Episode 7 - AI as Artist’s Assistant.

In this episode of Mind the Gap we talk with the President of Harvey Mudd College, Dr. Maria Klawe, who has had a distinguished career as a theoretical computer scientist, scholar, and college administrator.  Dr. Klawe is also a gifted painter -- a watercolorist.  And she has been painting for longer than she has been a computer scientist.  In this episode of Exploring AI Matters we explore with Dr Klawe her widely admired paintings and her use of technology to help her focus her artistic vision. [2022-11-14]

Transcript

SPEAKER_03 0:01

Welcome to Exploring AI Matters. This podcast series, previously known as Mind the Gap, Dialogues on Artificial Intelligence, will continue to appear in the ABA series to the extent that. In addition, all of the episodes, old and new, will now appear under our new podcast name, Exploring AI Matters. Thank you.

SPEAKER_02 0:40

As humans evolved, we developed forms of expression that transcend the limits of language. We can express ourselves in drawing and painting, and music and dance. And we have done so at least since humans painted images of large animals on the Lasco cave walls and ceilings over 17,000 years ago. Humans are more observant, more expressive, and more clear thinking the more they draw. Computer scientists and practitioners now use AI to create drawings and paintings and to restore missing parts of old paintings. But wait, is it a painting if AI created it or collaborated in its making? To decide for yourself, watch the online video entitled The Next Rembrandt. And now consider this. In late June 2021, the Rijksmuseum in Amsterdam announced the completion of a project in which museum staff used AI to recreate and restore large parts of Rembrandt's painting, the Night Watch. Those parts were cut off to fit the original painting into a venue in the 18th century, removing entire figures with the missing parts temporarily restored. We can see what Rembrandt's original looked like. In this episode of Mind the Gap, we will be talking with the president of Harvey Mudd College, Dr. Maria Klave, who has had a distinguished career as a theoretical computer scientist, scholar, and college administrator. Dr. Clave is also a gifted painter, a watercolorist. And unlike most of us, she did not stop drawing as a child and has been painting for longer than she has been a computer scientist.

SPEAKER_03 2:34

Hello, I'm Roland Trope, a national security lawyer. And I'm Mark Donner, a computer scientist. We are your hosts for this episode of Mind the Gap Dialogues on Artificial Intelligence. In addition, we have two more hosts.

SPEAKER_01 2:51

I'm Mama Adams, a national security lawyer.

SPEAKER_03 2:54

And I'm Charles Palmer, a computer scientist. Each episode will be led by two of us, with the other two adding impromptu questions and comments as the spirit moves them.

unknown 3:05

Dr.

SPEAKER_02 3:05

Klawe, can we start by asking you what prompted you to start painting?

SPEAKER_05 3:11

I have always painted. And it's just never occurred to me not to be painting and drawing.

SPEAKER_02 3:43

Since we cannot show your paintings on our podcast, can you describe what and how you like to paint and draw?

SPEAKER_05 3:51

So the first thing is I have a very strong interest in patterns and particularly patterns that occur in nature. So everything from flocks of birds to uh patterns of pebbles under ripples in a shallow stream to waterfalls, to clouds, I can just go on and on. So that's that's one part of it. And and I think there is some overlap between my interest in patterns in mathematics and patterns in nature, but I also do uh a lot of portraiture. I um more and more as I've gotten older, I've been doing paintings of children, of adults, of dogs and cats, um, all kinds of things. And um, and then I also do abstractions that are motivated by uh any of the above or just by shapes and patterns.

SPEAKER_02 4:51

Well, I noticed, for example, in your paintings, you have one of Ollie in the Gardner, which is a child portrait, but you also have one entitled Reassembly Required, which it looks like a self-portrait, and yet it's broken up into several patterns so that it's a combination of abstract and figural. How do you choose which one you're going to go with?

SPEAKER_05 5:14

Well, in that particular case, that was during my um second year, the first semester of my second year as President MUD, and I was working with an executive coach because I had gotten some feedback from faculty that they didn't like my style. They thought I was too blunt, too outspoken, um, a variety of other things. And so I was working with this executive coach, her name is Sabina Nawaz, she's fantastic. She said, Maria, as you work through this, there are going to be times when you feel a fair amount of frustration. So I think it would be good for you to do a painting that you can work on throughout this executive coach's coaching process so that you can express how you feel about it. And so the painting reassembly required was just this idea that I was going to be taking the person I was, taking the way I interact with the universe, and modifying pieces of it. And so the painting that I did was, in fact, I think seven uh different uh self-portraits from the same photograph uh in slightly different colors. And the painting starts out that I'm happily smiling, and then my head explodes, and the pieces go all over the place, and they sort of gradually come back together. Now, my son looked at this and he said, you know, it actually would look much more positive if you read it from uh instead of from left to right, which is the way it was painted, from right to left, because then, you know, here's your face that's a mess with lots of different pieces, and it comes back and it's completely back in one place. And I said, uh, yes, but I'm I've hung it opposite a mirror. And so you can look at either one. And in fact, it's hung in the hallway of the president's house uh at Harvey Mudd College, and you can either look at it from left to right, which is how it was painted, which is called Reassembly Required, which basically my head exploded and then it had to be put back together, or you could read it the other way, where my head was completely sort of misaligned and in pieces and it came back and it was perfectly together.

