Ned McGowan - AI in artistic research

Mikal: Welcome to the podcast HKU and AI.

Mikal: In this podcast we will talk to the HKU and alumni about how they AI in their work and

Mikal: process.

Mikal: In every episode we will talk about one aspect of AI, the reason for a concrete experience

Mikal: of experience.

Mikal: Hello, I'm Mikal van Leeuwen.

Babette: Hi, I'm Babette van Rheden.

Mikal: Welcome to another edition of HKU and AI.

Mikal: Today we have in the studio Ned McGowan from the Conservatorium.

Mikal: So welcome, Ned.

Mikal: Thank you.

Mikal: Before we start, could you introduce yourself?

Mikal: Sure.

Ned: I'm a composer and flutist from the States, but I've been living here for 30 years.

Ned: And I teach composition and artistic research at the Conservatory.

Mikal: And how did you get to the HKU?

Ned: Well, I was invited to teach a class in Indian Rhythm and that got extended.

Ned: And then I was teaching advanced rhythm in several departments at the conservatory.

Ned: And I started my PhD.

Ned: And then recently I was asked to coordinate the research approach at the conservatory

Ned: on the basis of my experience in artistic research.

Ned: So if I can just jump right in right there.

Ned: Of course.

Ned: I would say that my context for looking at AI is all based on how we can use that in our artistic research for the students at the school.

Ned: Yeah.

Ned: And that's something I'm also pursuing myself in my PhD.

Ned: It's a PhD in artistic research, which is very different than scientific research.

Ned: But, you know, it is proper research.

Ned: And the art has a lot of different methodologies to expose new informations about practice and vision and individual insights and experience.

Ned: And the AI can be a useful tool for helping us to communicate, access, explore those ideas.

Mikal: Yeah, so you use the AI yourself, but you also teach your students how to use AI in artistic research?

Ned: Yes, we are exploring that.

Ned: AI is a multi-tool, so I'm using it in a number of different ways,

Ned: and we can talk about all of those.

Ned: But in general, things like brainstorming, having discussions about topics,

Ned: doing initial research into contextualization,

Ned: what is the current knowledge out there related to topics.

Ned: It'll take anything you throw at it and give you some kind of answer.

Ned: And sometimes it'll come up with angles and ideas that you never thought of.

Mikal: Yeah.

Ned: And that can be very useful.

Ned: You know, you can have a discussion about the structure of your thesis, for example, and it'll give you some feedback on it.

Ned: And then you can reflect upon that feedback and see if you want to take action.

Ned: Of course, proofreading, checking if your comma placement is correct.

Ned: It's good for these kinds of things.

Mikal: Yeah.

Ned: And it can also look up potential sources that you can use to cite in your research about ideas.

Ned: And, you know, one of the main things which probably we will keep coming back to is that everything that it gives you needs to be checked.

Ned: It needs to be thought about.

Ned: You cannot just copy and paste.

Mikal: Why not?

Ned: Well, let's look at that.

Ned: We all know that experience, right?

Ned: You're typing a question and it gives you a response.

Ned: And your first reaction is, oh my God, that's amazing.

Ned: I could not have thought of that or written that.

Ned: And that kind of has two different functions.

Ned: First one is it's now opened up a space, right?

Ned: So now you've learned about perspectives and things that are out there, which you did not

Ned: think of.

Ned: Like just for example, if you're a musician and you go to a new school, you're going to

Ned: meet other musicians who are going to play and do things that are amazing, amaze you or you didn't

Ned: know could happen before. And then knowing that that's possible makes it much quicker for you to

Ned: to go there yourself. So in that space, it's like the brainstorming function. It's opened up a space

Ned: from which you understand ideas or things in the world. Now, the thing of it is, though,

Ned: So coming back to that initial wow moment, if you then start to really look at what it's saying, then you say, well, you know, that's not exactly what I mean.

Ned: Or no, that's actually referring to something which I don't, which I'm not focusing on or what I don't want to focus on.

