▶︎ 0:00 Welcome to Lecture 6: What Philosophy Can Do. So far in this series of lectures, I've been arguing that we should give up the conception of metaphysics that thinks of it as if what it's doing is discovering deep, hidden, worldly truths, or trying to. And I've suggested that those who follow this method of metaphysics have tended to ignore the diverse sets of functions that language serves in human life. That means it's not always trying to describe some features of the world.
▶︎ 0:38 I've also argued that then what we should do is to start a step back and look at the functions that the relevant form of language, whether it's the language of describing the creatures in the world, or of doing mathematics, or of talking about possibility, or about morality. That we should first step back and assess the functions that this form of discourse serves and how it works before we jump into any metaphysical questions about what the things are, their nature, and so on. And so I've been also arguing that we should rethink of this work in metaphysics as work in conceptual engineering, including both a reverse engineering phase that involves understanding how our terms and concepts work and what functions they serve, and a constructive or forward-looking phase that involves asking what we should do with them now. And I've tried to show how this makes a difference to thinking anew about old metaphysical problems about numbers, properties, morality, and truth.
▶︎ 1:43 Now, in each of these cases so far, this has mainly been work in reverse engineering and understanding how our talk about morality or truth or properties or numbers works. I haven't done much in terms of talking about the forward-looking project of evaluative and constructive and reconstructive conceptual engineering. So I want to start there in this lecture and talk about how that work can go, and this will lead us naturally into the closing theme of this lecture series, into understanding what philosophy can do and why it's important to human life, and why it's never done.
▶︎ 2:26 First, let's talk about constructive and reconstructive conceptual engineering. If we're going to replace traditional metaphysics with conceptual engineering, how should we do it? I mentioned in a prior lecture that we should think of conceptual engineering in two phases. One, a reverse engineering phase of understanding how the language or concepts work and what they do for us, and two, a constructive phase of what we should do with them. I've talked some about a method for this kind of reverse engineering. We've seen at least two examples of this kind of stepwise approach to understanding linguistic functions and how they get layered up across different linguistic and grammatical developments.
▶︎ 3:11 The question I want to address now is how can we go forward in doing reconstructive and constructive engineering? Again, some critics will say, "Well, look, you can't do this kind of constructive engineering without first doing metaphysics. You've got to figure out what persons really are to figure out how we should use the term person," for example. But if it must really rely on metaphysics, we can't replace metaphysics with conceptual engineering. Others say, "Well, look, if you're not going to base it on a metaphysical understanding of, say, what persons or freedom or whatever really are, then it's just going to be anything goes. It's just going to be a power game. You can just reconstruct concepts to meet your own interests, and there'll be nothing more to be said."
▶︎ 3:58 I want to speak against both of those. That is, I want to show how we can do a sort of forward-looking, constructive, and reconstructive conceptual engineering that really is pragmatic. That is, it doesn't rely on metaphysical claims of the sort that I reject, but it's also not just arbitrary or anything goes.
▶︎ 4:19 So what is that method? Again, even if you're looking towards forward-looking, constructive, and reconstructive work in conceptual engineering, I do think you typically have to begin with reverse engineering, with understanding what functions discourse of this type, broadly speaking, has served in the past. Because only then can we better go ahead to evaluate whether these are functions we still should be serving or still want to serve, or that are still being served well, even in our current context, where that context is always changing socially, historically, technologically, environmentally, and so on.
▶︎ 5:05 Now, in the prior work I've done in this series of lectures, we have engaged in that reverse engineering work to understand how concepts of properties, numbers, morality, and truth work. In each of those cases, this kind of reverse engineering work has been vindicating. That is to say, it's vindicated our use of those concepts, shown why it's actually a really good idea, why it would be a shame to give up our talk of numbers or properties or truth or morality. But it doesn't have to go that way. Conceptual engineering is not inherently conservative or revisionary. You have to look at it on a case-by-case basis and see, once you've seen what these functions are, whether they're things that should be preserved going forward. Sometimes understanding what the functions have been or still are can be a prompt for revising our conceptual scheme.