SPEAKER_02 7:34

In the telephone call we had with you a few weeks ago, you told us a fascinating story about walking in a lake and seeing two ravens and then painting them. Could you tell us about that?

SPEAKER_05 7:46

Yes. So um, one of the things I love to do is, and I started doing this when I was a marathon runner, and uh to run marathons, you have to do a lot of running per week. And my legs really sort of couldn't take it. So I did half of the running actually in deep water. And one of my favorite ways to exercise is to actually run uh in deep water in a lake and then be able to watch the birds. And uh there's a lake called Lake Arrowhead that's about an hour's drive from Harvey Mike College, and I was uh in late summer of last summer, I was uh running in water, deep water, and there was this pair of ravens that were just doing the most extraordinary aerobatic uh display. And uh and I later looked it up. I I was pretty sure it was a mating dance because I've just never seen ravens come so close together and do so many complicated maneuvers. And uh and so later I did uh a pair of ravens who were um to demonstrate what these ravens were doing, and I sort of tried to uh articulate the motion by doing lots of sort of geometric designs that came out from their feathers and their wings and so on. And it's called raucous ravens.

SPEAKER_02 9:14

You you mentioned you were painting from your earliest age, you can remember, and that you had a painting of autumn leaves that you did at school. But the from our previous conversation, I understood that um at some point you stopped letting everybody know, especially your computer science colleagues, know that you painted, and then later that changed. Could you describe that process or journey for us?

SPEAKER_05 9:38

Yes, that is definitely correct. So um, so first of all, I'm female and I'm a mathematician, computer scientist, and it's just a little bit difficult uh given the era in which I um you know became a researcher to be taken seriously as a woman. And I have been the first female to hold my job for the last 33 years, and so I it just felt like um it felt in the early stages of my career, once I had finished my PhD, I mean, I was openly painting as a graduate student doing my PhD and uh and actually had an exhibition and all that kind of stuff. And then when I actually started work, I just decided, oh, it's hard enough being female in these fields. They don't need to know I'm an artist as well. And and so I really, I mean, definitely my family obviously knew, my very close friends knew that I painted, but uh it was not something I ever talked about at work. And the year that I turned 40, on my 40th birthday, I had 13 paintings framed, and I hung six in my office. I was the head of the computer science department at the University of British University of British Columbia and the others in our house. And we had a lot of department events at our house. Um, and uh it was just sort of my way of saying, okay, I'm here, and I I want to say I turned 70 uh on Monday of this week.

SPEAKER_06 11:14

Well, happy birthday.

SPEAKER_05 11:16

Thank you. And uh it's really nice to think that that was 30 years ago that I did that. And I will say what was interesting to me was the dean, whom I reported to, the dean of science, came by my office and he looked at these paintings and he said, Oh, you must find it so relaxing to paint. And I went, No, this is just as intense a creative activity as doing research or doing anything else. It's something that is just as important to me. And he sort of like backed out of my office. He clearly knew he'd put his foot in his mouth.

SPEAKER_02 11:54

Now we've heard that you paint during meetings. Meetings are an intense activity, and I would think a lot of people, if they're seeing you paint or draw during a meeting, were wondering: is she listening? Is she participating? How does that help you? And why do you do it?

SPEAKER_05 12:10

Well, this started when I was became dean of science, um, which uh I took that role at UBC in 1998, and I had always done my paintings on weekends, and unfortunately, um deans of science across the country got together for you know every several months for meetings on weekends, and I found it intensely frustrating that I couldn't paint, but I also have always had trouble keeping myself from talking too much. And so I decided that I would try painting at the meeting as a way of keeping myself more quiet. And I was very lucky that the person who is the Dean of Arts and Science at the University of Toronto, whose name uh is Carl Amrain, he's the kind of person who's very twitchy at meeting. And he found that if he sat beside me and watched me paint, it calmed him down. So it kept me to be both of us to be quieter by the fact that I was painting. Now, um, for those of you who are not Canadian, Canada is sort of in Canada, the University of Toronto is the number one university. And so the fact that the dean from the University of Toronto clearly thought the fact that I was painting at meetings was totally fine, made it okay. And it somehow became something that I just did at pretty much all meetings. And so, for example, uh on uh on yesterday we had a cabinet retreat for the senior leadership team in Harvey Mudd, and of course I was painting throughout this retreat, but what was interesting about this particular one time was that I had been asked, we just uh lost one of our vice presidents to Washington University in St. Louis, and she asked if I would do a painting for her as a going-away present that would be a Zoom of all nine people on cabinet. And so basically a small, you know, like a three by three portrait of each of the nine people arranged as the windows of a Zoom. And so, of course, at the meeting, I was doing those portraits of the people, a cabinet. So um yeah, it's everyone sort of takes it for granted now. The one place where uh people were not enthusiastic about my doing paintings were on uh two public company boards, Microsoft and um Broadcom. And uh and so I didn't paint at the those board meetings, but I joined another uh startup company board about not quite two years ago called Glowforge. Uh the founder, Dan Shapiro, is uh a muddlum. And I told him about how much I love to paint at meetings, and he said, of course you can paint at Glowforge meetings. So Glowforge meetings, I show them what I'm going to paint. And at virtually every other meeting I'm at, um, even if I'm doing it on Zoom, I show people at the beginning at the end what the painting is that I'm working on.