Ned: Or it's pointing in a direction which I don't mean.

Ned: And so all too often, its words fall apart under intense scrutiny.

Ned: And when you're going to put a research paper out there or report with your name on it, you have to stand behind every word and every comma.

Ned: And so those all need to be checked.

Ned: Every word has to be a decision from yourself from this is what I want to say.

Ned: So the result is that when you start going through it and finding these problems, then you have to go back and then you have to start working on it and say, well, no, I don't mean that.

Ned: I mean it slightly different, or I mean it in the context of this,

Ned: or I want to actually bring this point across, not that one,

Ned: and then you have to rework it.

Ned: So then you're going through multiple iterations of the same text

Ned: in order to get it to say what you want to say.

Ned: And this is where I think it's an effective tool,

Ned: because then you can spend your time focusing on ideas and arguments

Ned: and not necessarily on the nuts and bolts of which words to use.

Ned: It kind of gives you that meta function where you can just deal with the ideas.

Ned: And I think that's really important when we're talking about artistic research,

Ned: is being able to understand what is it you want to say, what is the argument,

Ned: how are you relating your practice and your vision and your intentions in the world of practice.

Ned: I think it can help you come to much clearer communication of those ideas.

Babette: Yeah, and how do you teach students this?

Babette: How to use AI?

Ned: Yeah, it's a really good question.

Ned: So what we've done is we look at the different facets of the artistic research,

Ned: which is about, you know, what is your research question?

Ned: Who are you?

Ned: What instrument do you play?

Ned: What is the topic you're interested in?

Ned: What is the contextualization, the state of knowledge related to that topic and that question?

Ned: And what are some methods that you want to do?

Ned: And with each of those steps, you can focus on that with the AI and say, well, what do you think about this question?

Ned: Is it a proper artistic research question?

Ned: And it knows what artistic research is.

Ned: And if it doesn't quite know, you might say, well, I'm looking at the European model of artistic research or something like that.

Ned: So then what we've done in the class is, for example, a few weeks ago, I said to students who had already written their research proposal, I said, put in your research question and ask it, what kind of methodology do you think I could carry out to answer this question?

Ned: And then everyone goes into their laptops and is like thinking.

Ned: And then we'd go around the room and with every person, they said, well, yeah, I'd got some things wrong, but there were one or two points that were really useful for me.

Ned: So then, okay, I feel like we've found a way to work with it.

Ned: Of course, when you talk about AI, what kinds of ethical questions come up and the students are becoming more and more aware of those.

Ned: So we don't shy away from that.

Ned: We talk about all those things.

Ned: The other thing is, if I'm going to want feedback on my research question, of course I can ask my mom or my buddy or my friend or my teacher, but I can also ask the AI.

Ned: And that's really accessible and easy.

Mikal: So what is the difference between using an AI chatbot and using a collaborator or a co-researcher who's working on a similar project?

Ned: Well, for me, they're really similar.

Mikal: Yeah.

Ned: I really feel like the AI is a sort of buddy who has the same interests as me, plus a bunch that I don't have and can bring in some new ideas.

Ned: What is also similar to talking with your buddy, to use that example, is you can go into detail, right?

Ned: So like if I'm going to talk to my running buddy about the running watch or the shoes, we might get into real detail about the types of shoelaces.

Ned: just to name something, the shoelaces are how tight you put your watch or whatever, you know?

Ned: And what I like is that these, the details, these are the sort of the real level at which we are

Ned: working. Okay, we're thinking of big ideas and theories all the time, but we're also working on

Ned: a very detailed level. And in terms of the artistic research, that's where I think the new knowledge

Ned: comes out is in the details. And so when you can talk about exactly where you place your finger on

Ned: the violin to get a certain vibrato, which helps you achieve the interpretation you want to make

Ned: with some piece, this is the kind of detail that is real knowledge, in my opinion. And those are

Ned: the kind of details you might talk to with our teacher or our friend. And also AI, we could put

Ned: all those things together. And it'll come up with some brainstorming on how that could be looked at,

Ned: which is for me really similar to the kind of conversation

Ned: you might have with your teacher or your friend.

