anything about their biology.

So, they're completely passive. And this environment, which means all of the ethical responsibility falls on the bioinformatics driven approach. Second thing a set of affect aware technologies to gauge and respond to student emotional states that aims to make emotional life machine readable. And to control engineer reshape and modulate human behavior.

Again, same kind of thing extent in except instead of using biological information to manage a person's education, where using emotional and cognitive states to manage a person's education. Same risks, supply. And as in a site this is where I think a lot of the objection to learning styles theory comes in because learning styles are thought of as being applied in the same way, right?

You look at somebody's learning styles, you study it through a test, maybe you find out that their INFP or whatever and then you shape the education, they will receive based on that running style and the argument is on the part of the critics that there's no improvement in their educational outcomes based on that method.

And that's seems to make sense to me. Similarly it makes it seems equally likely that neither of these two projects that then Williamson describes would produce differences in outcomes. Either it doesn't fall from that though, that there are no bioinformatics about a person. It doesn't follow from that. There are no students, emotional states and it doesn't follow from that that there are no learning styles.

They were just the response is out of proportion to the criticism but the problem isn't him? No learning styles. The problem is in the model of learning that is based on the idea that you can do some kind of set of measurements or evaluation or determination of the physical state of a person and determine for them what they need and it's a problem.

Not because you can't do it. It's a problem because you have now assumed and ethical responsibility that you can't fulfill. Even if you can shape the outcome, you are not in a position to know that this is the right outcome. You don't have that ethical capacity to make that call.

Why? Well again we go back to the ethics of care and all of that, the people who it involves are, the only people who have the right to make that call. Right. Agency is necessary for ethics, but agency is also essential or ethics is essential to. Ah, try to see that in the written version.

All express that better. But the idea again here is that a person is responsible for the wrong learning only if they have control over their own learning,

We could think of agency as a way a culture evolves. Now, this is an idea of agency as something that is socially constructed. We could say that there isn't really a thing out there called agency, even if we can reduce it to things like, basic elements, of coherence and autonomy and we can point to the objective existence of coherence and autonomy.

But really, we've built this set of artifacts around that as a society which are perhaps expressed by the the conditions that Ellenuel describes, you know, it, perhaps described as this executive function, but this executive function properly. So called is a cultural construct and that's where we get point of view like Kimleys quote, human life waste dependent on cognitive capital that has typically been built over many generations.

This process of gradually of gradual, accumulation produces and adaptive fit between human agents and their environments and adaptive fit that is the result of hidden hand evolutionary mechanisms, a bunch of stuffs going on here. So, one thing we have is this idea of cultural evolution. Now cultural evolutionists tricky, if we apply it directly to, you know, if we apply the analogy directly to culture, then basically cultural evolution is the survival of the fittest, a applied to cultures.

Which means that the cultures that survived were by that fact the fittest which kind of runs counter to our intuition that you can't really compare cultures that way. It's not a case that they're competing for survival. Although you know, I mean, certainly there's this you know, this strong street of social Darwinism that that runs through society and the idea that, you know, not just cultures but say languages or other cultural aspects, you know, everything from from games to social more race to types of democratic organization.

All of these are in some way shaped by this survival of the fittest. The other thing that's going on here is the idea of the the hidden hand, which as we know is a reference to Adam Smith's hidden hand of the marketplace. And again, the marketplace is depicted as something.

That's competitive. And so once again you have the idea of culture as evolving as a consequence of competition,

There are other ways of looking at this but, you know, we have this model. So we have, you know, a contrast here, the human genetics, which is simply the relationship between parents and the offspring, but then human culture, which has all kinds of influences on the offspring, as a result of their peers, social norms parents relations etc.

And the question that comes up is is actual evolution of properties like agency the results, only of the one type of evolution, the human genes type of evolution or is it also a product of cultural evolution? Well, the idea about it is known as the duel inheritance theory. So dual inheritance theory models, cultural evolution, I'm reading here closely on genetic evolution and cultural inheritance closely on biological inheritance.