SPEAKER_02 15:33

I asked one last question before I hand off to Mark, which is you've been you've been painting with knowledge of your colleagues for 30 years. Has there been a significant change in your style during that time?

SPEAKER_05 15:47

Oh, that's a really interesting question. I I think there is a significant difference. Um I I think there's I I would say this, I don't want to say abstraction, but uh I would say in earlier days I was doing things that were more landscapes or portraits, and they were I was focusing less on the pattern. And I would say over time I've tended to do series of paintings about one particular thing. So I have a series called flocks, which is of flocks of birds. I have a series of uh snow trees, which is uh just after a very heavy snowfall. I have a series of um pebble patterns. And so I would say the the thing that has become more pronounced as the years have gone by, I'm I'm typically working on somewhat larger works, and I'm working on things that uh where the patterning is more explicit than it was in the past.

SPEAKER_03 16:55

How have you been using or have you been thinking about using AI in your uh composition and painting?

SPEAKER_05 17:02

I think the way that I so first of all, I have not used AI yet, but I have talked to actually some other artists that are used in AI. And um and in fact, I was so intrigued with what A.J. Kumar is doing at CalArts that um I think when I retire from Harvey Mudd, I'm gonna go take his class at CalArts so I can actually try some of the things he's doing. I think the thing I would love to do is explore using AI to look at different kinds of patterns to be able to provide. So, for example, um a painting I'm working on recently is of a starling murmuration, which is for those of you who watch birds, you often see huge thousands of starlings coming together in all different kinds of patterns. And I'm very interested in the shapes you get from overlapping birds in sort of both uh both what you get uh in the foreground, what you get as in the background. And um I would love to actually do uh a visual an AI computer recognition program uh on a variety of these shapes that I particularly like, and then ask uh the program to generate similar kinds of shapes and explore just exactly what would show up by doing that.

SPEAKER_03 18:33

Your your comment about watching birds flying in large numbers reminds me of some I I saw a bunch of galas uh swarming in in the sky in Australia some years ago, and it is uh astounding. The uh I don't I can't describe it, it's just I was mesmerized for like an hour just watching them fly. Um so I get a sense of what we're talking about. So, how how do you think um AI might create opportunities or tools for painters?

SPEAKER_05 19:08

So uh I think the reality is that uh AI is very good at recognizing patterns. What I've been talking about is doing paintings that are very much based on patterns. And I think, I mean, I think it's really possible to use AI as a tool that looks at a set of patterns that I find interesting and have it try to generate patterns that I would also find interesting that are different. So it's sort of like taking my imagination and pushing it way out, and then letting me have the opportunity to, you know, be in feedback with the AI to say, well, I love this set, but that set I don't like at all, and to sort of fine-tune it. And I think there's just an opportunity to explore much more quickly in more depth and more broadly using AI. And you know, from talking to AJ, it's pretty clear that this is being used uh both in music and in visual arts uh pretty significantly right now, and uh certainly in areas like animation and you sort of virtual worlds and video games and all those kinds of things, but also by people who are really focusing on doing art for the sake of art. And I think this will just become more and more pro prevalent as time goes on.

SPEAKER_03 20:43

So, you know, some years ago, um, computers conquered chess, and it it seems like the interest in chess has uh died a little bit, though maybe Queen's Gambit has has changed that. It's hard to tell. Uh, do you worry about the use of AI reducing people's interest in painting?

SPEAKER_05 21:03

Uh well, it's not going to reduce my interest in painting. Um, not really, because I I guess the thing I would say is that I I think what most people that I know who are artists are really interested in is trying to understand how they can create visual representations, whether they're of an abstraction or of reality or something in between, that they think are communicating something. And I don't think, for example, like initially, I think a a lot of representational art you would have said, well, now that we can take photographs, I mean who cares about art? I think the taking of photographs just uh made things much much More wonderful for artists. So, for example, I used to do all of my painting outdoors, and I was largely doing landscapes. And one summer I was on Pender Island in British Columbia, which is at the time where we spent most of our time in the summer, and it was really hot, and you could not do watercolors outside. I was intensely frustrated. And this is back before digital cameras. But I went and I got my camera and I took all kinds of photographs and had, you know, got prints of them. I said, fine, I'll print indoors. I'll I will paint indoors. And there's no question that the experience was different. But it wasn't worse or better. It was just different. And so what I think right now is that yes, the fact that AI is being used to create art of a wide variety of types.