▶︎ 6:10 Let me talk you through a few cases in which I think that has happened and can happen, and should happen. Here's one prominent example. Race terms. What are their functions? What functions have they served? And for whom? Well, look, grammatically we can say already that race terms like the Caucasian race or whatever, that these are nominalizations. We've seen that, and we know from the work in linguistics that nominalizations generally serve generalizing functions, enabling us to form the kinds of generalizations we need to do work in science, scientific laws, explanations, predictions, and in bureaucracy, formulating laws of state, general principles of action, general regulations.
▶︎ 7:05 Now let's go back to history and see what those might have been. Race terms, as we know them in our European languages, were introduced as a system of human classification, classifying what were thought of as different kinds of humans, and were really developed and made popular in Europe in the 17th and 18th centuries. Now, at the time, they purported to have certain kinds of theoretic goals, scientific goals, of enabling us to develop generalizations about different types of people. Anthony Appiah, the philosopher, in his Tanner lectures, talks about this and traces some of this talk to prominent texts like by Thomas Jefferson and Matthew Arnold. And he shows clearly that they both assumed that identifying different races could help us to form generalizations. Generalizations that would enable us to infer the moral, intellectual, and cultural characteristics of people from what race they were classified as being in.
▶︎ 8:10 Now, Appiah also argues that if that really was the function, we actually have really excellent empirical evidence now that that was a failure. That is, there's good empirical evidence now that we know so much more about genetics, for example, that there aren't any such empirically justified generalizations available. As Appiah puts it, each of us is the product of a huge range of genetic and environmental factors, and here's a quote, "There is nothing in the theory of evolution to guarantee that a group that shares one characteristic will share all or even most others." How do we classify people racially? Well, in this country, we mostly do so by superficial characteristics of their appearance or also assumptions about where their ancestors came from or predominantly came from. But again, as Appiah says, here's another quote, "While there are some characteristics that we're very good at recognizing, skin color, hair, skull shape, that are very unevenly geographically distributed, the groups produced by these assignments do not cluster much for other properties," and certainly not for those that Jefferson and Arnold thought we could generalize in terms of.
▶︎ 9:26 So if that's the true function of racial terms, if they were supposed to serve us in forming generalizations that would entitle us to make inferences about other characteristics, intellectual, moral, cultural, of people, that's a failure, and we should just eliminate these terms, it looks like, just like we should eliminate fish, as Carnap said, if we can no longer make good generalizations about all fish. Now, maybe there's some reservations to be brought in there. Maybe there's some reasons that we need at least some replacement terms for certain population groups to the extent that this is relevant to doing scientific work in population genetics or in medical genetics. Occasionally, there are medical treatments that seem to work differently for groups of some ancestry than other. Perhaps we'll need to reintroduce some terms in the vicinity for those properly scientific purposes if the generalizations pan out. And Quayshawn Spencer, the philosopher, is doing some interesting work along those lines. But on the whole, they certainly aren't going to serve the kinds of functions that people like Jefferson and Arnold had presented them as serving.
▶︎ 10:38 So you might ask, wait, so they were failed if they were supposed to be theoretic terms. They just didn't work. But let's look back to where these terms were introduced. I mentioned in an earlier lecture that you can often get historical evidence about what were the needs that were felt by those in power in developing and promoting the concepts and language at the time. Doesn't take much reflection to remember that in Europe in the 17th and 18th centuries, we were at the height of the slave trade, slavery, and colonization. Well, what other functions might racial terms have served, since they weren't doing so well as scientific terms, that would explain why they persisted, why they remained a central part of our way of thinking, even if they failed at underlying theoretic generalizations?