The central idea is that there are features of human psychology, most obviously imitation learning and language and features of human social environments. Most obviously long periods of juvenile dependence that result in a high fidelity flow of ideas from parents to their off school offspring.

And I find this in interesting theory and I find this a dangerous theory to be quite frank.

Because again, it's not clear to me that cultural survival is the same as biological survival and it's not clear to me that the same mechanisms are in place the same mechanisms that would inform one inform the other. But also two the way this picture is drawn, you have a mechanism to go from what the culture is around to the individual survival of the organism.

And so, you get the idea that some psychological traits, say, imitation learning are the result of cultural factors. I have a bit of a hard time with that but I can't prove that it's not true either.

And instance of this.

Yeah things I guess I could use the word an instance of this is the idea of being inheritance or you know the cultural or the evolution of means Richard Dawkins for example writing about the selfish gene or Susan Blackmore. Pneumatics. Does provide a useful way of understanding cultural evolution.

So both dual inheritance and mean-based theories. Share, the idea, the cultural transmission is both important and accurate. Okay, Dawkins calls the information that is copied. The replicator and contrast this with the vehicles that carry the replicators. So, the vehicle would be a human and the the idea the information he says, in my view, misusing already information but that's in the side that the replicated thing that becomes replicated.

For example, Garfield loves the lasagna. So, humans carry the idea of the Garfield. Loves lasagna. The idea that Garfield gloves was on you compets against other ideas out there in the world and the most fit of those ideas survives. Now, on the dual inheritance series, the biological fitness of the human agents.

Had explains the spread of these memes or of cultural variance, right? But most of culture can't be explained that way as we can ask, what is the fitness for a short? Melody, is it the ability to survive and reproduce of? There's so much of culture that really has nothing to do with the survivability of the humans.

In that culture, including the idea of a Garfield loves lasagna. Right. A person that believes, or at least contains the idea, Garfield loves lasagna, is neither more or less likely to survive than a person that doesn't. And the idea itself that Garfield loves lasagna has no more. Eminent survive.

The ability than the idea that Garfield loves pizza. All right, so the three it just doesn't seem to apply. And I think the reason why it doesn't apply is because the model of competition for survival is it's self incorrect when it comes to culture.

This whole picture is kind of questionable but I need to bring it out here to make it clear. So this image it's shows up all over the place. I got this one from Douglas, Halchin in the article, teaching the evolutionary sort of the evolution of morality status and resources.

And it's this kind of, it's not reductionist but it's sort of reductionist model of how evolution maybe applies to each of the different levels of entities in the world from matter, to life, to mind or psychology to culture. And we could talk about these as Daniel Dennett does, as, you know, different levels or domains of discourse.

We could, for example, when talking about mind, take what he calls the intentional stance. So we could, for example, talk about culture by talking about the social stands, but when we begin making inferences up and down this chart here, from culture to mind from mind, to life from life, to matter, and even vice versa, that's when things become a bit questionable because there's no reason to believe.

For example, that statements that we make about culture are reducible to statements that we can make about mind. For example, language as big consigned showed is a property of culture. It's a thing that exists out there in the world. It does not follow from that, that language also exists in the mind and certainly does not follow that language also exists in biology or that it exists in matter, you know, I mean quarks don't behave the way propositions.

Do we could perhaps, impose a requirement of coherence on all of this, or perhaps even say create a semantics of culture to mind a semantics of mind to life, or something like that, to draw relations between them. But this becomes a very tricky enterprise. It's not clear. Certainly that this enterprise helps us understand the concept of agency.

First of all, I'm the concept of morality generally, you know, and Douglas Alchemy. Here's the criticism, the ultra reductionism and implicit promotion of competition. That once dominated the field yielded to more balanced perspectives and more nuanced interpretations reductionistic bias, varies, but the text basically omit the concept of emergent properties, or new levels of organization the physiola or at the psychological and social levels.

They do not. For example, describe how social rewards or sanctions can regulate quotes selfish behavior or individual quoting as. For example, observed in food, sharing among vampire bats.

