Loading summary
Transcript6 lines
- [00:01]
Progressive Insurance Announcer
This episode is brought to you by Progressive Insurance. Do you ever think about switching insurance companies to see if you could save some cash? Progressive makes it easy to see if you could save when you bundle your home and auto policies. Try it@progressive.com Progressive Casualty Insurance Company and affiliates. Potential savings will vary. Not available in all states.
- [00:28]
Jaron Lanier
I work in AI. I'm the prime scientist at Microsoft, one of the principal organizations bringing AI out into the world. And so you might think, well, if I kind of believe in people more than AI, what am I doing? And the answer to that is that I think of AI in a particular way that I find is more pragmatic. It makes me a better engineer. It makes me more able to bring benefits to people who I think should be the beneficiaries. So the way I think of it is as a new form of human collaboration.
- [01:04]
Amari Edwards
Welcome to Philosophy for Our Times, the podcast brought to you by the Institute of Art and Ideas. I'm your host, Amari Edwards. Today's guest is Jaron Lanier, computer scientist, visual artist and authority. Lanier has been named by Time magazine as one of the world's hundred most influential people. Today he'll be discussing why artists shouldn't fear AI. Few people are better placed to explore these questions than Lanier. He spent decades at the intersection of technology and human experience, helping to build the digital world we now inhabit and spending just as long warning us about its dangers. In this talk, Lanier draws on his latest artistic projects to ask what originality means in a digital age, why artists should resist the temptation to fear AI, and what role human creativity will play in shaping the future we are building together. This is Jaron Lanier on why Artists Shouldn't Fear AI.
- [02:11]
Jaron Lanier
Foreign. Is a starting point for talking about creativity in the future. Here is why. In the past of human thought and culture and philosophy, there has often been an element of faith that relates to God or what happens when you die, all sorts of things like that. I think that faith has shrunk. The meaning of God has for most people, shifted into something that's more unified with reality or perhaps more abstract rather than a particular entity. You ask for things from not everybody. I think we've in general adapted more to the natural world as our home. As science has progressed and modern, the modern world has accomplished various things as a result. And yet what I would propose is there's a new frontier of faith which we need to accept and lovingly show allegiance to. And that is specifically a faith that we exist and that each other exists. What I Mean by that is that we have to believe on faith, not based on science or on reason or on anything verifiable or falsifiable. We have to believe that people are special. Now, why is that belief suddenly important? Well, the reason why is that we are in an age where the most influential process in our life experience is increasingly temporary. Technology. And it'll probably be even more so in the future. Now, technology is different from pure science and it's different from mathematics in the following way. Science and mathematics are both definable and sensible, without concern for what people are. Now, one might disagree with that. One might say, well, quantum theory seems to run into this question of an observer and some interpretations. So maybe that requires us to say something about what a person is. Perhaps. However, in general, the process of doing science is one of increased knowledge. And it's not necessarily the case that knowledge in one domain must proceed in order for knowledge in another domain to proceed. So therefore, this question of what's really going on with humans, is consciousness something special? Is there anything special about people? Is really separable from any other question in science and might not be a scientific question at all. My view is that it is not. As I said, I think it's a faith question. Similarly, with math, we can ask, would aliens have the same math as us? And my personal answer to that is that they might have a math that's so disjoint that then we would never notice any bridge between them. But they could not have a math that is contradictory. And at any rate, whatever we think of what we are or what the aliens are would probably not influence what the math is. You know, so far as we know that is the case. But, but, but let us consider technology. I know of no way to define technology without also defining a beneficiary of the technology. So, for instance, if I were simply to make a random configuration of materials, of rocks and silicone and carbon fiber, so, like weird, random conglomeration, I could define some weird alien for which that would be a useful machine. It might be rather hard to do it, but such an alien is plausible. In order for technology to be a technology, it has to be coupled to a beneficiary. Otherwise, it's just a senseless notion. Any computer might be interpreted by some arbitrary alien as a sort of a fancy lava machine, as the thing that makes heat and patterns if they have no means to interpret the patterns. So the thing is, in order to be a coherent technologist, in order for technology to be a meaningful thing, we need to have A beneficiary which has traditionally, and in a sort of an obvious. And in an obvious way that need not have ever been identified. It's people. When we work on new medicines, they are for people, or maybe for something related to our food supply or the survival of our ecosystem or something, but ultimately for people. We might define the ecosystem or animals as beneficiaries. That is perfectly legit. But the point is there is a beneficiary. Now, a problem occurs if you define technology as its own beneficiary. So. So for instance, I'll give an example of this. I shared a meal the other day with a young man who had come to a conclusion that he would not have biological children because of a fear that if he did so it