full interview_marcus du sautoy_2.mp3
Marcus du Sautoy [00:00:00] My name is Marcus du Sautoy and I'm a Professor of Mathematics.
Speaker 2 [00:00:06] So why do you love math?
Marcus du Sautoy [00:00:09] I fell in love with mathematics really because I began to understand that this is kind of the language in which the universe is written, that to understand the kind of chaos and mess around us, that there are actually patterns hiding underneath these. And if you can understand these patterns, it gives you a real insight into where we've come from, more excitingly, where we might be going next. So I think it's the. Aspect of pattern searching that you know if I had to describe what is a mathematician I'd say there's nothing to do with arithmetic and numbers and long division and things like that that's arithmetic mathematics is really I'd take something like the science of patterns and and so for me I think that was why I was drawn to it because I love the idea of finding the underlying pattern in something which looked at first sight very chaotic I think one of the other reasons that I fell in love with mathematics as opposed to say some of the sciences is that I was always interested in the arts as well, music, theater, literature, poetry and for me what I found was that mathematics was this wonderful bridge between the two, that somehow mathematics is the language to understand the universe and how it works. Yet also mathematics has a very creative side to it. You can make worlds that have nothing to do with the physical universe, yet they are as interesting for a mathematician as any other one. And for me, that sort of creative side of making imaginary worlds sort of resonated perhaps more with the idea of a novelist creating a new kind of world. And also... The ideas of patterns that I found interesting, sometimes they wouldn't have anything to do with the kind of physical world around us, but the patterns would be genuinely interesting in their own right. And I think very often I'd find that those patterns were ones that a composer might have been interested in or a painter. So I think because I was always interested in both science and art, and I really resented our education system asking us to choose one path or the other. And I said, but I actually would quite like to do both. Don't I? I think I found that mathematics was this great bridge between the two and I thought ah this is my way of doing both being a scientist and being an artist at the same time.
Speaker 2 [00:02:38] Sorry, I have to ask another question.
Marcus du Sautoy [00:02:42] You're happy to ask anything again.
Speaker 2 [00:02:44] Right, mathematics is beautiful, talk about that please.
Marcus du Sautoy [00:02:51] People often talk about the importance of beauty in doing mathematics and I think that's kind of interesting because for a mathematician I think trying to find beauty in the ideas or the equations is often a fantastic guide to finding what actually ultimately is going on, the truth of these equations. So I think you know, you go back to Keats who said you know beauty is truth, truth is beauty, and I think... You know, maybe how do we define something as beautiful? I think because it causes some sort of resonance in the body of a recognition of something important. It's kind of evolutionary trigger for us to take notice. So I think that that's why we are calling the mathematics beautiful, because very often it's a signal that we are on the right track. That's not to say that some mathematics is quite messy and ugly. And is necessary for understanding perhaps how the weather works, for example. But frankly, that's not really the mathematics I'm interested in. So I steered away from anything that didn't have that kind of simplicity about it sometimes. It's not always about being simple, and I think beauty is slightly dangerous, because sometimes it's about an extraordinary journey that you go on, which has an esthetic to it. But it may not just be about beauty, it may be about drama and tension and resolution. So I think beauty often is a bit static in its kind of description of an art form, and I think very often mathematics has a kind of dynamism to it, and it's about the twists and turns of ideas which can have a beauty to it. But for me, I think. I am drawn to beautiful structures because I think they're going to have some significance for understanding perhaps the universe or something even deeper maybe.
Speaker 2 [00:04:52] So, mathematics, solving a problem, how does it feel? You wrote in your recent play, I'm like six, that pleasure lasts for days. Tell us about that.
Marcus du Sautoy [00:05:07] People often, I think, react against mathematics because they find it difficult. But actually the difficulty of the subject is part of its charm. For me, what I really enjoy is wrestling with a problem for hours, days, months, years. I mean, I'm working on a problem at the moment which I've been working on for 15 years and I still can't crack it. But the moment when I do suddenly see how to make this thing work... That moment is the drug that we mathematicians kind of thrive on. It's such a surge of adrenaline. I mean, people talk about the ah-ha moment. It's that moment when you suddenly get a glimpse of what feels like the secrets of the universe, because these are universal truths. This has nothing to do with culture or this particular moment in time. You are seeing things which are universal truth across the universe and that is an extraordinary. I've had it a few times in my life when I've realized, oh wow, I've just created the first time that anyone's seen this particular structure, a symmetrical object for example, which nobody has seen before and I've somehow summoned it up out of my equations and I still remember those moments when I had those breakthroughs and you have to remember them because they're rare. And you just want to repeat them as many times as you can in your life and most of the time is spent just getting nowhere but you go I remember that time and I want to have it again so that's why you carry on even when things are really difficult.
Speaker 2 [00:06:47] So you have a wide range of interests beyond your core work of mathematics, music, playwriting, food. Is that unusual for mathematicians?
Marcus du Sautoy [00:07:01] So I'm, you know, fascinated in things beyond just mathematics, you know, I enjoy doing music, it's been very much part of my life ever since I fell in love with maths, I was learning the trumpet for example at school, I love the theater and it's interesting because mathematics in some sense requires a very narrow focus and I think it's probably not surprising that there's There's quite a lot of aspergic, autistic traits that, er... We mathematicians display and we need that ability just to kind of completely focus, dedicate yourself to doing, working on a problem. But I think many of us actually need something else, some other stimulation in order to kind of let our subconscious just carry on working on our problem and I think for me that's often what I do, you know, if I'm working at my desk, I get stuck, I've got just not getting you anywhere, I will then just pick up my cello. Play a Bach suite or something and then I come back and it's as if my mind has been slightly reset. I've got a slightly different perspective on the things, Bach especially, of course, you know, the mathematician's favorite composer. It's all about pattern searching and interweaving things and it has done something to my brain which gives me a fresher outlook when I come to the problem. So I think you'll find, for example, maths and music, a lot of connection there. So I kind of mathematics department could probably muster up an orchestra out of its members.
Speaker 2 [00:08:36] Does mathematics inform your work in the humanities and vice versa?
Marcus du Sautoy [00:08:45] I've been really struck by the time I've spent with creative artists, visual artists, musicians, even theater practitioners, that the kind of things that they're interested in, the structures that they like embedding in their work, just resonate so strongly for me. I recognize they're exactly the same things that I'm obsessed with. The difference is that they've got a different language to kind of explore these structures. So a composer might be... Using sound to explore the symmetries of an object and I'll use the language of algebra to do it. But I think that's why I've been so excited over the last kind of few decades when I, you know, first of all just spent all my time doing maths and then I went out and started talking to more creative artists and time and again I said well that's exactly a structure I'm fascinated by but you're looking at it from a completely different So I really enjoy just seeing that we're all interested in the same structures, just from different perspectives. And the exciting thing about that is their perspective might give me a new insight into the things that I'm looking at. And this is why I think this dialog is not just one way. I think very often the art-science dialog is often about artists plundering the scientists' cabinet of wonders to create a new piece. What does the scientist get out of it? For me, I do get a lot out of these collaborations because they can sometimes reset my view of the things that I'm looking at. I'm not saying they're going to solve the Riemann hypothesis or the problems or conjectures I'm working on, but they might give me a new question which I hadn't even considered asking about my structures which have come from their world. So that's why I think it is genuinely can be a very rich dialog to help scientists see things in new ways.