And that's the weakness of using an evolutionary account of agency. It's not clear that evolution explains agency because it's like clear that agency is based in reduction. You know, it's not clear that we can reduce agency to biological properties and more to the point. It's not clear that agency is I mean inherently competitive thing.

Our agency does not entail competition with other agents.

And when we look at agency as you know, as a tool of morality, when we tools the wrong word there, we're looking at it, not as something that develops naturally, on its own note of evolution and survival of fetus and competition. And all of that. Again, this is a lesson that the ethics of care teaches us.

Agency in so far as it is described as power or described as capabilities or feelings or ownership is instantiated or expressed differently in different people, depending on their own conditions and it's the sort of thing that if we want it to exist. Which if we're interested in ethics, we want agency to exist, then it needs to be supported in some way.

And that's why this whole module is about ethical practices. What are the practices that we would undertake to support agency, to allow people to develop their own ethics and not just make up ethics. But actually develop ethics Well, based on something. I wrote a while back called what piece means to me.

This was based on my experience, listening to stories of people who are impacted by the war in Colombia. Describing what the war took away from them and what it took away from them was their agency. And so I came up with these things based on that that described the sorts of things that we should be supporting that support agency.

Now again it's it's a two-dimensional thing. It's not a deep theory or anything like that but it's a first blush. Look at the kind of practices that we could undertake socially that support agency and therefore, support, ethics, ethics and individuals and maybe even ethics in AI about. I need to argue for that.

So, the first of this is security and the people who are in these war environments said basically, the first thing they take away from you is your sense of security. Three ways of doing this by directing, directly threatening violence, or by, threatening your sense of ownership, or your relations over property, or people, you'll love or the threat, that somebody else will inflict violence either against you or against your property.

That's, that's a subtle one. That's the idea of inventing an enemy that might harm you. In order to take away your sense of security, because you don't feel security, you feel compelled, to comply, you lose your agency. And to my mind, maybe I'm wrong here, but to my mind, the only response to this is to remove the effectiveness of violence to make it not possible for people to harm, you harm your property harm, your business, your loves etc.

And that is based in solidarity or trusting community. I think it's a hobby and argument, isn't it? We form a society in order to secure ourselves that part seems to be true. Now, informing a society. We we have the capacity of doing more harm than good, but the first instinct is a good one.

Second thing they take away is identity identity is, but maybe saying right is too strong. But identity is the capacity to define and to be who you are, but they said war reduces each and every one of us to one thing a tool of one side or another for fighting the war.

This is something that we've observed. I read along way gone. I watched blood diamond, the whole story of child soldiers, seems to me to be powerful evidence of this. And when you become a tool for fighting the war, then conformity becomes mandatory, There's a range of conformity. But minimally you need to identify yourself as being on one side rather than the other whether through face markings, or those little toughs that go up or the wings on the winged hassurs, whatever, speak the same language is or the same codes often supporting the same religions, but it's not just conformant.

Conformity, it's obedience, right? You are a tool, you will obey and, and you no longer have agency. You are no longer responsible morally, responsible for the outcome. And we saw that right in the neuromegrid trials, where people say, well, I was just following orders because they believed they no longer had agency.

The only thing they could do was all orders. So, you know, by the time you've lost your identity in this way. It's almost too late and, and perhaps you external intervention is required of some sort. And the time to prevent this is before this happens by protecting and promoting the right to define and to be who you are either as in the individual or as a culture as in the case of the distinct society of Quebec, or as in all language speakers, right?

You know, the people who speak, I don't know. French.

The third is voice and voice is more than just making noise. Although it certainly has making noise. It's part of the expression of identity but it's also in the mechanism that we use to create community, develop ideas, create future. Visions possibilities etc. It's how if you will. We actually create this cohesion that allows us to have individually agency and also collective agency Voices as they say the right to say and it's not just the expression of an opinion but the idea that what you say can and will be measured in weight.