▶︎ 11:31 Well, again, I think we can get a clue from this historical situation of origin. Part of what people were looking for was a justification for the enslavement of some people or for the colonization of others. And then we can get this clue again from linguistics. What these kinds of generalized terms, nominalizations, do for us is they enable us to form both scientific and bureaucratic generalizations. Racial terms may have failed as scientific terms, but they played a really central role both in our informal practices. We have concepts of who's of what race and how they are accordingly to be treated and to be regarded, and what sorts of jobs they should have and what social standing they should have.
▶︎ 12:19 And also official bureaucratic systems like laws. Laws that laid out different treatments for people, especially in the antebellum South, but also in pre-civil rights era laws. And there were even these very fine-grained racial terms that now have thankfully fallen out of use, that played a role in our legal systems, terms like mulatto and quadroon and octoroon and so on, that have now fallen out of service but played a bureaucratic role in establishing different rights, permissions, behaviors, laws governing people who are categorized in these different ways.
▶︎ 12:54 Once these functions are made explicit, that these terms were failures as terms to serve in scientific generalizations. And what they were really used for, is for enforcing these kinds of differential bureaucratic and social treatments of people who got categorized in these different ways, in ways that couldn't be justified by supposed differences in their characteristics that we could infer scientifically from their race. Then we can look at that and say, "We've got really good reason to think about revising these terms and deciding that we should no longer try to serve the function of having terms that will enable us to treat, bureaucratically, differently, people on unjustified grounds."
▶︎ 13:36 So what should we do? Should we just eliminate racial terms? Well, I think there's some temptation to that, or at least to eliminating racial terms as they have been used and as we have thought about them, or racial concepts more broadly. And yet I think we also have to be sensitive. Maybe we'd all have been much better off if we'd never had them or never thought in these terms. But where we are now in our historical situation, where we have a kind of enduring legacy of the history of slavery and colonization, we might still need some terms along these lines in order now, hopefully not to bureaucratize in the old-fashioned ways, but rather to try to help us identify and eliminate racism and other related forms of bias. Sally Haslanger made something like this point.
▶︎ 14:31 So what should we do with these terms? Should we eliminate the old terms, introduce new ones for population genetics, or terms that will enable us to try to identify and reduce or eliminate forms of persisting racism? These are tricky questions. I think there's no doubt that some revision is in order. Exactly what it should do will again depend on what functions we want these terms to serve, whether in our evolutionary theories or instead in our bureaucracies in ways that will help us to identify and root out racism or in our social scientific theories to track racism and how it's changed over time. I'll leave that investigation to be continued by people who work directly on that topic. But at least what I want to say here is we can get from that an example of how reverse engineering and identifying the functions needn't be vindicating. Instead, it can give us really good reason to instigate changes, to consider different kinds of revision in our language or conceptual scheme.
▶︎ 15:39 And that is at least one kind of case. If you reject the function, if you now come to see the function as reprehensible, if that's not a function that we should serve going forward, you've got grounds for revision.
▶︎ 15:52 Now, that's not the only grounds for conceptual revision. Sometimes, we can also have terms that failed at the function, but we just want to get rid of or revise on those grounds. We talked about the ways in which Carnap thinks we might need to revise the term fish in biology and what we might need to do with it now.
▶︎ 16:14 Phlogiston was a term in chemistry that was supposed to be a way of identifying the substance released during combustion. Turned out not to work that way. We've gotten rid of it, except for if we're talking about the history of science.
▶︎ 16:25 But there's also terms that once served their function pretty decently but can no longer serve it well, because the environmental, social, knowledge context, or technological context has changed.
▶︎ 16:40 A great example of this that I alluded to earlier is the concept of death, and there's a really neat paper by Bernard Gert and some others on this topic. What does the concept of death do? You might think of it as functioning to mark a kind of biological change, and that's one thing it does for sure. But it also is important for social and bureaucratic reasons. It tells us when to stop using medical care. It also tells us when to start preparing the funeral, when survivor's benefits can go into effect, and so on. There's a whole lot that hinges on it legally as well as biologically.