Speaker 2 [00:10:40] Do you have a specific example of a moment when something from the world of art or humanity did, in that moment, inform your mind?
Marcus du Sautoy [00:10:51] There are a couple of moments that I've had in my work where talking to an artist has raised a new question for me or even a new structure that I hadn't actually thought about before. One in particular, I'm obsessed with a Greek composer called Xenakis who loved embedding mathematics in his composition and there's a piece I've been working on at the moment called Nomus Alpha for solo cello. Where Xenarchist uses the symmetries of a cube to control the variations at work inside the piece. So there are 24 variations, and each variation corresponds to a symmetry of the cube, and then the cello kind of reads off the different sound qualities the cellos can make according to where they are now on the corners of the tube. So my own research is about symmetry, so I was intrigued that Xenachist had tapped into today. Constraints of the symmetry of the cube. But what I hadn't understood before was that he was actually choosing a pathway through these symmetries which was based on a completely different bit of mathematics called the Fibonacci numbers. These are numbers which you get the next one by adding the two previous numbers together. So artists love these because they give rise to something called the golden ratio which Leonardo loved for example. But, what Xenakis had done was Thank you this kind of algorithm for numbers but was applying it to symmetry and he was moving the cube around according to this kind-of Fibonacci light rule. Now I have never seen that idea used in mathematics ever before. But once I started to think about it was, well, you can start to explore that for different symmetrical objects. And Sinakis incredibly had found kind of the longest pathway of these symmetries where it would come back to the beginning again. So it started a research project for me, which is, yeah, what about these Fibonacci-like journeys through the symmetries of objects for other things other than the cube? So that for me is, he discovered a new structure, which through musically, which we've never seen mathematically. So that's really exciting. How many other kind of structures are hiding there that the artists will be able to see that we have? I've got another one if you want.
Speaker 3 [00:13:15] Please.
Marcus du Sautoy [00:13:17] I, my other passion as well as music is theater and I created a piece of theater with an actress from a wonderful theater company, Theater to Complicite. The play was eventually called I is a Strange Loop, but I loved the idea, actually one of the inspirations was Borges, another artist who loves threading science through his ideas, exploring through narrative the idea of infinities, contradictions. And also the nature of shape, the shape of space. So there's one story inside there called the Library of Babel that he wrote, which is about a library made up of hexagons. And the hexagones, like a beehive, cover the floor of the library, but then there are layers of hexagon going up and down. And so I use this as an idea, well, why don't we use the stage a bit like a little library? And the rule that Borges discovers is that when you go through this library, you keep on coming back to the beginning because it's a kind of circular. Both on the floor and in the kind of ceiling coming through the floor. So this interconnected nature of the library I sort of realized in this piece of theater. But the question then became, in this library, there are two doors in and out of the rooms. So it actually means it creates a labyrinth inside here, because you enter one room through one door and you exit through another door. The challenge came, and this isn't explored in the Borghese, but it got me thinking, well, yes, what sort of path can you make through this hexagon, labyrinth of, through this beehive of hexagons, such you can get to every room? Is it possible to draw a path through all of the hexagones such you could get to everything, every room, by this labyrintheine path? And once I did that, you know, I've got a notebook full of these little like beehives and But then again, there was a question I hadn't ever considered about this very simple kind of structure, the beehive, hexagons put together. But how can you draw paths through there such you meet every room? And it actually highlighted that it was one position which was almost like a center to the whole thing. And Borges talks in that story about a special room. So had he actually solved this himself, or? Was this just his imagination, there should be some special room with an index of the whole library. But actually the mathematics when I explored it, suddenly indicated there might be a special center to this whole structure. But that idea of the pathways through these kind of lattices of structures, again, wasn't something that I'd ever seen before, but it was Borges who got me kind of, and this theater project, got me thinking about a new kind of mathematics.
Speaker 2 [00:16:04] So you talk online about what you say when people say there is a God. God is a mathematician.
Marcus du Sautoy [00:16:13] Aha. I find it rather extraordinary that our universe, when you start to analyze it, it's not a mess. It initially looks incredibly chaotic and messy and unclassifiable, but as you dig down, you suddenly just find this beautiful abstract mathematics at the heart of everything. You know, physicists have understood atoms pull apart to electrons and protons, but they put a part to these things called quarks. And named after novelist James Joyce, named these fundamental things in physics essentially. But yet, when you try to understand these, you realize they are just a piece of mathematics. So why is mathematics kind of... I think many people say, oh well, God, the creator of all of this, must be a mathematician. Well, I'm going to reverse that, because I don't believe in a God. But what I do believe is that we have to try and give some answer to where all of these came from. Why is there something rather than nothing? Why is it so mathematical? If you want something to be created you need to create something a reason which needs to be outside of time. You need something which doesn't have its own moment of creation. Now for me, mathematics is the perfect sort of example of that. There are moments when we create and discover new things, moments in time, but the mathematics itself I believe has always been there, isn't a moment of creation. Mathematics is about structures and patterns that are possible and aren't created. So, I mean, I'm a true Platonist at heart. I'm revealing my philosophical kind of nature, but for me, mathematics is always there. So what we are seeing, I think, is a physicalized piece of mathematics around us that at some moment. The kind of void became, you know, out of the void quantum physics or the mathematics implies that well zero, nothing, well that's also equal to one plus minus one so out of nothing you can start to create something and actually this is the beginning of a new play that I've written, the Axiom of Choice, where we start with nothing, the empty set, just nada, sunyata and uh in modern set theory we now say that is the beginning of the whole of mathematics because you take the empty set and then you you've got something there you've got nothing so suddenly the set of nothing is something and that's the beginning that's one and using that idea you can start to build up the whole of all our numbers and from numbers you get mathematics from mathematics you get the whole science and i believe art as well so really for me mathematics is right down there at the bottom of everything. And for me, that's why artists are secret mathematicians. They are actually sort of exploring these structures themselves. I mean, I think that's, all of us are trying to interpret the natural world around us, which I think is basically mathematics. And so we're all finding different ways to realize that. Maybe it's musical, maybe it's a composition, maybe it an abstract art form, or maybe it is a story. That they I think that's why so often they don't think they're mathematicians But when we begin to talk they say well If you're saying this structure is mathematics, then perhaps I am more mathematical than I thought I was
Speaker 4 [00:19:43] The Press Okay, can I just get...
Marcus du Sautoy [00:19:46] It really short because your answers are so long and what the fuck are we going to do with this in the edit? Yes, sorry about that.
Speaker 4 [00:19:53] Off in a great direction. That's great, but just the whole thing about the universe and being, the math being there all along. Is this a really, I mean very profound? Yeah. It needs to be said a little bit more simply. Yeah. You know what we're asking. Yeah, okay.
Marcus du Sautoy [00:20:12] For me, I think the universe is basically a physicalized piece of mathematics. And that's why we're seeing mathematics all over the place. And I think that's certainly in science, you're seeing lots of mathematics help you explain things. But I think also at the heart of why art often is exploring mathematical ideas because they're trying to use their art form, music, visual art, storytelling, to understand the world around us. And if the world is mathematical at its heart, it's inevitable that they will be creating their own versions of these mathematical structures.
Speaker 2 [00:20:47] So math is everywhere.