SPEAKER_06 23:06

That's not going to make human paint humans less interested in painting.

SPEAKER_05 23:12

It's give it going to give human humans who want to be artists a greater set of tools.

SPEAKER_02 23:19

What about the group of humans in between? Not the ones who regard themselves as artists, but just want to draw sometime or paint sometime. Do you think AI will be appealing to them as tools also?

SPEAKER_05 23:31

Well, so first of all, I think absolutely everyone is an artist. And I think I might have told you the story about my younger sister Katrina. For her 50th birthday, I had taken her to uh the Mayan Riviera. And um, I of course took my paints with me because I paint everywhere. And uh as I'm taking my paint out, we're about to go to the beach and I'm gonna paint. And she says, I've always wanted to be an artist, but I have no talent. And I said, Of course you have talent, everyone has talent. And here she is, she's uh 66 this year, four years younger than me, and she is a phenomenal artist, and she has spent that time, the last 16 years, developing, taking lessons, uh doing exhibitions, she's a carver, she's a uh dozer filix and oils. Um I I'm just so delighted that that she said this because if she hadn't said it, I never would have had the chance to say, Katrina, of course you're an artist. So whenever somebody tells me that they're not an artist, I just say, you know, you just haven't given yourself a chance. Like art art is like everything else. Um there's there's stuff to be learned, there's stuff that some people will learn faster than others. But the reality is that everybody is capable of creating things that are of interest to them and to others. And so anyone who wants to draw, yeah, I think AI tools could actually give you an opportunity to learn more quickly. I mean, it feels to me, I haven't tried to do this, but I'm gonna I'm gonna learn more from AJ, and maybe that's what I do after I retire from Harvimud. I start to teach AI and art. Um I think it's entirely possible that using AI, one could empower people to learn more quickly and to explore more broadly, even if they don't have a lot of experience drawing or painting or just doing things visually.

SPEAKER_02 25:52

It sounds like you think that painting is a deep part of your identity. Is that correct? And if so, what does that mean to you?

SPEAKER_05 26:00

Oh, I can't imagine not painting. I mean, I just it's I mean, just as you said at the beginning of this, it's a way of expressing things I find difficult to express in other ways. And it's a way of exploring ideas that I find difficult to explore in other ways. And I mean, I do draw, um, but I like painting better because I've become so accustomed to painting in watercolors that it's like it's just part of my thinking about how if I want to explore a particular kind of thing, I'm gonna try it with watercolors because that's what I'm accustomed to using.

SPEAKER_02 26:46

Does your college or any of the other Claremont colleges encourage students in the arts to explore how they might use AI in their work?

SPEAKER_05 26:55

I honestly don't know at this point. Um we have a very um talented visual artist as a faculty member, Ken Fendel, who uh was the chair of the photography department at the Art Institute of Chicago, and uh decided to move to Harvey Mudd. And you know, I asked him when he came, I said, you know, that's one of the top photography schools in the world. Why are you coming to MUD? And he said, because I want to be teaching people who don't necessarily think of themselves as artists, which I just found incredibly interesting. Now, I do know that Ken co-taught a course on fluids with a math professor who taught the dynamics of fluids and he taught the photography of fluids. I don't think they were using AI, but I think they were trying to understand the relationship between the mathematics of how fluids flow and the visual effects of how fluids flow. So I don't know whether we have taught a course like that at Harvey Mudd, but I know for sure they're virtually all art schools are teaching them now.

SPEAKER_02 28:20

And do you think students seeing that you paint that your engineering students themselves are encouraged more to try to do that during when when they have time?

SPEAKER_05 28:32

I I think one of the things about Ahavi Mud that has nothing to do with me, but I'm a good ambassador for it, um, right from our founding, it was believed that it was very important, even though uh all of our majors are STEM-related, it was believed to be very important for students to take a significant number of courses in the humanities, social sciences, and the arts. And the reason for that was that we wanted to graduate engineers, scientists, and mathematicians who would understand the impact of their work on society. And it was felt that if you didn't have a strong background in humanities, social sciences and arts, it's hard to do that. So, because of that, we attract students who love some area in we refer to as HSA, humanities, social sciences and the arts, as well as uh and areas in science and engineering and mathematics. And so it's that we take it for granted that people have passions that extend across that particular bridge.

SPEAKER_03 29:48

So uh the classical masters were famous for having workshops full of apprentices to whom they taught their skills. The output of these workshops were generally signed by the master. Uh contemporary art historians have pierced the veil a bit and given us some insight into this practice. Um, have you uh worked with an apprentice or an assistant? Uh would you?