▶︎ 17:24 Now, once upon a time, 100 or 200 years ago, death was defined as the cessation of spontaneous breath and circulation. And you'd hold a mirror up to the person. You see the old movies. And this was fine when most deaths happened at home, and death as a biological process is a gradual one. It's not instantaneous. There's a whole series of changes in the biology of the person affected. It used to be fine to leave it vague, to specify the day of death, without it having to be down to the minute.
▶︎ 18:05 Well, what's changed? Some of what's changed is changes in medical technology. Cessation of spontaneous breath and circulation used to be okay, but of course, now we have ventilators, and we can keep circulation and breath going often for a very long time after there's no longer the kind of brain activity that is typical of living people and that enables interaction as persons normally interact. Also, most deaths now happen in our culture in a hospital, and people are hooked up to very expensive life-saving devices, which can bankrupt a family if they don't have the insurance to cover it. There's reason to not carry those on after the person is dead, and how do we define that? There's also the possibility of organ transplants, and for that possibility, the sooner you harvest them after death, the better the chance of success for the receiving party.
▶︎ 19:08 So all of these changes put pressure on the medical community to try to revise and reconstruct the concept of death and the criteria for death in a way that would be more precise, so that we can not waste money on medical treatments when the person is no longer going to benefit from them or is dead, so we can harvest organs sooner with better hopes of success, et cetera. So that's a great example of how technological changes can prompt a need for conceptual change.
▶︎ 19:45 Here's another case where conceptual revision is needed. Now, not so much because of technological change, but rather because of social change and change in our knowledge about the social world and public knowledge. Think about the concept of marriage. Again, many people, certainly in the days when, in this country, debates over gay marriage were prominent. A lot of people wanted to look at it metaphysically and asked, what is a marriage? What is a marriage really? What is the nature of marriage?
▶︎ 20:17 And this line of argument was most typically used by people who wanted to say, "No, look, what a marriage is, it just is a relationship between one man and one woman." As if they were making some metaphysical or perhaps religious discovery. Again, I don't think there's really a prospect for staring deep into the world and discovering the metaphysical nature of marriage. Marriage, certainly as it was being argued about then, is a social concept, and it's one that's embedded in more than 1,000 laws in this country, endowing different rights and responsibilities to people based on whether or not they are married: tax laws, visitation laws, laws about caring for children, and so on.
▶︎ 21:04 So what are the functions of the concept of marriage in our contemporary society? Well, again, you can look back historically and see that in some historical periods, it was to give the woman over as property. I, for one, would not endorse that anymore. If those are the functions of the concept of marriage, I think we'd do best to revise it or eliminate it. But these days, though it may have some of that legacy, we can see the legal functions of the concept of marriage as giving a kind of legal protection, and legal obligations and permissions to those who are involved in what are generally life-shaping close relationships that often involve the merging of lives, the raising of children, and so on.
▶︎ 21:48 And as I say, there's over 1,000 laws in the US that refer to this status. Given that, and given the social knowledge, which has become more and more public, that a whole lot of these close relationships involve same-sex couples or could, if this is socially permitted, then you have to ask, what could possibly justify not revising the concept to extend those same rights and benefits and requirements to those couples as well? And so in this new social situation, a situation of social knowledge where these kinds of relationships are no longer hidden, we can see that we have reason to revise the concept on pragmatic grounds when we consider its function. Not because we've discovered the true metaphysical nature of marriage.
▶︎ 22:41 In both these cases, death and marriage, and even race, so all three cases, we can see that once we really pay attention to the functions of the concept, we can see grounds for revision, either because the function is objectionable or because it's no longer working so well in the current technological, social, or knowledge context. But these changes also, though they're not metaphysically motivated, they're not at all arbitrary.
▶︎ 23:13 With that much in hand, that shows us ways in which we can reconceive the work of metaphysics as conceptual engineering, including both reverse engineering and forward-looking reconstructive engineering.