Marcus du Sautoy [00:20:49] I really believe that mathematics not only is everywhere, but is everything. There's an interesting example actually of an artist who kind of, you know, one of the interesting things is do artists realize that they're creating mathematical things or are they sort of discovering them almost by chance. So for example, Jackson Pollock, flicking paint around, couldn't anyone do that? But no, he's actually doing something very special. And he's creating a sort of chaotic structure with the way that he painted and what's happening on the canvas is a kind of fractal structure beginning to emerge and he didn't realize he was creating this geometry called a fractal but when we mathematicians looked at it we were able to analyze that has fractal structures. If you zoom in on it, it never simplifies. I mean at some point you hit a sort pixel of paint or something but the beauty of painting is is you get lost in them because you lose a sense of how close you are because of this scale, this nature that's at work in how he created these things. But what is he doing is he's producing an abstract version of the natural world around us because if you go to his studio, it's surrounded by forest, trees. When I went there, all the leaves were down and you just saw all these branches going up and they are just fractal structures. So he was abstracting. The natural world around him into this abstract canvas, but was ultimately creating something which we as mathematicians recognize as a fractal.
Speaker 2 [00:22:30] So, I'm going to switch to music. Kurt, you touched on it a bit. How do math and music come together in your head?
Marcus du Sautoy [00:22:42] For me, mathematics and music came together because I fell in love with them both at the same time, when I was about 12 or 13. My maths teacher at my school excited me about new structures that I'd never seen before, and then my music teacher got me learning the trumpet. So they've always been a sort of parallel paths for me. But I think, you know, many people have talked about the connections between maths and music. Leibniz used to say music is the sensation of counting without realizing that you're counting. But I think it runs much deeper than that. Sure, Pythagoras understood that the notes that we find harmonic are mathematical in nature because they are whole number ratios that define a perfect, fierce, an octave. So the ingredients of music, rhythm, harmony, are very mathematical at heart. But for me, it goes much deeper than that and it's about what you do with it. Those are the... The kind of letters and words, but what are the stories you're going to write with that? And for me, where the true mathematics meets music is in the overall structure of pieces that composers will make, and that for me is much more exciting than the kind of the ingredients of note or rhythm.
Speaker 2 [00:23:58] We filmed a wonderful emotional sequence with a music historian in New York about Bach's prelude number one in C major. And he talked about how that music brings the listener along on the journey. Can you tell us about the mathematics of that piece?
Marcus du Sautoy [00:24:19] I'm not sure about the mathematics of that piece, but maybe I can choose another piece of Bach if that would be all right.
Speaker 3 [00:24:28] But you know the taste, right?
Marcus du Sautoy [00:24:30] Yeah, I mean it's not as interesting as other pieces.
Speaker 4 [00:24:40] Can you say that? From a mathematician's standpoint, it's not as interesting as other problems. Say it to her. Yeah.
Marcus du Sautoy [00:24:47] I think from a mathematical point of view, that prelude is not as interesting as other pieces that Bach wrote. I mean, if I was going to choose a piece of Bach that really celebrated mathematics in his work, I think you'd have to go for the Goldberg Variations, which is a piece you know people love. It's very kind of an emotional journey that you go on, but and that's the interesting thing you see. Many people Say if you're comparing math and music, oh, I don't want you to do that because you're taking the emotion out of the music Well, that's completely wrong. What I'm trying to say is that mathematics has huge amount of emotion in it and the journeys we go on mathematically are very similar to the one you'll go on if you listening to the Goldberg variations and In the Goldbeg variations Bach is playing the just most extraordinary games of symmetry at work. I mean it's It's already has a circular structure. You begin and end with this aria that all the variations are based on. So the fact that Bach wants you to end with the same piece you began with, there's a circle there. And even halfway through, he calls the 16th variation an overture, which is usually the beginning of the piece. So already the circle is saying, well, circles don't have beginnings or ends. But there's also another circular structure in the the canons that he wrote every third variation is a cannon and the there's a The second voice gradually climbs up each new variation until it reaches the octave and you have this kind of resolution again of a circular structure So and the way that Bach does variations so much symmetry at work The voice might go up and the second voice will cascade down. There's reflection or symmetry and the rhythm structures he make sure that each variation covers any combination possible of quavers triplets semi-quavers two three four beats in a bar i mean it's structurally just so beautiful and satisfying and i think when people listen to it they don't know that but what they're they're experiencing it and the satisfaction of the piece you know it's an hour and 15 minutes with all its repeats. It's a long journey in a concert, actually. But I think you're experiencing just the journey around these kind of symmetrical shapes that Bach's embedded in there and the feeling of arrival at the beginning again, for example, is very satisfying.
Speaker 2 [00:27:13] Is there an emotional need that we all have for symmetry?
Marcus du Sautoy [00:27:18] I think that symmetry you find so often, just in the natural world and also in the artistic world, because symmetry, which isn't just about the simple left-right reflection of symmetry of a face. Symmetry is about some structural relationship going on in an object and it's gotta be there for a reason and so I think that we've become very sensitive to symmetry because we know that that's a signal of something of significance, something we should take notice of. So I think we've become very sensitive on a kind of evolutionary trajectory to symmetry because it's helped us to survive in this world. I mean think of us in the jungle, the chaos of the jungle and then suddenly see something with symmetry. Well that's probably an animal, it might be a tiger that's about to eat you or maybe it's something that you can eat. So those humans who recognize symmetry are the ones that survived in this world.
Speaker 2 [00:28:13] What are the symmetries in AI?
Marcus du Sautoy [00:28:17] Gosh, so artificial intelligence at its heart is basically code. What is code? It's algorithms implementing if-then instructions. And what are algorithms? Well, that's basically just mathematics. So when everyone is getting so excited in this kind of particular time about artificial intelligence, actually they're excited about the power of mathematics to make decisions and so understanding the mathematics is going to give you an ability to create the most extraordinary artificial intelligence. So you know, which bit of mathematics is hiding in there? Well there are lots of different sorts, so for example, you know there is symmetry at heart in, for example, even the Google algorithm. I mean, it isn't artificial intelligence, but it's pretty amazing at finding the websites you want. But what Larry Page and Sergei Brin understood is that a piece of mathematics that they learned about how to understand movement and symmetry and something called eigenvalues, very technical bit of math, but opened up the ability to navigate the internet. So, you know. If you know the maths, you're going to be able to understand just the extraordinary power of this new artificial intelligence.
Speaker 2 [00:29:39] So artificial intelligence would seem to be about following a set of rules, algorithm, programmer, but isn't true creativity about breaking rules?
Marcus du Sautoy [00:29:52] So I think in the past, people just thought artificial intelligence really, how could it be genuinely created? Because it's just implementing rules that humans have written down and it might be able to do it faster, deeper than a human can, but it's not really producing anything new. It's just implemented the ideas of the human. But that's really changed in the last few years because of a new sort of way of coding cool machine learning, deep learning. And this is where code is able to sort of mutate and change, rewrite itself. There's a sort of meta-code that is saying, okay, we failed that time round. If we re-change these parameters, we would have got that right that time. So it's mutating and changing. And this gives the possibility that if you give it data, it will learn from that data and become something very different from what the original programmer wrote down. That gives the possibility that the eventual thing that appears will be doing things that the human didn't expect it to do. And that for me gives the possibly of creativity. Because creativity for me is about surprise. It's about, oh, that I wasn't expecting. It's an emotional resonance. Weirdly you don't think a piece of AI is emotional, but it can learn what will trigger our emotional reaction. And the fact that this changing, mutating algorithm... Is no longer doing what the human told it to do means that it could be created.