would evoke an emotional commitment to biological humans. Whereas the more important commitment in the future was to make it safe for AI entities who are more likely to survive than biological humans. To me, this is senseless, because if we're saying that an AI is an entity and a technology is to benefit that entity, we are not defining a separate beneficiary for the technology. And we end up in a kind of a self referential, very claustrophobic, tiny space that just doesn't make any sense. This happens sometimes in the sciences. Many have argued, for instance, that modern physics has sometimes gotten lost in theories that are kind of for themselves instead of related to empiricism or the natural world. Some have said this about string theory, for instance. Similar criticisms have been made of other sorts of intellectual pursuits, that they are too lost in their own little ivory tower place and, and no longer connected to the practical world. And what, you know, it's arguable to what degree that's a problem in basic science. But with technology it is certainly a profound problem, because if the technology is its own beneficiary, honestly, it can be anything, any fantasy can apply. I can declare that some AI entity wants this or that. Who's to disprove me? Essentially, it's a backdoor way of somebody expressing a faith in what they think the technology is or should be. It's essentially a new form of religion, but they're pretending that they're being technical. So if we need faith to talk about what technology is for now, which is certainly the case for AI in the, in the hard AI sense, that we're creating some new entity or some new consciousness, then let us consider our choice of faith with a little bit of an eye towards rationality. And there's this very peculiar thing that happens where the Only rational choice is to choose where to compartmentalize your irrationality. So this is kind of an interesting thing. I think whenever you try to make a perfectly rational way of encountering reality, you cannot. This is formally true in math, where there's always some little. I call math a. An immaculate house with a messy closet. That's how math is. And I think, therefore science is like that. But the thing is, you can choose which closet gets the mess right. And so with technology, the flaw that prevents us from having a perfectly closed, crystalline approach to reality is this question of what we put faith in. And I would argue the only sensible place to put it is this rather irrational and indefensible choice of having faith in ourselves, in each other, in ourselves, because we experience consciousness. So there's that. There's this weird other epistemological channel and in each other. Well, this is where faith comes in. You know, do we believe in each other or not? Now, this is something I'll get incredible pushback on. You know, a lot of my colleagues think this is a horrible, antiquated way of thinking and even unethical, because once again, supporting the future AIs is more important than supporting the people, since they're more likely to survive, et cetera, or they might just be better or something. That's an attitude I run into. But the degree of irrationality required to make AI the beneficiary is vastly greater than that required to make humans the beneficiary, because the AI is entirely a creature of our own fantasy. I can make up an AI stuff has any arbitrary wants or whatever, because it's all our own fantasy being projected. So this might sound like it's a rant that is a little bit off topic. If my topic is to be the future of creativity, and I believe that's the topic. So let me explain where I see the connection. I work in AI. I'm the prime scientist at Microsoft, one of the principal organizations bringing AI out into the world. And so you might think, well, if I kind of believe in people more than AI, what am I doing? And the answer to that is that I think of AI in a particular way that I find is more pragmatic. It makes me a better engineer. It makes me more able to bring benefits to people who I think should be the beneficiaries. So the way I think of it is as a new form of human collaboration. Consider the Wikipedia. The Wikipedia is made of contributions from people. The people are generally operating under pseudonyms. This leads to a feeling that the Wikipedia is Sort of an emergent collective voice. I personally don't like that about the Wikipedia, because I think there are many topics for which there really shouldn't be. One overall perspective, you know, it doesn't quite work. There's some for which it works, but the way it's presented might not be quite right, for instance. But anyway, I won't go into the Wikipedia. But the point is, you can think of large language model AI has been quite similar to the Wikipedia. It incorporates contributions, source documents, training data from many, many people. The data is amalgamated and in some cases extended in various ways by algorithms. But it's ultimately similar to the Wikipedia. What you get out is different in that instead of a single preset collaborative article, it can create compound articles combining subjects that the Wikipedia could never do. So, if you want to know, I don't know if you want a biography of Ludwig von Beethoven, but in a pirate voice, it can combine those two qualities statistically. And that is a new thing in the world. And sometimes it is immensely useful. It turns out to be really useful for generating code in computer programming. It turns out to be useful for various things. Many people are fascinated by it. Okay, so that's large language models. Now, one of the problems with the term AI is that it can mean anything, because as I was pointing out, it's an arbitrary act of faith. So somebody might say, well, what about AI that isn't just a large language model? Well, there are many gradations of AI and there will be many more in the future. And I don't wish to address all of them right now, except to say that they all ultimately come from people. If you choose to believe in people as being special, which I propose we should do. Now, if you share with me a