Mikal: Yeah.

Mikal: And do you see any difference in the truth level

Mikal: of what's coming out of that conversation?

Ned: Well, see, in this case, I would refer back to this idea

Ned: of the artistic research.

Mikal: Yeah.

Ned: And why we practice artistic research

Ned: is because it's not getting at universal truth,

Ned: but it's trying to divulge personal truth

Ned: how does it work for you what is your artistic practice what are your visions artistic goals

Ned: and how do you achieve them and that's that's not universal everyone's different because it's

Ned: personalized knowledge and just like we might go to a master class of a great artist and they might

Ned: explain the way they do it and people can ask very detailed questions these are the kinds of

Ned: knowledge is that you don't see so often in books, and we can get to that with artistic research.

Ned: So coming back to your question about how truthful is it, in the end, it can create theories based on

Ned: what you've done. It can summarize those things, so reduce those details into theories. But in the

Ned: end you're the one who knows the details and it can propose things or propose ways of looking at

Ned: it but you have to say if it's right or wrong what what it is or what it isn't just like we might with

Ned: the buddy who's talking about shoelaces of our running shoes you might say well you know those

Ned: are too tight or they're not elastic enough for me so i use this one

Mikal: what are we talking about if we're talking about artistic research in your context what are the students doing and maybe you

Mikal: can give an example of a research question that they have?

Ned: Yeah. Artistic research is where you ask a research question which is related to your artistic practice. Research in and through the arts.

Ned: So as opposed to a scientific research where you might want to ask a question related to science

Ned: and like what's the speed of light or something and then you might come to some sort of universal

Ned: truth about what that is. Artistic research is an artistic question related to your own practice.

Ned: In this case, because we are performance majors at the conservatory, we are practicing and

Ned: performing artists. And so you will ask a question like, how can I create my interpretation of

Ned: this piece by Debussy? And then what's essential in that question is, first of all,

Ned: the researcher themselves. How can I create my interpretation? So it comes to the vision

Ned: and the artistic practice of the researcher. And then that whole process is then, like I said

Ned: earlier, is weighed upon what is that vision. And it's about growing that vision. And what's also

Ned: essential in artistic research is the artistic practice itself. So we're talking about recordings,

Ned: videos, annotated scores.

Ned: It's not just texts.

Ned: The research is not taking place on the written page.

Ned: It's taking place in the practice.

Babette: And is it mandatory that the students use AI in class?

Ned: Mandatory? No.

Babette: No.

Ned: No.

Ned: So, of course, anyone who's a good writer,

Ned: they don't need to have AI write their texts.

Ned: And some people philosophically are opposed to it,

Ned: and then they don't want to use it,

Ned: And that's totally fine.

Ned: So I don't force it at all.

Ned: We're just exploring it.

Ned: And we're talking about the issues.

Ned: And, you know, we have this new tool.

Ned: So let's see what it can and what it can't do.

Ned: I think that it really can help for most people in some situation.

Ned: Like either, like I said, brainstorming about methodology.

Ned: But people who don't speak English or Dutch really well,

Ned: they can write in their native language.

Ned: then have that translated. And the translations are pretty good these days. But I think this thing

Ned: about like improving the way you communicate is a pretty big point. I even know for myself,

Ned: I mean, I'm a native English speaker and I come from a family of lawyers who are

Ned: used to writing very clear speech. And, you know, we worked on that and I write some texts,

Ned: which I think is pretty clear. And then I show that to my PhD supervisors and they say, well,

Ned: they tear it apart.

Ned: So despite all of whatever advantages I might have,

Ned: there's still a lot I can improve in the way I communicate my ideas.

Mikal: Aren't you afraid that you're opening Pandora's box?