Speaker 4 [00:31:27] Are we there yet?
Marcus du Sautoy [00:31:29] And in fact, I think that we are already seeing examples of true machine creativity. Moments when it's taken data, learned, understood something new inside that data, and has fed back to us things that have surprised us, excited us, things with value that have taken us as humans on a new creative journey. For me, it's the most extraordinary collaborator because it can see things at a level that we can't. If we look at huge amounts of data, millions and millions of data we don't have the sensual equipment to navigate that. This AI is a bit like Galileo being given a telescope and suddenly he's able to see things in the solar system beyond that we've never seen before. That suddenly unleashed new science or a microscope seeing what's actually going on at kind of molecular depth. For me, the new AI that's emerging is like a telescope into the digital world we're producing, helping us to see things which are hiding there, which we can then kind of exploit in our own creativity.
Speaker 5 [00:32:39] But the telescope wasn't creating something for itself. It was not doing something improvising itself, where something, what we're doing here might be a little bit different.
Marcus du Sautoy [00:32:49] I think at the moment, you see, artificial intelligence is still dependent on the data that we give it. So what it's doing in some ways is understanding things that we've missed in the data that has inspired our own creativity. You've got to remember, how does an artist work? It doesn't come from nothing. It comes from being exposed to all of the art of the past and sometimes you'll see artists emerging through a movement. You'll see, yeah, this is a continuation of... The esthetic that has been established and then you'll get those amazing moments of break when you know Picasso appears and starts doing things that have never been seen before but Picasso didn't come from nowhere he spent many years replicating the styles of the past in order to be able to break those and I think what's exciting is that you know there's nothing mysterious about this it's about breaking rules so why can't you implement that as an part of the code. Of the algorithm doing the creativity. Understand what's going on, then break something and see if something new and exciting emerges from that. So understanding AI creativity I think is helping us to understand human creativity in a new way.
Speaker 2 [00:34:01] Can you speak to that point, specifically, AI and, excuse me, can AI creativity exist without the human element?
Marcus du Sautoy [00:34:13] I think that it's, you know, tabula rasa learning, where you don't give the machine anything to start with. You know, it's possible, but you know what's the point of that? We don't want it to... And I think, you see, the problem with not giving it the data that we've already generated as artistic humans is that if you let a computer off on its own without being sort of exposed to... Our world, you're going to get something which just starts to resonate for us at all. I think what you want is an AI that's going to show something just beyond the horizon, not something which is so alien that we don't even know how to engage with it. So that's why I think that this new AI which is taking our data, pushing it a little bit further, finding new things in it, finding things outside of it. It's really important because it connects to the world that we have already created. Sorry, let's just let my daughter...
Speaker 2 [00:35:16] Do you want to stick it to your teeth while they're doing it?
Marcus du Sautoy [00:35:18] Well, I actually finished my tea, but, um, am I sounding a bit, uh, okay, okay.
Speaker 3 [00:35:21] Okay, okay. Yeah, yeah.
Marcus du Sautoy [00:35:25] Yeah, yeah, you can join if you want, she's, I was explaining she went to a film college for her sixth form for 17 and 18 and she's doing film studies next year at university. Yeah, if you've got any openings. Good luck with that. Yeah. I didn't want to say. All right, um.
Speaker 4 [00:35:55] Yeah, I'm trying to remember what I was saying too.
Speaker 2 [00:35:57] So we seem to be in a moment where generative art is more about technology, people are going to have these sort of visual experiences, rather than so much of thinking about the meaning of the work. Do you think this is the case, and do you think that will change?
Marcus du Sautoy [00:36:22] I think, actually, this is more than just, you know, interesting new ways to generate visual worlds. In fact, I think what's happening is that the art we're seeing is really an exploration of a really important moment in history when we've got this new technology appearing. And the art that is being produced is perhaps our best way to really understand what is this artificial intelligence that we're creating. For me, I think that it may not be so much the visual esthetic that's important, but the insight it's giving us into the code that's beginning to emerge. And, you know, Master McLuhan always said that art is our early warning, distant early warning system, telling us what is about to hit us. So I think the art that's being produced by AI... Will help to give us an insight into how the AI is thinking. Because the code that it's producing is now so complex that we can't go through it line by line, understanding why it's making its decisions. And so I think that's what's more exciting. We're getting a commentary on a very important moment in history where we've got this new technology appearing.
Speaker 4 [00:37:39] Could you just say the rational flow of mind again? Yeah.
Marcus du Sautoy [00:37:42] Yeah, yeah, no, I get it. Martian McLuhan always said that art is our distant early warning system, telling us old folk what's the new things that will be appearing on the horizon. So for me, that's why I think the art generated by AI is actually helping us to understand perhaps what will be emerging in the next few years with the impact of AI on society.
Speaker 2 [00:38:09] Do you think there's anybody right now, a generative artist, who we will look at 40 or 50 years hence, where you looked at Picasso or Jackson Pollock?
Marcus du Sautoy [00:38:23] I think that it's hard to identify a particular artist that we'll say is, you know, like the Picasso of this AI movement. But I do think that there is a style emerging. That's what's very interesting. When I do talks about AI-generated art, and more and more people getting more familiar with this particular style that is emerging from. The AI and I think that will be recognized and it comes out of a particular type of algorithm called a generative adversarial Network or a GAN and this is really two algorithms working Almost in competition against each other one is generating new ideas the other one Critiques it and says no that's too close to what we've seen already or no that gone way too far What wouldn't they're not going to recognize that as art and the two working together is producing a style of art that I think is genuinely of its moment is what we will recognize as AI art. And it has an interesting quality to it which I think it's interesting going forward because it has a complexity about it that AI really enjoys kind of making things more and more deep, complex, intricate, and I think we're starting to see AI beginning to create art that we humans are finding a little more difficult to engage with because we don't have the sensual equipment to really appreciate that complexity. And this is happening in music as well, where the sort of music that's appearing, we almost can't listen to because the complexities of the interconnecting rhythms just aren't close to the beat of the heart that we're used to. So I'm beginning to wonder whether we'll start to see a drift away of AI art, where it's not being really produced for humans, but actually this is AI art. Created for other AI rather than for humans and
Speaker 2 [00:40:15] That sounds very human in and of itself.
Marcus du Sautoy [00:40:19] Yes, I suppose it does, but I think, you know, we have limits of our senses for how to engage with the world around us. And those limits are exciting and constraints that we enjoy exploring, but it also means that there are things that we will never be able to do and think. I mean, my worry is, some of my mathematics, ultimately the proofs are so complex that My human brain, its physical nature will not be able to ever... Achieve that proof and maybe it will be in combination with an AI that is able to do things faster or at depth and it'll be the combination of us that um will make the progress and for me that's the exciting thing this is not about competition it's all this whole story is often couched in the terms of you know when is uh AI going to take over when it's going to be better than us the singularity the moment when all of this uh you know But for me, it's not about that. It's about, wow, we've got this amazing new collaborator that's able to do things differently to us, but in combination, we can both go much further than human or AI individually.
Speaker 2 [00:41:30] So we don't have to worry about human creativity being diminished as a result of it.