And again, in the ethics of care, the person who has this, say, first and foremost, is the person who will be most impacted by whatever is being talked about. And also, the person who has the least agency, the least power, right? Because you're trying to promote agency and power or agency, or the the feeling of agency, the feeling of self efficacy, the feeling of ownership over the outcome of something.

And this means that people need to learn to have their voice. I mentioned this earlier, when I talked about autonomy, people need to learn how to be autonomous. Similarly, people need to learn to self-organize, to make decisions resolved, dispute, etc. So promoting this is to promote support for people who are trying to learn these things finally opportunity.

And this by this, what we mean is a path or at least a potential for a path to get you from where you are to what you aspire to achieve. This is the actual idea of free learning as opposed to controlling the idea of that you can actually choose or even better.

Not choose, but forge, or create this path for yourself. People who will have their opportunity denied to them in this way. Other took peace. And this is something I think that it's important to recognize when we have an environment of control learning, when we have an environment where we have an artificial intelligence say, or an educational institution, or even a teacher, no matter how ethical.

And how behind actually making all the decisions for the person, their opportunity has been eliminated, their ability to go where they want to go, rather than where their instructor or aorta. I want something to go and it puts them in a position of struggling against them inevitably but maybe not inevitably, but but frequently, I don't want to say, you never really do nothing, really inevitable.

So what needs to happen? Especially for people in marginalized or vulnerable groups is support for self-determination support for them being able to do whatever it is that they want to do, there are caveats on that, we'll leave about the side. But the idea here is that, instead of directing them, instead of telling them, what they ought to do, these mechanisms, support them in what they believe themselves is the right thing to do.

And so that requires an infrastructure and educational system of the appropriate sort. But also, you know, all the rest of society, for example, a commercial infrastructure that is robust and trustworthy. I didn't. And it's funny, you know, it's these are the sorts of things that enable agency and it's this agency that enables people to be quoting quote self-made men, right?

But without this opportunity without a social infrastructure that actually allows them to pursue these objectives. Their pursuit of these objectives would be impossible. I'm gonna wrap up the section on agency. By looking at Jose rest repos, four steps to peace. And this again, was presented in Columbia the same session that I attended with these survivors of the conflict again, like my thing, it's not authoritative.

It's not deep. It's intended to provide a framework or a mechanism for thinking about these ideas, you know, and to outline the sort of practices that might be the sorts of things that we want in order to foster a more ethical approach to analytics and AI. So these are the past to peace.

First of all redistribution, you know, for example scholarships for students from low income families and areas of conflict and increased capacities in communications information technologies learning.

There's a lot of discussion about the whole ethics of redistribution. There's, there's two actual major ways of thinking of redistribution one is to take away. What? People already have their house, their money, their possessions, whatever, you know, the basically an act of expropriation, some sort, the other is to take away something that was potentially there but never actually became.

There's and this is what taxation does. This is what management and control over public properties. Such as as a mineral resources or hydroelectric resources or lumber or whatever, that sort of thing. And it's interesting. The psychological effects of the two mechanisms of redistribution are different, not surprisingly. Because when the government comes and takes stuff from you that impacts you very differently from when the government simply prevents you from taking more.

And I think that might reflect some of the objection that some people feel to taking the money of the very rich and making it the money of the very poor. You know why they would represent that as a kind of theft but preventing them from ever getting the money in.

The first place is very different and it's true that some people can feel that they have an ownership over these opportunities and only ship over these revenue streams, but the sense of theft isn't as strong there, as it is an outright seizure of property for read distribution. In many events is certainly an area of discussion because, you know, we live right now in a world that is very unbalanced, in terms of income skills.

Capacities access to power, and access to resources. Etc. And there is very good argument for read distribution, not just to preserve the peace or to win the peace and conflict areas, but just in general, and society in general recognition is a second one. And recognition is raised and a lot of social political and ethical contacts.

For example, in discussions of gender equity and diversity promoting things like remembrance truth and reconciliation recognition, you know, I put here a focus on training of indigenous communities and excombatants and in even that isn't really how I want to say it because recognition isn't the result of training. It's something different.