Mikal: I mean, you're asking your students,

Mikal: put your research question in

Mikal: and ask it to give suggestions on how to answer this.

Mikal: It's not a big step to then say,

Mikal: could you write my dissertation on this question?

Ned: Of course, it'll write a paper for you.

Ned: and it'll have an introduction and a main body and a conclusion

Ned: and those things will all work together.

Ned: But it'll say a bunch of things which you don't want to say

Ned: or you don't mean or that are not true.

Ned: It just doesn't know you.

Ned: So that's why in the context of artistic research,

Ned: I'm just not at all afraid of AI.

Ned: I think it's a really great tool to work it out what you want to say

Ned: and to make your research better.

Ned: It's very much a personal thing,

Ned: which is related to your specific situation.

Ned: And that just takes a lot of effort.

Babette: But you think just every student

Babette: has to experience something with AI?

Ned: Yeah, absolutely.

Ned: I think people should try it out.

Ned: It's like if you were trying to cut things with your hand

Ned: or tear things apart and someone gave you a knife,

Ned: then all of a sudden you could cut things cleanly.

Ned: And okay, a knife can also be really dangerous,

Ned: but so it has to be used responsibly.

Ned: But this tool really enables a lot of things.

Ned: The other thing which I think is that these AI tools

Ned: are going to become ubiquitous in so much of what we do

Ned: and how we do it.

Ned: And not only is it going to be integrated into your iPhone

Ned: and all that stuff, it's just going to be a process.

Ned: And with this integration, I think that it's going to be unavoidable in some way.

Ned: So we better get used to it.

Ned: Is that it?

Ned: Well, yeah, that sort of forced connotation is not the way I would put it.

Ned: I think we should accept it as a real advancement in possibility.

Ned: Yeah.

Ned: That's the way I look at it.

Babette: And when you look at their future work, do you think they need this kind of skills working with AI?

Ned: Yeah, somehow the need feels like too strong of a word, right?

Ned: Because we've proven that it's not needed.

Ned: But you can walk to Paris from here or you can take the train.

Ned: So it can get you there faster and maybe more effectively in some cases.

Ned: Actually, in a lot of cases.

Ned: I think in the case of the research, I think it can help us to make more effective, deeper and better research.

Ned: Another thing, right now we're navigating acknowledging AI in you, right?

Ned: Because you should put in there that you used AI and not only that you used it, but how you used it.

Ned: But I have to say, I think in a few years it's going to be irrelevant.

Ned: I think it's going to be as relevant as citing Microsoft Word for your thesis.

Ned: We don't do that.

Ned: It's just a tool.

Ned: And, of course, it's really fantastic that a lot of people are spending a lot of time and energy thinking about security and ethics and also the energy consumption.

Ned: Because those are really important topics to be talked about.

Ned: And there needs to be controls put into how it's used.

Ned: Absolutely.

Mikal: So do you see your students using it in their research?

Ned: Are you asking me if I catch them using it?

Mikal: Yeah, or if they mention it.

Ned: Yeah.

Ned: Having worked with AI, of course, I start to recognize its way of thinking, its way of expressing ideas, its way of formulating things.

Ned: So now I start to recognize some of those in the research reports.

Ned: And we'll have discussions with the students and say, well, did you use it?

Ned: How did you use it?

Babette: And you ask every student or only the students when you see it in the report?

Ned: Well, like I said, we are now riding the cusp of that wave, right?

Ned: So we're figuring that out.

Ned: So I can't put a rule to it, but certainly we're talking about it

Ned: and we're going to be in some exams the rest of this week.

Ned: And if I think they've been using it and they didn't mention it,

Ned: I will ask, not in an accusatory way,

Ned: but a discussion about how that works in the process of making a research report

Ned: and acknowledgements and search.

Mikal: Yeah, okay.

Mikal: So AI usage is then part of the assessment.

Mikal: How you use AI and in what way?

Ned: Well, I mean, in the end it comes down to,

Ned: are the ideas being communicated from the student researcher in a clear way?