SPEAKER_05 30:17

Oh, that's a great question. So, first of all, um no, I have not at this point. If I had uh if I had a student or another, you know, any other person who said they wanted to try doing something jointly together, I'd be happy to try it. But it wouldn't occur to me not to uh have their signature on it as well. And and I would think I would want to do it less as I'm the master, you're the apprentice, and more as let's discover what we can do together as a team versus what we can do as single artists.

SPEAKER_03 30:59

Yeah, I I think the uh the motivation behind it in the in the classical masters was uh commercial. You wanted to sell more paintings, and so you just needed them to look as much like the master's work as possible. Um describe the use of AI as augmenting human intelligence, uh, to borrow a phrase from the late Doug Engelbart. How would you view an AI in the role of apprentice or assistant?

SPEAKER_05 31:29

Oh, well, I guess I think there's sort of two ways of viewing it. One of them is as a tool that's just like using photography or you know, various other kinds of things. Um, you know, I I will often work with Photoshop in terms of being able to size something that I want to or to compare things together or whatever. And so I to some level it's and I haven't tried to check whether Adobe offers AI in some of their tools, but they probably, if they don't now, they will soon, and they may well do that. It's not something I've particularly gone looking for, but uh there's a certain extent to which um so I happen to be crazy about cats, and I'm also crazy about my grandchildren, and I think of them as providing a different perspective on life, and I would sort of think about if I were partnering with AI, I would think of it as adding a perspective, not so much being my apprentice, but as giving me a way to see things that I wouldn't see without the help. And and that's very much the way I feel about having the joy of being able to interact with grandchildren, is because they're and my grandchildren are two and a half and one, because they're at a different age, they just see to things things differently. And if you're paying attention, it's an enormously enjoyable learning opportunity.

SPEAKER_03 33:22

What do you view as your most important tools and techniques in painting?

SPEAKER_05 33:27

I think it's important to me to be able to control color and the balance of light and dark and shapes. I would, I mean, I think one of the things that's important to me is that if I want to make something look like something, then I can actually do that. Now, I don't, I mean, I've done these crazy, crazy pattern paintings. These are the uh pebbles under pebbles in the shallow stream under ripples, and they don't look like anything. And I'm just I'm just sort of like mystified by I can't come up with a mathematical explanation for why the certain certain kinds of repetition of patterns is going up. Um, but it is important to me that I'm actually able to create something the way I want it to look. So I, you know, I'm uh when I do portraits, that's incredibly important because people either recognize people or they don't. And it's never entirely clear to me like you'll be working on something and it won't really look like somebody, and then all of a sudden it will flip into looking exactly like that person. And it could be that it's a shadow on their face, it could be that their hair is slightly different, it could be their ear, it could be their neck. I mean, it's bizarre. And so this ability to sort of have an idea of what you want to see, and then knowing when you can see that, that's something that's important to me a lot.

SPEAKER_03 35:09

Yeah, recognizing people's faces is a fascinating thing. Some years ago, I had a full beard and had had it for many years, and then one day, uh, prompted by uh my my girlfriend who's now my wife, um, I shaved it off. And I shaved it off with no warning. And it was fascinating because some people didn't know who I was. Some people said, Do you have new glasses? And some people didn't notice that there was a difference. And I I have to assume that each of them, you know, used different cues to recognize me.

SPEAKER_05 35:46

Well, I have to say I had the same kind of reaction. So I have a very close friend who uh uh Charlie will definitely recognize the name. His name is Bob Tarjan, he's one of my best friends, and he's a very well-known theoretical computer scientist. And um he used to have long hair and a beard, and one day all of a sudden he had short hair and he was clean-shaven, and it took me like days to get over the difference. I mean, I could tell it was the same person, but how could he possibly have the same personality when he looks so different?

SPEAKER_03 36:26

Charles has a question here. I I uh this is I'm just sitting here awestruck, but one of the challenges we have with AI is you you gotta train it, you gotta teach it. And you know, as as we as you said earlier, uh it's kind of hard for an artist or any sort of expert to put their finger on exactly what it is that allows them to do what they do. So I don't think we'll bother going down that uh path when training AI for painting or art. But what we seem to be doing now is training them uh you know, here's a bunch of examples. So we might show them all of uh Maria's art and uh you know, okay, go make another one, and maybe it would learn or not. What troubles me, and part of my question here, is just like when we're trying to train them to to recognize stop signs or cats versus golden retrievers, sometimes they pick up on stuff that you know humans would not have picked up on and use that as the differentiator between a cat and a golden retriever. So I'm wondering, do you with all this challenge, is AI gonna if it starts doing art, are we gonna find whole new paths that may or may not be appealing to anybody but the AI?