Recognition. Well, I mean, Charles Taylor talks about two types of recognition and this is this is where some of the, the distinction and some of the difficulties come up. First of all, recognition of dignity. And this is the idea of a person as an ethical agent. A means or sorry.

And rather than means, and that carries with it, the sense that everybody should be treated the same, at least that's how it's expressed, right? The same rules should apply to everyone, but there's also recognition in the sense of recognizing differences, different culture, different heritage, different language, different belief system, and respecting that.

And that the the two are at odds with each other or at least, you know, and they seem to be a large with each other. They seem to suggest that we can't do both. We can't treat everybody the same, but treat some people as different. I personally think this is a fairly narrow view because of course, there are ways to treat people the same and to treat them differently because treating a person quote unquote isn't a single thing.

It's a whole set of things. So we can treat, everybody is being governed by the speeding loss on the highway. Even in cases of emergency, you can't go above the speed limit. But you can give some people special privileges on the highway like high occupancy vehicle lanes or bus lanes.

So here we have a case of we're recognizing everybody, you know, the dignity of everybody, the right of everybody to be safe on the highway by virtue of the fact that nobody's speeds. But also the recognition that different classes of people, different groups of people have different needs and different rights to different parts of the robot.

So in the trick here is making come work. You know, there's you know I mean, Charles Taylor wrote a whole book about the politics of recognition, which I can't even come close to summarizing much last the much wider discussion. But there is a sense but you have to recognize the other person as equal in order to have in order to well, no, I don't want to say it like that.

You have to recognize both the dignity and the distinctness of the other. If the other is to have agency, if the other is to, in your eyes, have ethical standing representation working together to promote. And this is this is again the way of promoting peace and Columbia, that was suggested working together to promote interdisciplinary.

Research to develop capacity to rebuild society sharing particular on human rights and writing textbooks to teach new generations. That kind of the control learning way of doing it. Representation to me is to answer the question who speaks for us, right? I don't always speak for myself and and I won't always be able to speak for myself because my interests.

My capacity for agency is projected into many places where I have no knowledge, no influence. No, nothing. You know, some aspects of tax policy for example, but I don't even know. Exist nonetheless, require my representation, so who speaks from me, you know, the idea in representative democracy is that I have a voice in choosing, the people who speak to me, and to the extent that they do actually speak from me, I consider that system to be more democratic kind of funny how that works.

And and you know, again thinking about these things, socially constructed, how we've taken the idea of democracy to be such that one person's speaks for a group of people who are organized together by virtue of living in the same place. It's an odd concept. I would have thought that the one speaking for the people would have been people organized by virtue of agreeing on something, right?

And then you don't have the scenario where the person representing, you actually says the opposite of something. You believe it's the sort of thing that we need to think about and consider. But representation isn't only about governments and control. Representation is also about who conveys the idea of what it is or who it is that I am out there to the rest of the world.

So, for example, we have media representations of culture, which can be quite damaging to the culture by the perpetuation of stereotypes or mistaken beliefs or, you know, even substituting their own voices, their own views, or even their own selves, for representations of that culture over and above the members of that culture themselves.

And that's how we got, you know, the scenario of white people playing Indians or Japanese people in the movies, right? The right to be represented was simply taken away from these communities, at least in these contacts and given to somebody else. Representation ultimately means people being able to determine for themselves who speaks for us or who speaks from me.

I'm running out of voice here. Reconciliation. Finally is the mechanism of promoting peaceful conflict, resolution and alternative past to success self-expression and community support. There are different processes in different societies that have been used. Apology memorials, truth telling, amnesties, trials and punishment, lustration, which if you wonder this, I did means vetting, people who are seeking political office today with respect to the things that they did while they had political office in the past reparations.

Forgiveness and participation in delivered of processes. So these things again are all practices cultural practices but can promote agency and therefore, promote a societies capacity or the members of a society's capacity to define for themselves and to be responsible for the ethics, or the morality in that society, this gives us this concept of shared agency.