Ned: And you can know when they've had their hand in all the words

Ned: and all the ideas that are in the research report,

Ned: what we want is the student to also develop that kind of sensitivity

Ned: towards real authentic personal expression of ideas.

Ned: And then to be able to weed out and change

Ned: when it's not what they want to say.

Ned: And sometimes you have to think about it.

Ned: Sometimes it'll say something, you think, wow, great.

Ned: And then you sleep on it for a couple of days

Ned: well, you know what, actually that's not what I want to say.

Ned: And then you go back and you revise it.

Ned: And I think that's essentially the process of research anyway.

Ned: And now we just have a tool to generate more options

Ned: and have a little bit more flexibility to work with ideas.

Ned: And so having a sort of malleability to what we're saying

Ned: and how we're saying it in an easy way,

Ned: I think is really effective for communicating artistic research.

Ned: So then we're working on the level of ideas and how we're generating arguments instead of, well, is it this word or that word?

Mikal: So you say for the artistic research, they write a report, but you also allude to that you have a discussion with them.

Ned: Right. Yeah, they present their research.

Mikal: They present their research. So that's both part of what they are being assessed on.

Ned: Yes. It's the live dissemination of the research where the student will talk about what they did and what their insights are.

Ned: They will play some recordings or play their instruments.

Ned: So it's real, the human, a human person representing what they did.

Ned: So that's a really essential moment.

Ned: What I think we're going at with this human versus machine quality, right?

Ned: And I think this is what AI is helping us get to, which is where is the human qualities?

Ned: What does it mean to be human?

Ned: And I think the AI is helping us define that better.

Ned: But I think what it is doing is through reflection, it's helping us learn to make choices.

Ned: And so if I come back to what I was talking about, this process of choosing what your text is going to say and going through multiple iterations, that process is a process of choice.

Mikal: Yeah.

Ned: And that's where it comes down to, is the human choice.

Ned: And the AI, it cannot choose.

Ned: I mean, of course it can, but in terms of your own artistic practice,

Ned: you are making choices in the way you play that piece,

Ned: the way you play your instrument.

Ned: And that's the essence of, well, artistic research.

Ned: I also think that that's the purpose of art, is to show human choice.

Ned: And when we experience art from others, we experience someone else's choice, their taste.

Ned: And that teaches us something about human existence.

Ned: This is for me what the meaning of art is, is to explore the human condition.

Ned: So what does it mean to be a human?

Ned: How can you think about things?

Ned: How can you express yourself?

Ned: How fast can you move your fingers on the violin?

Ned: These kinds of things.

Ned: And if you hear someone do those things, say, wow, it's amazing that someone can do that.

Ned: I could never do that.

Ned: Or I could do that.

Ned: And then we learn something about the way people can take decisions and do things.

Ned: And then that becomes a way of self-reflection about our own identity and who we are.

Ned: And I think that's where we grow from.

Ned: And so with this AI tool, when used properly, it's helping us get to those questions.

Ned: What is it that I want to say?

Ned: What is it that I believe?

Ned: And now I have a tool to be able to work with ideas and explore them until I get to what it is I want to say.

Ned: And then I think that's always going to be evident in the result.

Mikal: My final question for everybody is, do you have any reading or listening tips about AI?

Ned: No, I don't have any literature about AI.

Ned: But I have to say that it's a constant topic of discussion among friends and colleagues and family.

Ned: And even with my 12-year-old son, we're talking about it and exploring its uses.

Ned: So in that respect, we're forming our own opinions based on our experiences with the tool.

Ned: So that's how I'm researching it now.

Mikal: And you're suggesting his father's will be go and try and see what it does?

Ned: Absolutely.

Mikal: Yeah.

Ned: Yeah.

Mikal: All right. Thank you very much.

Babette: Thank you.

Mikal: Thanks for being here and see you around.

Mikal: Thank you also.

Mikal: Yeah.