▶︎ 23:26 Let me step back a little bit more broadly to talk not just about metaphysics, but about what we can do in philosophy, and how we can do it, and whether it's still worthwhile in human life. Let's go back to that quote from Stephen Hawking and Leonard Mlodinow that I mentioned in the very first lecture. Just to remind you, it goes like this. People have always asked a multitude of questions. How can we understand the world in which we find ourselves? How does the universe behave? What is the nature of reality? Where did all this come from? Traditionally, these are questions for philosophy, but philosophy is dead. Philosophy has not kept up with modern developments in science, especially physics. Scientists have become the bearers of the torch of discovery in our quest for knowledge.
▶︎ 24:25 Now, look back at that quote with a critical eye and notice what's missing. He talks about questions of how the universe behaves, what's the nature of reality, where all this came from. These are all descriptive questions, and they are all empirical questions for the sciences, if you ask me. But notice what's missing. What's missing is all what you might think of, and certainly what philosophers would call normative questions. These are the should or ought questions. Now, in my view, the very most basic philosophical question is simply a normative question.
▶︎ 25:07 What should we do? Or maybe better, what should we do now, given that we're in a constantly evolving historical, environmental, technological, and social situation? These kinds of questions, these should questions, are obviously central to a lot of the areas of philosophy, and they're not questions that science alone addresses.
▶︎ 25:35 Think, obviously, about questions in ethics, in moral philosophy. They ask questions of what we should do, how we should live together, how I should live in order to have a virtuous life, what makes a good life for humans, how we should treat each other, and much more. The questions of political philosophy are also obviously should questions. How should we arrange a government? How should we arrange a system of laws? What human rights should we recognize?
▶︎ 26:13 Not just the questions of ethics and political philosophy, but even the questions of epistemology, the theory of knowledge, can be easily thought of in terms of should questions. How should we engage in scientific investigation in order to achieve the truth? How should we engage in reasoning in order to preserve the truth? Those are questions of logic as well. What methods of inquiry or reliance on testimony or reliance on the knowledge of others should we trust in order to acquire knowledge? These are all central questions of epistemology and logic, and they're also all questions of what we should do.
▶︎ 26:54 All those kinds of areas of philosophy are really easy to think of on the model of asking these kinds of normative questions, questions about what we should do, how we should live together, how we should reason, how we should investigate. And none of those are questions you're going to get answers to from a laboratory or from empirical investigation alone. The view I've been presenting in these lectures enables us to see metaphysics as falling under that banner, too.
▶︎ 27:27 Once we think of metaphysics as engaged in conceptual engineering, we can see it, too, as addressing should questions. What concepts or language should we use? How should we use them? Should we change them? Should we revise them? Are there some we should keep no matter what? And this requires, as I've suggested, a first step of reverse engineering to understand how they do work.
▶︎ 27:51 Some of these questions of conceptual engineering also aren't just metaphysical, but also take us into issues in applied ethics and philosophy of science, in feminist philosophy, as we ask questions like what concepts should we use of harm or of privacy in applied ethics. Or in philosophy of science, what concepts of species should we use in biology, for example, or of confirmation should we use overall? This is one of the questions Carnap was interested in. Or in feminist philosophy? It's been really heavily interested in questions like this. What concepts should we use of gender or of rape or of consent or of misogyny? And there's been really interesting work trying to clarify and rethink many of those concepts in feminist philosophy.
▶︎ 28:42 So if we take this route and we think of the work of metaphysics and a lot of the work of other areas of philosophy, too, in terms of conceptual engineering, we can make it functional again, in both the senses that it will function well and transparently, and also in the sense that it won't make the mistake of assuming that all of our discourse functions in the same way. So understood, philosophy generally, and even metaphysics specifically, is centrally relevant, world-changing, and it's a perennially needed human project.
▶︎ 29:20 Let me say why. Those who oppose conceptual or linguistic ways of understanding what metaphysics has done or can do, or what philosophy can do, often do so by treating it as shallow, unimportant, just local or parochial. And they'll say things like, "I don't care how our language or concepts happen to work. I want to know about the world." But I think that's a mistake. Our concepts and language aren't unimportant, and they're not even unworldly in a sense. Because what concepts and language we have and the ways we use them come with and can provoke changes in the world and are central to the ways that we act and live.