Marcus du Sautoy [00:41:35] We really don't have to worry about AI in creativity, wiping out human creativity. That's just, it's a Hollywood kind of story rather than the reality. And the reality of it is, is that I'm seeing artists so excited to see new ways of using their ideas that they, in a funny way. Artists often say, oh, I feel that I've just got stuck in a particular mechanistic way of behaving I'm repeating things I've done before They're behaving like machines and it's the AI that is pushing them out of behaving like a machine and suddenly reigniting their human creativity
Speaker 2 [00:42:14] Can you, simply for our simple audience, describe the loveless, the love with chest and conclusion, and whether it ever will be possible to pass it?
Marcus du Sautoy [00:42:28] So one of the first people to consider code-making machines do interesting things is Ada Lovelace, this Victorian woman who her mother took her along to see some of the scientific experiments of the day. And she saw this thing, the scientific analytic engine by Charles Babbage. It was just created to do simple multiplication. But she said, no, I think this can do more interesting things. If you give it instructions, it I might even be able to compose. Music of a scientific nature So first person really to think of machines being creative, but she had a word of warning She said yeah, but you really can't call this the creativity of the machine Because this is the creativity if the person who wrote the instructions to make the machine do something I think for many years that was absolutely true. So we now have this thing called the the lovelace test You know can a machine produce a piece of work? You know, the original programmer doesn't actually understand how it did it. So we should really call this genuinely the creativity of the machine rather than the human who programmed it. And I think up to a few years ago there's no way that could be passed because code was written in a very top-down manner. A human wrote the code, the machine just implemented it. Maybe it's speed and depth that a human couldn't, but it's really the creativity of a human. But because the code is now changing, mutating, becoming something new, like a child growing up being exposed to the DNA of the parents, it's code, but it's also got new experiences which make Picasso not, Picasso's work is not the creativity of the parent, it's the creativity Picasso. So the Lovelace challenge is, well is Lovelaces wrong that actually this really should be called the creativity of the machine, because it's doing something the human who programmed it doesn't really understand. So have we passed the Lovelace test? I think we have. I think are seeing moments of things which are being produced that humans are saying, I don't know where that came from. I didn't program it to do that. And in fact, I would have probably deleted that bit of code as a kind of silly thing to do, but it's found a way to use that as a moment of new kind of discovery.
Speaker 2 [00:44:42] So this is sort of, this is basic. How does that happen? Code generating code, the human somehow is out of it. How will our audience understand that?
Marcus du Sautoy [00:44:58] So the idea of machine learning is that when, for example, if it's playing a game and often games are where AI kind of first cuts its teeth, when it makes a move, for example it has an algorithm to play noughts and crosses, tic-tac-toe, and it makes a moves and it loses the game. And then it says, okay, I'm going to actually change the weight of, you know, perhaps it was randomly choosing paces. That seems to be a weak move, so I'm gonna, next time, the loss triggers the program to say, okay change the move you made, which seemed to be the one that caused the loss. Tone that one down next time. Let's lower the probability that you make that move. So each time it loses a game... It rejigs the kind of decisions it's going to make and increases the ones that seem to be more successful, decreases the moves that are least successful, and in that way the thing can start to change from just being sort of a blank sheet of, okay, all moves seem to be equivalent, and starts to weight moves as more successful. The wonderful thing about this is, and I think it's a great message for education These algorithms only improve... When they get things wrong. Now we are so obsessed in our education system that kids have to get things right all the time. They don't learn when they're getting things right. They learn things when they are getting things wrong and understand what it is about the nature of failing that time in order to be able to improve the next time. I think you learn so much more. Getting things wrong and that's what the AI is taking advantage of. Every time it gets things wrong it re-jigs the parameters inside the program to try and be successful the next time it meets that scenario.
Speaker 2 [00:46:47] I'm starting to feel there's something very anthropomorphic about AI.
Marcus du Sautoy [00:46:54] Well, of course, you know, AI's beginnings with Turing asking the question about whether machines could be intelligent was ultimately about trying to understand our own intelligence. So we are using these machines as ways to experiment with trying to understand our our own thought process. And that's why I think it's exciting and we have this feeling maybe we're creating something related to us. Is this an evolution of our species. But ultimately I think that's the wrong way to approach this. We know how to produce human intelligence. We don't do it in a lab, we do it the bedroom and we have children. Much more interesting is to produce a new sort of intelligence. So actually I often try and translate AI not as artificial intelligence, artificially creating our own intelligence, but maybe augmented intelligence or additional intelligence, alternative intelligence. Much more exciting to create something that is unique to the machine world and is different from us because then the combination is going to be much more powerful.
Speaker 2 [00:48:01] I'm going to go back to art for a second because you were talking about your robot artist friend who had created a sexy robot to do the art, would you speak to that?
Marcus du Sautoy [00:48:19] Artificial intelligence is kind of intriguing because it's sort of in a black box. It's about algorithms and implementation But often you then want that AI to make something perhaps make a new painting And so then you'll see the interaction of AI with robotics the limitations of a Physical arm perhaps that is going to implement the algorithm and that for me is quite an interesting interface because You've got challenges now about the abstract world meeting the physical world Sorry. So, I think there are some very interesting projects where kind of AI artists are actually putting these inside robots and making a painting and the limitations of robotics is being revealed. I mean, Conrad Shawcross, a friend of mine, lives around the corner from here. He loves this interface of art and science and he likes pushing algorithms such that actually the robot can't cope with what the abstract idea is trying to tell it and it kind of breaks. And he loves that moment when it fails and it breaks down, it does something weird. But I think we need to be very careful with this, because some of the projects I've seen are about trying to, I think it's almost like trying to trick the viewer because they've made a robot that looks human, got a beautiful face. And there's something very unsettling about this because I feel that this is sort of Robot looks human, so the art it's producing should have some resonance for you. The face has nothing to do with the actual art that's being produced. Yeah, the robotic arm and the AI, but putting a face on it, frankly, that's just ridiculous. Oriented to this guy but there's a guy that we sort of keep on...
Speaker 2 [00:50:13] Do we need that shorter?
Marcus du Sautoy [00:50:14] No. Oh, right. All right.
Speaker 2 [00:50:16] How would you define creativity?
Marcus du Sautoy [00:50:20] Defining creativity is a tricky one. It's almost like trying to define consciousness. It's a very slippery term. But I think I quite like a pretty user-friendly definition which is you're trying to make something which is new, which is surprising. Engage your emotions makes you go, oh, that's making me see the world in a new way. But shock value just for its own sake isn't creative. So it has to have some ultimate value. And surprise and value are, you know, hard things, they're not, you know, they are very subjective and will vary from period to period. And of course, creativity varies from period to period, so that's one definition. But I think actually, I mean, I wrote this book all about creativity and artificial intelligence. It was quite a useful definition to start with. But for me, ultimately, I liked Carl Rogers' psychologist's kind of exploration of creativity. That's Creativity is the tool that we humans have produced in order to understand our own consciousness. I really need to explore what's going on inside my internal world, more I want to share that with you. And also I want you to see, are you seeing the world like I do? I mean that moment when consciousness suddenly emerged in the human species and I've been such an extraordinary shock, I'm wanting to know, are, are feeling fear like I am or pain? Um, and of course that's one of the great unsolved problems of science, is that how can I ever know what it feels like to be you? I think perhaps the art and creativity that we have is our best tool for trying to solve the hard problem of consciousness. So for me, creativity is this wonderful tool that humans probably produced the same time we became conscious as a human species. I think creativity probably emerged at the same in order to explore that inner world.