sense that AI is a new form of collaboration, a new form of bundling the creative outputs of many people into a sort of a new source of compound creative outputs. There's another interesting way to think of it, which is a way of making the efforts that people have made in the past more available in the present. So here's what I mean by that. If I wish to do something like I wish to say, how can I use this particular peculiar contraption? May be one of the weird music machines sitting around me here in order to accomplish a particular sound. There's no guarantee, but there's a chance that a large model will be able to do that by amalgamating and interpolating between all the many, many things that have been said on forums and social media sites by musicians and There's a very large volume of that stuff, so it actually might come up with something, and it happens. It's not a guarantee, but it's pretty cool. So essentially all the efforts of those people are more available to me than they traditionally have been. Now, every form of human communication has done this, has brought the past more into the present. Writing obviously has done it, so have movies and so on. But this really does more of a dragnet to capture previous creativity and make it available now. There are good things and bad things about that. It has often been pointed out with regard to human creativity that in a sense, high quality past efforts by people can be discouraging or overwhelming or invalidating to present efforts by younger people, by newer people. So, for instance, if somebody tries to write a novel now, are they going to write at the level? And here one could choose different people that you care about. Maybe some of you would think of somebody like Nabokov or Joyce as exemplars of the best possible novelists, and others might not. But whatever it is, there have been a lot of good novels written in the past. And so the question is, is a person in the present kind of having to compete against these incredible peaks of achievement from the past? Is that a problem? And there's an argument that maybe there should be a little more forgetting and that maybe a degree of haziness about the past has been a bit of a benefit to creativity because it puts it under less stress in the present. That's not an easy topic, honestly, because I think creativity is always fresh, because reality is always fresh. Human experience is always fresh. And the world in which we're being creative is not the same as the world in which past masterpieces were creative. And so therefore there's a new kind of relevance, new kind of freshness that comes up. And that is why there are new styles. And I don't think this is a trivial thing. I don't think you can just say, well, it's all just churn. And just because right now we're in an era of a certain kind of popular music that in a hundred years will be in a different one and then a different one, and they're all just kind of the same and they don't really matter. I don't think that's true. I think they do. And I think there is a cumulative process, but it's hard to articulate exactly what it is. Anyway, the reason I'm going through this whole question about legacy is that this is going to be an increasing problem now for all Future creativity in the sense that creative people are increasingly feeling that they're sort of in competition with what can come out of the large models. I've interacted a lot with filmmakers using tools like Sora video synthesis tool or just text, generative text tools or image tools. And there's this sort of feeling that this ability to recombine past efforts, often with a sort of a surreal sensibility where you say, you know, can I have surfing cats on Mars or something, and then this thing comes out and it seems to be a sort of an instant creativity. And so then what do you do with that? Where do you go with that? It's particularly difficult at the moment economically, because some of the systems that have created economic sustenance within either a world of government support or institutional support, or in a market capitalist world, are under threat as more and more resources go into the central hub of whatever the biggest computers are that run the biggest algorithms. A trend that I think will continue. I guess what I have to say about that is it's an urgent matter for creative people to view this as a challenge that must be met and can be met, although only with great difficulty. I think we will be entering into difficult times and yet I think people will come out of them well. I think our sense of what is novel, of what is is creative, will necessarily evolve and surreal combinatorial forms of creativity will probably become deemphasized, even though, particularly in 20th century modernism, they were profoundly emphasized almost in anticipation of AI, or perhaps as part of the motivation that caused AI to come about as it ended up coming about. The new creativity that will emerge is by definition impossible to predict, but I think it will be dissimilar from 20th century modernism. I think in retrospect, a large Number of 20th century artists who were in part differentiated because of a sense of novelty that had a combinatorial quality, might have less of a legacy than we would have anticipated, because that sort of thing will come about so readily. I suspect that the new creativity will focus a little more on intimate authenticity, which is something that's of course been known in the arts. That's not a new thing. But I think that sort of creativity will become more important. I think a sense of radical honesty will be important. I think an element of personal style that is intent, that is subtle enough that it's hard to capture, hard to copy, will become a thing. I think lifestyles that reject an AI first way of life, people who don't let their data out, who have a kind of a privacy about them, might start to become the vanguard of a new aesthetic. This is an interesting, an interesting moment for that particular dynamic. So I'm a rare person who's both a public person, but also has never had any social media accounts. Somehow