SPEAKER_05 37:54

I I suspect yes. I mean, I'd be surprised if if if not in some sense, because I think we are very conditioned by I mean, there's just no question that we're conditioned by what we see as being quote-unquote beautiful art or important art, or you know, I I I can come up with lots and lots of different adjectives, but basically, so for example, when um I took a fine arts course as an undergraduate, I think I was a second-year student, and uh it was a printmaking course, and the faculty, not the people teaching my course, but the faculty who were in charge of the different sections of that printmaking course, they were really into sort of cloud-like soft blobby objects. And I wasn't interested in doing that at all. And of course, I got a horrible grade in the course, and and which is fine, but it was sort of like I was sort of going, like, why, if you're gonna do prints, which you can do all kinds of different things, why are soft blobby objects the thing that's thought to be beautiful right now? So there's absolutely no question that over time uh society views certain kinds of things as wonderful art and other kinds of things as trash or sprite or whatever. And there's just no question that that influences how we see things. So I'm pretty sure that once we once we actually start training, well, I mean it's already being done, but that as the training of AI on various kinds of whether it's animation or whether it's watercolors or abstraction or whatever, we're going to see that there are things that that AI is identifying that's not that we had not thought about. And I think that's good. I think it's it's just like um now we we know that AI can do a better job of reading x-rays on average. And uh we're gonna deter, we're gonna learn new things, and I think that's good.

SPEAKER_03 40:27

I I I I'm triggered by one of your words, you use the word beautiful, and I have to say that I some of my favorite art uh is not beautiful. I don't I wouldn't call it that. Think of the scream, okay? It's immensely powerful, it's very moving, I adore it. I would never accuse it of being beautiful.

SPEAKER_06 40:49

So I wouldn't either.

SPEAKER_03 40:51

So I I I try to avoid the word beauty in relation to art because I think that trivializes the message to a certain extent in my mind.

SPEAKER_05 41:02

So I wouldn't agree with that because I think that beauty is just one component of what could make you value a piece of art. And emotion is another one, which I think is what comes across in the screen. And a statement of, I mean, there's a lot of social justice art right now that is really making a statement about society and about inequity and unfairness and all kinds of things like that. And so I think art can be, it can convey a variety of messages, but I don't think that, I don't think art has to be beautiful to be important. I don't think that art being beautiful means it's necessarily not important.

SPEAKER_02 41:51

I want to go back to something you mentioned earlier about how you've moved uh over time further away from representational art and more towards abstract and patterns. And it reminded me of uh an exhibition that my wife and I saw before the in the fall before the pandemic started in Paris. It was a retrospective of Juan Mirot's paintings. And at the beginning of his career, he did portraits, and then he moved further and further away from that into those wonderful abstractions with the gorgeous blue background. And it's a pattern that's similar to Mondrian. So the question that this leads me to ask is for artists, do you think AI is going to be most valuable by giving us, as you mentioned it, patterns to look at that the artist may then add to their painting themselves with their own brush, or having the AI generate the pattern and that that be then added to the painting either as part of a 3D printout or some other means?

SPEAKER_05 42:59

I think any of the above. So, you know, I think if I um just like watching the Ravens in the meeting flight, I can think about that influencing what I'm going to do in lots and lots of different ways. So, for example, uh, you Roland sent me a photograph of some eagrets. And um, and I'm uh I'm just starting working on that piece. And the fact that you sent that, I'm I'm thinking about doing a whole series of pairs of birds in flight. And what's actually most interesting in that photograph you sent me is the way their reflections are distorted in the water because of the ripples in the water. So just by sending me that photograph, you've made me think about a whole

SPEAKER_01 44:00

series of other paintings I might think about doing which really have to do with a pair of birds because uh because I find birds extraordinarily beautiful but also ways in which patterns either of their movement but in this case the reflection on rippled water causes you to see patterns and think about extensions of those patterns and and so when I think about using AI it's all about seeing suggestions of ideas and wanting to iterate on them and magnify them and morph them to just do things with those ideas and those ideas could come from anyone and yes I'm gonna send you a photograph of the painting when it's done um but it could come from you know it could come from a grandchild who is fascinated by my grandson is fascinated by water running down into sewers I can definitely think about doing a bunch of paintings that really has to do with waters going into grates um and and he would totally adore that but you could get this from anywhere and so I see AI just as another form of creating ideas suggestions images patterns whatever I think the way you framed it earlier um you know in all these podcasts we've done I haven't heard anyone say it quite like this but the way you said it in the sense that AI provides a different perspective right a different avenue or approach to thought and I thought that was a really clear way of kind of expressing the tool or the opportunities that AI could present particularly for an artist like yourself. And I was just curious listening to this conversation in a previous life I wanted to be an art history major you know uh my 11th grade teacher did not say nearly as kindly as you did that we are all artists as she was making separate projects for me because she did not deem me to be an artist. But what kinds of other perspectives are you interested in we spent some time talking about patterns you've given this great example about sort of the distortion of the ripples on the water are there any other perspectives that you're interested in pursuing in your artwork that you think perhaps AI could be a helpful tool or opportunity for you?