Speaker 2 [00:52:17] So, does your creative process differ from one of your interests to another, from math to theater to...
Marcus du Sautoy [00:52:26] Yeah, I mean mathematics is a hugely creative subject I think you know it's a lot about leaps into the unknown, making new things, producing things that will surprise my colleagues you know, it is about engaging their emotions. Wow I didn't expect that to be true. So the choices I make and that's what's important I think people underestimate you know surely mathematics is just Oh truths and logic and where's the creativity in it? No, the creativity is in choosing the things that I will eventually talk about. I'm not just interested in proving all the true equations of mathematics. I want to pick those that will move my fellow mathematicians. So for me, I think that's the connection with the creativity of the artist. So they're looking for something that is just gonna make people look at things in a new way. It's not just good enough to make a new painting. That's not creative. It's got to... I have some action in the world. And so for me that's also a key part of my own creativity that it's not just about the equation being true. Most equations are boring. It's the equations that make somebody go, oh, I didn't realize that was related to that. That's amazing. That has value because now I can do something with that. So I think there's a lot of connection with the creativity that an artist is looking for and the creativity that I'm eager to create in my own search.
Speaker 2 [00:53:50] And so, on that, what connects your passions, your two sides, your mathematics, AI side and your music and arts and beautiful work side?
Marcus du Sautoy [00:54:05] I mean, I do all of these very different things and, you know, music, theater, exploration of visual art, mathematics, mathematics applied to sciences and things, and I think what underpins all of them is my obsession with pattern, the thing that I fell in love with mathematics right from the beginning. I just enjoy the challenge of looking at a structure, trying to understand how it And then seeing the secret to the way the pattern works such that I can see it continue on and I think at its heart music is you know mathematics is the science of patterns and for me music is the art of patterns. And I think probably that's why for me, music of all the art forms is the one that really resonates for me.
Speaker 2 [00:54:52] So you have another creative side in the kitchen. Tell us about your amazing kiwi cake. Oh my.
Marcus du Sautoy [00:54:59] Oh my gosh yes yes oh you see my yeah so I love cooking but weirdly I don't think it's a place that I'm most the time creative because what I enjoy about going to the kitchen is that I enjoy following somebody else's creativity and I'm one of those cooks that is obsessed by following recipes and I have to have exactly the right amount of this and if I haven't got that in the storeroom cupboard I'll go down to the shop and buy it so Most of the time I use. My cooking as a kind of therapy away from the mental world, that I just use my hands and stuff and I follow instructions. But there was this one time when I saw this recipe, somebody made this wonderful fruit tart and I suddenly realized... It's quite nice what you've done with the kind of patterns of the fruit on there, but I can see something much more exciting There was a symmetrical structure of sort of these four-sided figures that if you put them together Would suddenly create this almost Escher like effect of cubes being hiding inside this Tart, so so I piece together all of these different colored fruits. So there's dragon fruit watermelon and kiwi and I took ages cutting these little bits of fruit out, tiled them up and put them there, and the effect was amazing. It was just like, suddenly these 3D cubes were inside this tart. And, you know, I tweet a lot about my science and my art, my mathematics and things. The tweet I put up about that cake that I made is the one that has been most viewed, most retweeted, and everyone's going, wow! I never realized you could do that with fruit.
Speaker 2 [00:56:45] What do you think that is? Was it a foodie thing, or was it a pattern recognition thing? Yeah. Yeah.
Marcus du Sautoy [00:56:53] For me, this is a wonderful fusion of some of my passions. I love cooking, but I hadn't really thought of that as a mathematical domain, but then I think the visuals of the food are so important, so it was so exciting just to use a bit of my mathematics of symmetry to create something out of foods that suddenly had surprising connection of maths and cake.
Speaker 2 [00:57:18] So is there something that you haven't done yet creatively that you would like to do?
Marcus du Sautoy [00:57:26] I think that there are a couple of things I, you know, we only get one go at this life, but, you, know, I wish we could put ourselves in quantum superposition and do many different things. And I suppose that's almost what I'm trying to do is, even though I'm a mathematician, I can write a play and call it sort of mathematics because I'm exploring mathematical ideas. I think... If I had to choose another path, I would love to have become a composer. And a lot of the work I do with composers is kind of vicariously following that path through their work. So I have actually written a string quartet with a composer, Emily Howard. Which is called Four Musical Proofs and a Conjecture. She really did the composition and I gave the proofs from which she then bounced off and created these pieces. Five little miniatures for string quartet. But it's kind of exciting because I do have my name on a score now published by a music publisher. But I would love to kind of explore that direction more.
Speaker 4 [00:58:29] Anybody want to talk about that? Yeah, I got a couple questions, and they're going to be somewhat random. But you were just talking about the symmetry and the cake. Yeah. One of our themes that's kind of running through, especially going to a lot of our early thinking about the show, we even thought of making an hour just about the idea of harmony. Okay, I was wondering if you could talk about, and our working theory was that most humans seek harmony in their environment, you know, or in their food, in their music. I don't know if you agree or what you may think is culturally conditioned, but the fact that you're interested in symmetry, it seems to be part of all of that. So maybe you could just talk about why you love symmetry, it's obviously incredibly important to you. What's the feeling you get when you look at something symmetrical, you can, you know, things that obstruct you, just talk, talk,
Marcus du Sautoy [00:59:25] Yeah, well, there are kind of two bits in there actually, but yeah, yeah. People are often drawn to things that they'll call harmonious. Why do we sort of search for harmony in our lives and of course harmony first of all one thinks musically or of notes that feel nice and right together and the interesting thing is that's related to mathematics because actually what people are I think after are interesting sort of symmetrical or simple arithmetic relationships. So the notes we find harmonic, the ones that Pythagoras discovered, are the ones who have this very simple whole number ratio associated to them. So I think that we are drawn to these, the things that have a reason behind them that we're able to understand. And we feel comfortable with being at home with the kind of reasons for why this thing is working. If something is too complex, we get lost in it, we get scared and kind of fearful and so and again I think we're always looking for some rationale to understand these hugely complex organisms. So I think that's what we're often looking for, those structures that we feel we can understand because they've got, They might be. Outwardly quite complex, but if we can understand that at their heart is something kind of natural and simple Like a one to two relationship or a two to three relationship. Those are the notes that we find harmonic if you do a like 17 to 19 relationship in two notes You'll react to that because it's just too complex for the brain to to navigate So so I think that's sort of what? Why things resonate is that we're the brain even though it may not consciously be doing it is that is understanding that there is some Kind of simple reason at work behind what still outwardly might look complex
Speaker 4 [01:01:40] I'm just going to check to see if your mic is...
Speaker 6 [01:01:46] No, I'll come here Oh, there I am.
Speaker 4 [01:02:05] I'm more than a good one, too. Love it.
Marcus du Sautoy [01:02:10] I'm always complimented on my claps by film crews, one of my specialities.
Speaker 4 [01:02:19] There seems to be, I don't know, it's a physical pleasure, a psychological pleasure. I don't know if you can agree or disagree, when one encounters symmetry in the world. Yeah, absolutely.