▶︎ 30:06 So think about the importance of how we understand some of these central philosophical concepts. The concept of freedom is something we want to preserve. But how should we understand freedom? Whether you acted freely or not makes a big difference to how you get treated in a court of law for whatever you did. How exactly should we understand and think about freedom? It's going to make a huge difference.
▶︎ 30:32 How should we understand a person and who counts as a person and thereby gets all the rights that come with being a person? How should we understand the concept of a right and what the rights are, whether there are fundamental human rights and what they are? How should we understand the concept of art and what really counts as art and what doesn't, and so what belongs in museums and history books and getting all that kind of valuation? What about the concept of justice? How can we understand justice or harm if you want to know whether a certain kind of medical treatment causes harm or killing off members of endangered species causes harm?
▶︎ 31:13 What about knowledge? How should we understand knowledge and what counts as genuine knowledge and what doesn't, and when it's achieved, and what's worth teaching in the universities? How should we understand the concept of mind? This is really essential, not only if you want to understand the minds of humans, but figure out whether we think that non-human animals have minds, and if so, which ones, and if so, what difference this makes to how we should treat them. Also important if you want to know whether AI systems could have minds or that related concept, consciousness.
▶︎ 31:48 So how we think of these central philosophical ideas makes a huge difference to how we live and act in the world. They also come with really important norms for how we are to treat different individuals, as I was suggesting, for who we protect as having a mind or as being a person, for what we value as art, for what we prohibit as a violation of freedom, what we rely on as knowledge, and so on. Philosophical concepts, in fact, new ones, especially developed by thinkers like John Locke, played a huge role in the founding of the government and laws and practices of the US, including concepts of liberty, responsibility, property, and personhood, played an enormous role in how this country got developed that still has consequences today.
▶︎ 32:39 Social movements with great social change have always been attended with conceptual changes as well. Changes in our race concepts, in the marriage concept. You can see also pressure for changes and historical changes in our concept of disability, of rape, the introduction of concepts like sexual harassment have all made huge differences to how we live together, to our world, and to our legal system. Scientific revolutions, too, often require and inspire conceptual change. I've mentioned a little bit before the ways in which changing our concepts of separate space and time to a unified concept of space time was crucial in the development of physics, and the concept of gene was crucial in the development of biology.
▶︎ 33:26 Changes in our concepts also come with changes in how we recognize, respond to, and treat various kinds of conditions. So think about what we would now think of as psychiatric kinds. Back a few hundred years ago, there would've been a supernatural explanation of certain kinds of behavioral differences. People would be treated as possessed and needing an exorcism. Then there's also conceptions of these kinds of differences that treat them on a personal responsibility model. Well, you've just got to take responsibility for yourself and cheer up or act normally or whatever.
▶︎ 34:02 And then there's the medical model that's been introduced, where we now speak of mental illness, and we treat this as medical conditions that then also require payment for treatments by medical insurance, having real-world consequences, as well as consequences in how we think of people who exhibit these differences. So our concepts, they're not just idle. They come with norms for how we think, for how we investigate the world scientifically, for how we design and enforce laws, and for how we live together. Nothing could be less shallow. Our changes in our concepts come with and promote, lead to changes in the world.
▶︎ 34:49 Conceptual engineering, I want to suggest, is a way to reconceptualize the work of metaphysics and at least a lot of the work of philosophy. And it's not an optional project. We have to use some language and some set of concepts to get on, to get about in the world at all. We can do it blindly in the way that Nietzsche criticized that I mentioned in the first lecture. Just accept the concepts and the language we're given and never critically assess them, or just letting other people or history or algorithms, for that matter, push us around into using concepts and language in certain ways than others and never reflect on them.
