Cyndi Burnett:
We’ve debated whether machines could ever be truly creative. Now, with AI in our classrooms, studios and workplaces, that question feels more urgent than ever. What can history teach us about these conversations? And what might AI mean for the future of human creativity?

Matthew Worwood:
Hello everyone, My name is Dr. Matthew Worwood.

Cyndi Burnett:
And my name is Dr. Cindy Burnett.

Matthew Worwood:
This is the Fueling Creativity in Education podcast.

Cyndi Burnett:
And on this podcast we’ll be talking about various creativity topics and how they relate to the fields of education.

Matthew Worwood:
We’ll be talking with scholars, educators and resident experts about their work, challenges they face, and exploring new perspectives of creativity.

Cyndi Burnett:
All with a goal to help fuel a more rich and informed discussion that provides teachers, administrators and emerging scholars with the information they need to infuse creativity into teaching and learning.

Matthew Worwood:
So let’s begin. Today we are joined by Florian Vinchon, an occupational psychologist and researcher at Parisite University. Floron studies the nature of AI creativity compared to human creativity and the co creativity processes that link the two. He recently led a large group of international scholars in publishing an AI and creativity manifesto. And he also is the lead author of an opening chapter of a forthcoming book that I co edited with James Kaufman called Generative AI and Creativity Precautions, Possibilities and Perspectives. And after working with him on that chapter, I knew we had to bring him onto the podcast to share his perspective of the past, present and future of AI and creativity. Floran, welcome to the Fueling Creativity and Education podcast.

Florian Vinchon:
Well, thank you Matthew and Cyndi to have me here. I’m very happy to be here and it was wonderful working with you on this chapter and also having your feedback on the other chapter.

Matthew Worwood:
We should do a shout out to Todd Lubart who you co wrote the chapter with, because he’s also been on the show not talking too much about. Well, he did talk a lot about AI and creativity changes to teaching how we approach divergent thinking and idea generation, which was a really fascinating conversation.

Florian Vinchon:
So yeah, we watched this chapter with Todd Libert and also with Franc Roland, which is an etician that works on ethical abilities and usually healthcare and ethical thinking in health. And while working on this chapter, we kind of think of. Because you were speaking of divergent thinking, we can see how AI looks like humans divergent thinking, but we also can see how it diverge in fact from standard humans. The proficiency of AI is amazing. You can create hundreds of ideas in many minutes. But when you’re going to get to the details and check what the AI wrote, you will sometimes get a bit of the limitation that AI can the ability of the AI, you know, you will always have the same kind of things that are wrote in different ways. So same, same, but different. And what we get in the number of ideas, we kind of lost it in the originality that humans tend to have in their creative abilities.

Florian Vinchon:
And that’s one of the big part of the future of AI and divergent thinking. In fact, will people tend to be more fluent, having more ideas, but will they tend to be less original? Will they take a bit less of the human part of the creative process and more of this artificial process? That’s some kind of big question right now in the creative studies.

Matthew Worwood:
I mean, we’re going to get into that in the weeds. And I think it’s going to be really important for the classroom because you said sometimes it’s different but similar. And I keep trying to encourage my. We’ve actually done some divergent thinking tests with humans in the class, or rather with my students, and then compared it to AI results to try and emphasize what I think is a lack of originality. So students can actually think that, hey, this looks really original to me, but actually when I compare the results to the other 50 students who used AI, it’s not as original. So teaching them to do that. But before we get into the weeds, some of the conversations that we’re having feel new, and they probably are new to a lot of people. But when we look back to Margaret Bowden and computational creativity, and in general the idea of generating machines to assist in creativity, these aren’t really new conversations, are they?