Marcus du Sautoy [01:02:35] I think if you, when you go to the Alhambra, there's just an immediate sort of buzz you have about going around the walls there, because there's this wonderful symmetrical games going on there. And it's interesting, why are humans so obsessed with symmetry, symmetry, making symmetry, being exposed to symmetry. I think it is because we as a human species have become incredibly sensitive to symmetry because that's generally an important message. That this thing isn't symmetrical by random. If there's a reason why it's symmetrical, it's got a message inside it. So even, you know, if you come to humans, if you give people faces and you've made some artificially symmetrical, most people will say that symmetrical faces are the ones they find more beautiful. Why is that? I think it's because it's communicating a message. It's hard to make symmetry in the physical world. Most of the time little disruptions occur, our faces are generally not symmetrical. You need very good DNA, very good upbringing to be able to create a symmetrical face. So what's being communicated is, you know, I'm a human being with good genetic heritage. Good environmental upbringing, I'm going to make a very good mate to have children with. So, you know, very evolutionary biology reasons, but for me, I think that's why we've now taken this on to kind of more interesting levels, that symmetry for us is always about structure and meaning hidden behind what we're seeing.
Speaker 4 [01:04:15] I have some more questions that a lot of people wanted to try to do
Speaker 5 [01:04:18] We can circle back a few, also random.
Speaker 4 [01:04:22] Yeah, that's right.
Speaker 5 [01:04:25] You talked about creativity in the spirit of wanting to sort of create something new that kind of surprises people, but we heard creativity described as a sort of tool to solve problems, and that you need a problem to solve. And I was thinking about you spending 15 years trying to solve a problem, and I'm sure that's happened a number of times. Can you speak to creativity sort of in that light, and how you approach a many, many years problem, yeah.
Marcus du Sautoy [01:04:57] Stravinsky always used to say, I can only be creative under huge constraints. And sometimes a blank page is really not very helpful and you need things to constrain you. And one of the things I find stimulates my own creativity is if I've got a very specific problem to work on, which I'm trying to solve and perhaps I've seen something I want to be able to do or some connection that I want to try and prove true. And I find that's very. Helpful that constraint if I need to solve this problem, so it isn't good just to make anything I have to make something which is going to make this thing work So for me often The most creative moment is coming up with that problem because then that problem creates the constraints which I'm going to work in so often in mathematics Really the person that should be celebrated is the person who's come up with the challenge in the first place rather than the... Person who solves it. And they are often the ones that get their names on the discovery. Fermat's Last Theorem isn't called Andrew Wilde's theorem because he was the one who finished it off. It's the person who first saw that something new was going on, a challenge out there for us mathematicians to work 350 years on to try and solve. The Riemann hypothesis, currently the greatest unsolved problem about prime numbers. Whoever cracks it, it will still be called the Riemann hypothesis, who first understood there was something really interesting going on with these prime numbers that we don't know quite why it's like that, and that is the constraint that we're on now trying to prove that. But it will require hugely creative moves in order to be able to crack these problems. What often happens is that you have to go off in a completely new direction, find connection to something completely unrelated to where you first started. Some of these great theorems that have taken centuries to prove, the journey we've gone on is almost perhaps more valuable than the actual discovery that we've finally come up with with a QED. And I think that's important. You know, mathematics is not just about knowing that an equation doesn't have any solutions. It's about why it doesn't have solutions. That journey of the story of this is the reason why this thing doesn't have any solution.
Speaker 2 [01:07:20] What's the real world impact when something gets lost?
Marcus du Sautoy [01:07:24] Who cares about the real-world impact? I really have no interest in the mathematics being applied to the real world. This is far deeper than just the temporary world we have around us. We are getting access to some of the fundamental truths of the universe, things that will be true on the other side of the Universe. When we meet aliens, how will we talk to them? We'll use the language of mathematics as our tool for communication. Yeah, I'm being a bit flippant, of course, you know, we get our funding because the things that I do help create new codes, help crack codes, create new digital platforms to be able to make films on, for example, but for me, that's not the motivation. There's something much deeper, which is about understanding how this universe works.
Speaker 5 [01:08:11] Yeah, at the same time, the way I'm perceiving it is a lot of the work that you're doing is that that's a kind of popularizing or demystifying or somehow making it more clear for a larger public, as in ready to play, as a copy of music, or whatever that happens to be, so you understand the underlying concept.
Marcus du Sautoy [01:08:34] When you've had an idea, I think that it really only starts to live when you bring it alive in the mind of the other person. If I prove something, it's useless, worthless, if it just stays in my own mind, and that's part of our shared consciousness. I want to share these stories with you. So I think a lot of the work that I do is, well, if I can make the ideas of mathematics come alive in as many minds as possible, then the ideas truly begin to bubble and perhaps go on to create new ideas. So I think a lot of my work is about not just talking to my own fellow mathematicians, but trying to get the public out there excited about prime numbers and why they're so beautiful and the things we still don't understand about them. The more minds where these stories get kind of brought alive, the more the ideas live themselves.
Speaker 5 [01:09:32] Yeah, I was curious if you are familiar with the Turk, yeah, yeah the story, the chest, the mechanical code, so when they just discovered that it was controlled by a person, it was wildly disappointing, and just in terms of AI today, I'm wondering, you know, it's always going to be someone, some man behind the AI, and what... What is the level we need to get to? The Turing test, notwithstanding, how do we, do we just want to be fooled? I mean, there's always going to be a wizard behind the curtain. Do you know what I'm saying? What do people really want at the end of the day? What's the level that we have to get?
Marcus du Sautoy [01:10:19] All right. I think there's all this wonderful technology being produced, but we have to always remember that it begins with humans. We are writing code or putting bits of kit together to do these extraordinary things, you know, fly to the moon, see new things, but ultimately what's I think really exciting is that these are all actually still human creations. Um, but, what's I think there's something different beginning to happen where actually the sort of artificial intelligence we're producing now might start to really begin to make its own journey where we will not be able to say like the mechanical Turk that yeah there's a machine playing chess but really it's somebody inside there, a human. I think we're beginning to see emerge a mechanical Turk which doesn't have. Human inside it. That is genuinely something which is growing, mutating and becoming something separate from the humans who started the creation. It doesn't really answer the question, but that's it.
Speaker 5 [01:11:31] This is a very long road we can walk down, which we don't have. It's more of a philosophical thought, but I'll move on.
Marcus du Sautoy [01:11:41] I think that one of our problems is that we consider things in too linear a way, so we're always looking at, you know, when is AI more intelligent than us, can do things better than us. And I think thats the wrong way to look at it, that really we need a much more multi-dimensional view of our relationship with machines and artificial intelligence. There are some directions in which the AI will be better than us. There are other directions which we will always be superior at. We are a physicalized, emotional body with consciousness. It's a long time until I think these things will be conscious. And it's about the multiplicity of all of these intelligences which is exciting. Not about this kind of linear idea of the singularity when one thing is better than us. That's always going to be the wrong way to look at it.
Speaker 2 [01:12:29] So this AI is a paintbrush of another Europe.
Marcus du Sautoy [01:12:35] Think you could consider artificial intelligence and its involvement in the art world a bit like the moment the camera appeared. It allowed us to do a new sort of art but I think it's more than that because the camera was still the tool of the human and I think we're seeing something a bit more exciting than that. It is a tool for humans to explore data in a new way but I genuinely think that learning process is making something which is sort of different from just the kit we put together. The kit has been put together and then it's kind of rearranged itself to do something a little bit different to the way we first put the camera together. That for me is really exciting.
Speaker 5 [01:13:21] I just wanted to be remiss if I didn't try to put a finer point on education and the way we teach our society.