▶︎ 35:34 To abandon the project of conceptual engineering is just to move forward blindly and uncritically without ever explicitly understanding our conceptual scheme, why we have it, why we would want to have parts of it, or if we would, and not understanding it or stepping back to evaluate it. And it's still making conceptual choices, even if you do so blindly, even if you just blindly push forward with our inherited concepts unquestioningly.
▶︎ 36:06 Conceptual engineering on this model also isn't a project that's one and done. It's a perennial recurrent project. On this model, neither metaphysics nor philosophy should be thought of as in the business of looking for eternal truths to deep worldly questions. I used to have students come in at the beginning of an intro philosophy class saying, "I don't really know why I should bother with this. Philosophers have been asking these questions for thousands of years. They don't have the answers. It's clearly a waste of time." But if you think of it on this model, what we're trying to do is not to write down the answers to eternal questions. And having not found them is then no embarrassment.
▶︎ 36:49 Instead, it's a project that requires understanding where our language and conceptual scheme have come from and how they work, and constantly rethinking what language and concepts we should use and how we should use them in ever-evolving contexts with ever-changing projects, ever-changing knowledge, and ever-changing technology. As Mary Midgley put it, quote, "Thought ought to be conceived dynamically as something that we must do and must constantly keep doing."
▶︎ 37:23 In fact, the need for rethinking our concepts and language, you might say, is more essential now than ever, because we're in such a period of rapid change, rapid social change, rapid technological change, especially, and even rapid environmental change. And this leads us to need to rethink many of our old concepts that, like the concept of death in the age of new medical technology, need to be rethought.
▶︎ 37:51 Think about the ways in which the development of AI, and even of other forms of digital technology, lead us to need to rethink the concept of privacy. The philosopher Anita Allen does some really interesting work on this. The concept of privacy worked fine when it was a matter of not peering into my window or pulling my documents out of my briefcase. But we have to rethink it when we also have all sorts of forms of digital surveillance, data, medical records, personal imagery, and so on.
▶︎ 38:22 The concept of bias, we used to think of just in terms of an individual having a certain colored lens through which they saw the world. But what about bias in algorithms, for example? Responsibility, we need to rethink the more we farm out our work to AI systems to make decisions, to self-driving cars, to move the cars around, and so on. And all this whole range of concepts like intelligence, originality, plagiarism, creativity, art, education, even friend, have to be rethought as artificial intelligence develops.
▶︎ 39:06 So philosophy, I think, once we see it as involving examining and sometimes reconstructing our concepts, is actually more needed than ever.
▶︎ 39:17 And understood in these terms, it's far from shallow or uninteresting. Far from it. Our concepts have enormous importance to how we treat each other, how we arrange our laws and our societies, how we conduct scientific investigations, how we engage in medical treatments, in short, to how we live.
▶︎ 39:42 So on this model, philosophy is not a second-rate, outmoded rival to the natural sciences who've taken over its functions. It's also not a shallow investigation of how we happen to think and talk. And it's not a failed project of failing to find answers to some deeply worldly questions that we haven't found the answers to. Instead, it's a project of trying to figure out how we should think, talk, and live. A project that's already had enormous impact to how we live, to how we govern ourselves, and what we do.
▶︎ 40:21 And it's crucially important going forward in guiding what we do. And it's always changing and is never going to be completed. But we should go into it clear-headedly, without the illusions that come from thinking of it on the model of scientific knowledge, however much we might want to borrow that respectability. Philosophy is not a rival to the natural sciences. It's doing something different, but something important.
▶︎ 40:50 We must go on using some conceptual scheme. The worst things we could do would be to just uncritically use whatever conceptual scheme we fall into without ever asking questions about it or reflecting on it, or to let some sort of skepticism about finding the true answers thought of in metaphysical terms lead us to just give up trying. Instead, we have to keep asking questions about what we should do, including what language or conceptual scheme we should use and how we should use it. Because how we think and speak of ourselves and others really matters to how we live and to what we do.