I've done okay. I'm not even exactly sure how that's possible. Neither is anyone else. But I suspect there'll be more and more people doing this sort of thing of saying I'm not going to be part of the data. I'm going to try to really chart out a less promotional, less platform oriented way of connecting with people, Even though that's hard and even if it's hard to even know exactly how it's working at all. But I don't know. We shall see. I think there'll be a lot of folks entering into a crisis of meaning where they wonder if they. Part of the reason for that is the tech industry, which will increasingly be the richest and most influential power center in the world, transforming everything and possibly being more, more influential than traditional states or all sorts of things, and might be there already. Certainly technology culture tends to ridicule, resent and belittle creativity in humans because of this moral imperative to make AI into the new creature. Why is that a moral imperative for so many? It's an interesting question. I think the right way to think about it is that in each era of human history, our means of survival have side effects in which human personality type tends to come to the fore and become the most influential. If we are in a situation of violent groups confronting each other, then we'll tend to have warriors become the most influential model personalities, Genghis Khan or something like that. We're in an area in an era where computer hubs are where the power is. And so we have a certain sort of nerd and slightly spectrumy personality becoming the most dominant one. And within that personality framework, the idea of the computers becoming alive has enormous appeal. I suspect it's a feature that makes more sense to many of my colleagues than a chaotic one of a lot of creative human beings. Nonetheless, as I say, the only way to be rational is, is to adopt an extreme irrational allegiance to humans that you can't really justify. That is a matter of faith. I don't really think there's any other path. That's the only place the messy closet can exist that works. So creativity, hopefully, hopefully in the future creativity will become the central definition of being human. It hasn't really in the past because in the past, you know, we were often a little lower down on maslow's pyramid, if you like. We were a little bit more about survival or wealth or whatever. But if we enter into an era of, and I don't know if we will, if we enter into an era where we're somehow not just killing ourselves constantly using technology, if we can end up in an era of more relaxation and reduced threat, I'm not sure we will. That evidence right now is meek, but it still is a possibility. Then there'll be a chance for creativity to become a dominant force. The question is, what does that look like? We're used to creativity being, let's be honest, a little bit of an elite thing. Even if you leave aside the formal arts. Not everybody can be a great cook at home, not everybody can be a wonderful gardener. But, you know, a significant portion of us have been creative in our identities. Somehow that will have to increase in order for mankind to be happy and healthy. I think the way it might increase is that rather than technologies growing in a uniform way, they might expand in a blossom of exponential variety, each of which creates new niches for creativity. Some of this will sound wild, but I don't think I have any choice but to be wild, to think about the future. Because if we survive and if technology continues, it'll get wild. Let us suppose in the future we have a technology that allows our bodies to reformulate themselves and change form biologically through some kind of a mechanism of rewriting stem cells or something. And you might suddenly grow tentacles or wings or some other sort of limb that we're not familiar with that never has occurred before, maybe will evolve to be able to free float in the vacuum of space happily, or who knows? But the thing is, whatever that is, there'll be a gigantic new form of creativity for people who invent the space of options in that. But there might also be a million, and then eventually 10 million things like that. And so I think the variety of forms for creativity could expand fast enough in the future for people to be mostly creative, even though there's a lot of us, and to not fall into a sense of being unimportant and redundant. So, you know, when a lot of my friends like to think about extremely radical technical futures, but they think of what that means is some AI superintelligence that replaces people. I think what it means is more and more niches for human creativity that we can't even imagine. And I think mine is more radical and more beautiful than theirs. But it's an ongoing, lifelong debate I have with people where you are. Thank you so much. For your attention. Foreign.
- [30:02]
Amari Edwards
You've been listening to Jaron Lanier and why Artists Shouldn't Fear AI Creativity in the Digital Age. If today's conversation sparked something in you, please take a moment to rate and review Philosophy for our Times on Apple Podcasts, Spotify, or wherever you're listening. Your feedback is really appreciated and it genuinely helps more people discover this podcast. For more talks, debates and articles on philosophy, science, politics and culture, please visit the Institute of Art and Ideas website. IAI tv from all of us at Philosophy for Our Times, thank you for listening. We'll see you next week.
- [30:49]
Progressive Insurance Announcer
Insurance isn't one size fits all. That's why customers have enjoyed Progressive's Name your price tool for years now. With the name your price tool, you tell them what you want to pay and they'll show you options that fit your budget. So whether you're picking out your first policy or just looking for something that works better for you and your family, they make it easy to see your options. Visit progressive.com, find a rate that works for you with the name your price tool. Progressive Casualty Insurance Company and affiliates Price and coverage match limited by state law.