Marcus du Sautoy [01:13:28] Yeah, I think that's a yeah
Speaker 5 [01:13:30] So this can be country by country, it changes, but generally speaking, how do you feel about how it's taught and what may or may not want to be improved?
Marcus du Sautoy [01:13:40] I think one of the tragedies of our education system is this silo mentality that we, after about 11 years old, that kids just find themselves going to the history class, the music class, the science class, and the maths class and they don't realize that they're connected. I think before 11, certainly here in the UK, it's lovely, kids are at school learning all of these things together and then suddenly we put things in silos and for me, my is that we will just break down the... The walls between these classrooms, and realize that there's so much interconnection. You know, mathematics has a history, the moment that zero first emerged in India, for example. That's a shock to most people, they thought zero was always there, but it had a moment in history where it emerged, the connection of mathematics to music, that a musician can help explore their music by looking at the mathematics behind it. So, I really. Thinks that, you know, the exciting thing is to find the connections between these subjects and our education system doesn't help find this fluidity between all of these disciplines.
Speaker 2 [01:14:51] Can I ask just one other basic question about algorithms changing and evolving? Are they, um, they're programmed to do specific things and they're also... Program to recognize differences. How do I I mean, I'm really But how do we make somebody understand the general audience what's going on AI that it can do replications and can make changes without being specifically
Marcus du Sautoy [01:15:40] Mm-hmm. I think this idea of code kind of rewriting itself and mutating and changing it is a little bit hard for people to grasp, I mean what is it doing when it plays a game and it gets something wrong, how does it rewrite itself? I think you sort of almost have to think of it as kind of two layers of code, that after it's tried to do something that it's tasked with and it gets it wrong, the kind of meta piece of code looks at what the story was of how it played the game and analyzes perhaps what, if I changed that what would the impact have been, the if-then is very much part of how an algorithm works. If I did this, then what would happen? And so if I now change this, and perhaps instead of when I encounter this next time I do something different, what effect would that have? And if it makes me more successful, then I will strengthen that connection. It's exactly how the brain works. The brain works by reinforcing synapses that are more successful in recognizing things. And as a child grows up, it throws away synaptic connections between neurons because they're not useful. So the brain, really what we're doing is modeling the way the brain evolves and strengthens certain neuronal connections because it helps in analyzing a particular situation and forgets others. I mean, this kind of. Or when you clip things in a garden. Pruning, yeah. So the idea is the brain very often is pruning itself and throwing away things which aren't helpful. And that's really what the code is just playing a similar game of saying, okay, that was not successful. I'm gonna cut that bit of the code out. This is actually helping me to. Thank you. Often identify the difference between a cat and a dog in my visual recognition software so I will strengthen that bit of the code and perhaps I need to ask a new question about the data it will introduce that as an extra layer and that's that's the kind of learning process that's going on which is very similar to the way the brain changes a mutate.
Speaker 4 [01:18:01] Yep, I've got a question. We've talked about this before. So in making the film, we're constantly coming across artists who have employed something scientific in their work. And we talked about, I just want to hear your thoughts about that.
Marcus du Sautoy [01:18:21] I'm of course really excited by this kind of interface between art and science but sometimes I'm a little nervous about it because it does feel sometimes like you introduce an artist into the lab, you sort of tick a box of we're doing art and science, the artist sort of plunders the ideas of the day or creates something which actually isn't communicating the science at all. But then puts the word quantum on it and that's somehow meant to make people think, oh this is a very deep thing about quantum physics and so I think one has to be a little bit questioning sometimes of these art-science projects. What is the purpose of them? Is it to communicate the science? That's a valid thing. Is it just to stimulate new artistic ideas and who cares whether the science is right or not? I think you need to be very careful if you're just using the science as a way of Brassing up the art as something, you know important and significant. So be very careful when you're using these words like quantum for example
Speaker 4 [01:19:23] There is something about, I was just thinking actually, it's a little bit like the way artists, and well all sorts of folks, obviously even if you're vile, looked at different, like sort of the Eastern religion. It's just something to kind of put on their art and add to it. But it's little bit sort of a credential. It's like a little like, and sometimes it seems that way about science. It's, like, well, I've been working with biologists, So, you know, my piece is more valuable for whatever it is, whether it's trying to explain the way I'll do it or not. Yeah. A lot of artists would say, well, I'm not trying to do it. No, exactly. I'm just sort of using it, and you said pondering, which is what artists do, right?
Marcus du Sautoy [01:20:07] I think one of the exciting things about these collaborations is that, of course, as scientists and mathematicians, we've uncovered interesting new structures, patterns, ideas that the rest of the general public haven't been exposed to. So I think there really is a point about bringing an artist into such a community and telling them about these, because it might take the artist in a new direction. They suddenly sing. Patterns and structures that they never realized were there and very stimulating for their own ideas and I certainly found this with the composers that I've worked with. There was a composer that made a piece about a sphere, Emily Howard, and then we started exploring different geometric shapes that she'd never even heard of before and she was able to make a piece about something called the anti-sphere which is a very different geometric shape that she wasn't aware of, so. So I think that's something really exciting. And she's not trying to explain the shape to the general public. It's not a tool for scientific dissemination. It was a stimulating new structure for her to spend time exploring, to create something that, because it is related to a structure which is there in the universe, means that that piece will probably have some resonance for an audience, even though they don't understand what the structure is. Because... It's still sort of tapping into something fundamental about our universe.
Speaker 4 [01:21:38] I didn't really answer your question. It's okay. Are you following up on that? I'm not sure whether I am or not. Well, it's this idea of artists and scientists. Wasn't there a time, the ages, eons ago, where the art of science wasn't siloed?
Speaker 3 [01:22:00] Yeah, yeah, yeah.
Speaker 4 [01:22:02] It's an obvious candidate.
Marcus du Sautoy [01:22:07] Of course the tragedy is that there was an age when nobody would see the distinction between art and science. I mean, in some ways I really wished I'd lived in Leonardo's time because he was quite happy to be able to do mathematics, science, inventions and his extraordinary art. And nobody, I think, really batted an eyelid that he was doing all of these things together. And it is kind of a tragedy of where we've arrived in our modern age that there is... This separation between the two. Maybe it's because each of them has become so complex that it's very hard to do both. And so, you need to dedicate yourself exclusively to one thing in order to be able to do it so well. It's quite hard to be both. Brilliant at mathematics and brilliant at music at the same time. So You know, maybe in the past it was easier to To kind of navigate these things because we haven't reached the level of complexity that we're at today
Speaker 2 [01:23:09] So, this is related, our project is currently named Confluence, Art, Science and Creativity and it reports that art and science, the intersection of art and science have informed creativity, culture and the world through the ages. Comment? Agree? Disagree?
Marcus du Sautoy [01:23:34] I think what we often underestimate is how creative you need to be to be a scientist. And I think that's why you'll find a lot of scientists actually that are musicians or amateur painters and things because, or even just love that world because it just helps their own creativity. So I think it's really important and it's I think something that I'm seeing in governments of basically saying, right, science is great, we're funding all of science, but. Creative arts is kind of a luxury. It's not a luxury This is what's helping scientists to make the great breakthroughs that they're being stimulated from an artistic creative Point of view so I think it's so important that these two Rivers continue to mix together because we will lose so much Scientifically if we if we abandon the creative arts as just some luxury for the rich