SPEAKER_05 46:39

Well I actually think that children when they first start to draw and you know we have these sort of ideas about what a stick figure looks like and what a house looks like and all those kinds of things. They're very traditional and I don't mean so much that but when my son who's actually the father of the grandchildren start to draw he was doing these amazing sort of abstract figures and shapes and stuff and I actually thought that he would become you know that he would pursue art and I could never persuade him to do that. And um you know he's a math computer science geek like his parents he does stuff that's at the interface of AI programming languages and software engineering. But you know I think there's I think there's a lot to be learned from what people produce when they haven't been influenced by what your teacher told you was a reasonable way to produce a a mummy and a child or whatever, right? And I think that just like being interested in what AI might do, I think there's sort of a there's a there's a a lot of possibilities there. And you know I remember um there's somebody I met in 1974 I don't remember it was 1974 but he does he's an artist he's his name is uh Danik Wzgensky and he's a a very successful Canadian self-taught Canadian artist who does sculpture and uh portraits and a variety of other kinds of things. When I first met him he had spent I think maybe two months up with the Haida and had gone there to learn their perspective on art and I mean I I just feel like we're not encouraging enough of people being willing to sort of go and ex just explore things from people who have different perspectives. Now obviously the Haida are are incredibly successful artists and um but this was long before their art was you know uh as cherished as it is today. And yet he as you know uh he must have been well let's see I was 23 so he was 22 at the time um had chosen to go spend months with them to understand their perspective on the world and of course many artists have done that by going you know I to spend time with indigenous tribes in various parts of the world um and so on. But I I just feel like think of AI as just another lens on the world.

SPEAKER_02 49:38

Can I ask you you've mentioned we we we we know that linguists posit that children have an innate ability to start to acquire language with your children and and your grandchildren now do you have to show them a little bit about drawing or do you just have to give them a tool and do they start drawing on their own initiative so to speak? I had Brahma's experience the a few teachers that looked at my drawing said it was terrible and I stopped but uh and then resumed it years later just because I found I looked at things more carefully but I've always wondered do children just have an instinctual desire to draw and you don't need to do much to coax them to it.

SPEAKER_05 50:26

So I I think it's impossible to answer that question because well at least within our culture because I think about how the time I spent with our own children and the time I'm spending with my grandchildren we spend so much time with books and those I mean the first set of sort of stories you tell them have beautiful illustrations of different kinds and so I think it's and and of course they're generally they are drawings or paintings they're not photographs. And so we are starting the education of our children at very young ages with with a variety of kinds of art forms. And and so I think it's almost impossible that they wouldn't have the idea that they might reproduce some of those kinds of things. And and if you look at I mean so one of my uh favorite uh artists is Eric Carl and um the very hungry caterpillar and um but there's there's there's also uh a set of um books right now from a um who's the n I'll forget the name of the author he's French but um that's about uh combining colors to get different kinds of things and combining shapes to get different kinds of things and so on. And so I feel like we can't ask that question given that we're exposing our children to so many forms of semi-representational very representational almost not representational at all art in the books that they're seeing. And so I think it's almost automatic that when they go to their first play school or you know when they're given uh you know a whiteboard and markers um they're also they're being introduced to things like shapes like circles and rectangles and triangles and all of those kinds of things and all of this is that's also about art. So I I I think we we are immersing them in so much visual representation that I honestly don't know what they would draw if they hadn't seen any of that or whether they would want to draw.

SPEAKER_02 53:07

Which makes it to me all the more puzzling that some boards would have told you they didn't want you to paint during a meeting well I think they thought it wasn't respectful enough.

SPEAKER_03 53:19

Wow yeah one one of my colleagues um in a rather well all the senior managers in the in the the larger department had this one operations meeting that we had to go to every week or two and uh she was likely the most brilliant person in the room and she proceeded to knit throughout these meetings and throughout as we went through one VP and the next one and the next one she continued to knit and was able to keep full track of what was going on in the meeting. One of those uh annoying students that can be fooling around with their phone and then you ask them well do you understand what I just said and they repeat it back to you. Well she could do that. Do you find that um the painting makes it even easier for you to focus on both I mean yeah I mean I definitely I'm a much better um listener if I'm painting there's there's just no question.

SPEAKER_05 54:21

That's amazing. But I think well but I think it's partly because if I'm painting I'm not trying to think of a question to ask. Yes I like focusing just what I'm hearing. So when I first went to Princeton as the Dean of Engineering well in fact when I was interviewing for that job I was talking to a very well known biologist and I asked him if he thought it was going to be a problem if I painted at meetings. And he said well I like to knit at meetings and I sort of went and he said the only problem is that sometimes people think of it as um an element of disrespect and so I sort of have to be careful to explain ahead of time why I'm knitting. And I thought that was actually helpful but turned out when I uh came showed up at Princeton the uh the president Shirley Tillman absolutely loved that I painted at meetings so it was fine and I think for her it was uh it was just representing that uh you know you can be a scientist or an engineer and also be an artist or a musician or anything else and that that's all fine.

SPEAKER_02 55:40

But it's interesting that you're comfortable having others watch you while you're painting and actually see what you're painting.