And that's an important concept in the sense that it allows for things that are not necessarily individual human beings to have more outstanding. But there's, you know, there's a lot of care that has to be taken here because shared agency implies shared responsibility and if responsibility is a company with punishment, you have shared punishment or the.

In other words, the phenomenon of collective punishment. The sins of a father shall be visited upon the sons, right? Or the idea of know, punishing a whole society for the misdeeds of some people in that society to considered a war crime. But at the same time, we want to allow for the possibility of that, a culture, or a people, or a nation or a community can, you know, express some sort of shared entities, shared agency and it's not simply as rough says it means Stanford encyclopedia philosophy.

It's not simply the aggregation of individual acts, rather, it's the structure of interaction among those individuals structure that has a distinctive normative significance. For those individuals with an impact most immediately on each individual's, intention based practical reasoning. It's the fact of being associated with other people in a certain way and the resulting behavior on the part of that person, as a part of that agency, or as part of that, that organization that creates shared agency.

We could extend this maybe to ethics and analytics, maybe we can have something like a shared agency between a human, and a computer, or maybe an artificial intelligence could have a shared agency as a representative of, as a group of people on who's behalf it is acting.

But it doesn't follow that it's responsible for anything over and above the specific conditions of that shared agency. So if for example, it went off and did something on its own that was not sanctioned by the groups by the people that it represents. It doesn't follow that. It, alone is responsible, doesn't follow that.

The people are responsible and they're questions here of who is the actual agent? Where is the ownership? Where is the power? You know where is the capacity? Located. I know all this was pretty complex and often not directly related to ethics and artificial intelligence, but I think it was important to have this discussion because we cannot talk about ethics without talking about agency for the reasons I've outlined here.

And when we start talking about agency, we start getting into complexities that aren't really covered by the ethical theories involved of, you know, consequentialism they ontology and the rest because of the many different flavors of agency that exist and it's also important to because not only humans have agency.

We can talk about the ethical standing of animals. You know who among you hasn't said good dog, right?

We can talk about the ethical standing of organizations like companies or nations or cultures, but that understanding comes with the caveat that, you know, as ethical agents, they have or need to be allowed to have those properties. That constitute things that can have agency at a minimum coherence and autonomy and maybe depending on how we construct, our concept of agency rather more.

And this does allow us to assign agency to things like artificial intelligence. But the extent to, which we assign ethical responsibility to our artificial intelligence only exists to the degree that we allow artificial intelligence to take agency away from us. And to my mind, the best mechanism for defense against unethical uses of not just AI or analytical power, but corporate power.

Institutional power etc is to refuse to allow these to take away from our own individual agency. That's does not dissolve in to and, and ran individual, all against all kind of thing. We looked at the idea of agency as an evolutionary concept based on survival of the fittest, but that's a concept that doesn't work for a variety of reasons.

It's not clear that cultural phenomena all reduce to a struggle for survival of the fittest.

It's not clear that cultural values or social values can be reduced to individual values. There is, you know, it's it's not a reductionist thing. A culture can have a property that is not the property of any given individual. For example, a culture can have a history that expands centuries or as the history of an individual, wholly applies for years and so on, and the properties that a culture has things like language, things like laws things like the capacity to move mountains are not the properties that an individual has.

And individual does not possess a language, lawyer culture does, and the individual does not possess knowledge. The way a culture does and two different kinds of things, they can both have agency. But talking about the agency of an individual is distinct from talking about the agency of a culture and we need to be careful not to blend.

You know, the the biological with the social in that way, you know, we can use the metaphor, the body, politic, but the politic is literally not a body. It is something distinct. And that's why our discussion of ethics is not going to be exhausting by a discussion of culture, or for that matter, practices or regulations etc.

The discussion of ethics ultimately is going to come down to the discussion of the individual person. We are ourselves how we come to have ethics at all of what that means and how that informs us with respect to our interactions with our culture, with our social political technological frameworks with the law and with the rest of the world.

That's the subject of the next and final set of discussions in this course and to them we turn. But for now, thanks for joining me. I'm Stephen Downs. I'll see you next time.

Force:yes