SPEAKER_05 55:47

You mentioned that the president of Toronto University sitting next to you I take it he was able to see some of what you were drawing that calmed him but it doesn't make you unduly self-conscious to have somebody watching you yeah I don't think it bothers me at all at this point and I have no idea why but it's just I'm completely I'm both immersed in the meeting and immersed in the painting and the fact that somebody's watching so I had an experience I got elected to the board of the AAAS the American Association for the advancement of science and when I went to the first meeting I was painting um a portrait of two grandchildren of the chair of the board of Pittsarge which is one of the other Claremont colleges and he had watched me painting at a number of board meetings we have a shared board for the the Claremont colleges where it's all the presidents and all the chairs of their boards and he'd watched me painting a number of times and he finally said you know I was wondering if he will paint it for me and I said of course it's a picture of his grandchildren and I was doing it the person sitting beside me at the meeting was uh Carolyn Ainsley who's the CFO of the Gates Foundation and she just sat beside me the entire time and was just totally amazed at watching this painting just appear and yeah I just don't think about it anymore. It's just you know something I'm doing.

SPEAKER_03 57:27

Would you go ahead Charles I I I didn't mean to interrupt I was gonna say your your comment about um technical people uh and artists uh it's just another expression of their abilities I've always followed the rule uh or realized the rule uh the best programmers are musicians or bridge players and to me that means patterns and uh I guess there's patterns in art as well.

SPEAKER_05 57:59

Yeah absolutely and well you know it's one of the it one of the things that I found really quite uh ironic when as I went into my career as a computer scientist and mathematician was that nobody thought it was weird to be musicians or bridge players for that matter but being a visual artist was weird and I think now it's less weird which is all good.

SPEAKER_02 58:29

Am I correct that you wouldn't want AI however to be commenting on your work or suggesting anything to you while you're doing it if it ever got that capability oh I don't think it would particularly bother me.

SPEAKER_05 58:45

I think I would just ignore it I mean I I would probably listen and I would see if I agreed and if I didn't agree I would just ignore it.

SPEAKER_02 58:56

Well can can I ask a question of sort of the elephant in the room which is if AI were to create a painting itself do you think it's important for the observer for other people when they come to look at the work to know that it was generated by AI rather than by a human?

SPEAKER_05 59:14

Do you think that it's important to keep that distinction or when AI has been a co-collaborator or co-creator of a painting that that should be highlighted rather than having the AI be given a name like Watson and somebody thinks that it's there for a person um I think I would probably want to know just because I would find it really interesting you know I mean I well let me talk about myself if I was to do a piece that I did in collaboration with AI I would definitely want to say this because I think it's that would be interesting just in its own right that that I had chosen to do that that this is what came out of it and you know I would probably also want to uh you know have some kind of narrative about how it was different from something how it had changed what I had produced and I think I would want to do that just to educate people that this is possible. And similarly if something was completely produced by AI now you know it's not going to be completely produced because it's going to be trained by someone you know I mean it's gonna come using some kind of system that somebody created but you know there's uh so there's a guy named Perry Cook who was uh a faculty member in the CS department uh while I was at Princeton and he's he's done some really interesting things and his students have done really interesting things so AJ Kumar did his um the person I mentioned earlier at CalArts in his page with Perry Cook Princeton and so his students have done things in between music and AI and in between art and AI and in between both and AI. And you know my sort of feeling is that one of the ways we educate people about how the world is changing is to let people know when AI is involved in a decision. And you know similarly so I'm on an uh an advisory committee for the Schwarzman College of Computing at MIT and um one of the things they are doing right now is a series of policy forums about between AI researchers and people in government that are implementing policy that might be influenced by AI. And so I had a conversation with one of those one of the faculty members at MIT who's involved in that and you know I think we just it's just important to it's important to be willing to explore what the possibilities are and to educate people about what the policies are or what the possibilities are and to actually you know it would be from my perspective it would be a mistake not to tell people if something had involved AI because that's a good way for people to learn about what AI can do.

SPEAKER_02 1:02:25

Dr. Clave you've been very generous with your time and even more generous with sharing your thoughts and observations in response to our questions and I think all of us have thoroughly enjoyed the conversation and we want to thank you very much for being willing to join us on this podcast.

SPEAKER_05 1:02:43

Well I just want to say thank you for having me I've enjoyed this conversation and of course you guys have stimulated my interest in actually doing some artwork with AI we thank the business law section of the American Bar Association for their generous sponsorship of the production of this podcast we welcome questions and comments from listeners.

SPEAKER_02 1:03:12

Send email to comments at MindtheGapdialogues dot com we read all comments and questions and we'll try to respond in the letters section of a future episode if you are writing about a particular episode please do mention a specific episode number please also do include pronunciation tips to help us properly say your name when we reply in a subsequent episode see you next time on Mind the Gap Dialogues on AI Thank you for listening to the AVA Business law section's podcast series to the extent that the section offers a robust collection of content.

SPEAKER_00 1:04:09

To explore more about this topic or to learn about joining the section visit ambar.org slash fist law that's P I D L A W