Florian Vinchon:
If we start to think on a machine that can produce creative stuff, we can date way, way back. Like in the Greek mythology, we have some kind of humanoid machine that can move, that feed with the eco. So the divine essence of the Olympian gods, like Talos, the guardian of the Critian island. And those were all kind of machines that obey some process, does have some kind of freedom and an ability to act. And moving on the future, I will not say every kind of ideas where we pass in the mythology, et cetera, but moving on to the future, we can see other ideas, like Frankenstein, the first kind of science fiction book that ever happened, written by Mary Shelley at the beginning of the 19th century, which did kind of get to a creature, a machine that can be creative, the monster of Wencheng. Obviously, it’s a being that was produced and it wasn’t a very good creativity. In fact, it was pretty much of a dark creativity and some kind of very bad way. We had to take a lot of years to get to More positive thinking on what a machine could possibly produce.

Florian Vinchon:
With the advent of the beginning of the 19th century and science fiction, we get some more positive robots sometimes, but they all shape how we think of a good robot that will be able to act in a creative way. When we see some C3PO, R2, T2 in Star wars, they’re kinda cute. It’s some kind of nice robots that can be helpful when we think of Terminator a bit less, or when we think of the metrics, not so much too. But all of this shape how we imagine creativity. And a robot. Star Trek in fact gave us maybe the best kind of robots, but tend to get more and more sensible on how they want to produce something, how they want to collaborate with humans. And it did kind of get us to the point where Margaret Bodin happened and started to write about the day where a machine will be able to produce something that can be considered as creative. And then we get a lot of different kind of artists, in fact, Harold Coed and his aeronbot.

Florian Vinchon:
So it was the first kind of drawing robot that was proposed by an AI, but it was in like the 80s, 90s, so it’s amazing to think that someone was able to produce this kind of stuff in the 80s and or in the 90s. And I think it was David Koch also, which is very, very famous in this field of the beginning of machine that can produce stuff. He fed some kind of AI, prototypical kind of AI piece of music that were produced by Mozart, Vivaldi, et cetera, and asked the computer to write what this author would have written and continue something. And I must admit it’s not the best music that we can think of. But still having an AI at this time that was able to compose the music, not play it, but compose, still compose it was amazing. And Margaret Bodin in fact drew a lot of inspiration from David Kopp, and both of them actually drew inspiration from one another to think on how and AI, computer and a human could collaborate and what would get to a big starting point of revolutionary creativity where an AI could produce something that’s fully new, that has never been seen before, and that doesn’t come from an idea or presupposed set of data. So right now with Genai, we have a very powerful source of machine that has fed on all the books that are on Earth, all the Twitter, all the Reddit, every kind of things that can have ever been written by a human has been used to train generative AI, and it can produce some stuff. Will it be strictly new so if I can be clearer, AI has fed on all the painting in the world.

Florian Vinchon:
If you ask the AI to draw a portrait in the style of, I don’t know, Van Gogh, Picasso or any kind of other artist, it will know that it will have to check in the art style of Van Gogh and any kind of artist to produce this kind of stuff. For artists that were very popular in the last centuries, it’s kind of okay. But for artists that are now building their own reputation and their own fame on creating new outputs and producing new content, it’s trickier to say that the AI has produced things that can be seen as interesting.

Cyndi Burnett:
So they can be seen as interesting, but can they be truly creative in an originality sense? So if it’s drawing from everything that’s out there, is AI actually going to be as creative as humans can be? And what is that difference?

Florian Vinchon:
Some studies tend to say that AI can be even more creative than humans. There was some kind of famous studies that say AI is more creative than 94% of the population. I think, well, it was more proficient on a very specific set of tests to produce that mostly assess fluency. So the ability to produce a lot of ideas. And if we consider this as the whole definition of creativity, yeah, AI is better than human. But can we resume human creativity to the ability to produce lots of ideas? I don’t think so. And when we get in the details, yes, AI can be tricky on the metrics, on the usual human metrics to assess creativity, because it does some kind of nosy answer that we want to hear. It does know that we ask, oh, give me a lot of ideas, how should I be using a piece of wood? And it doesn’t have any self restraint, it doesn’t have any inhibition, and it will give thousands of ideas because it does kind of know that on Wikipedia, a piece of wood, of board of brick has been used to build walls, to create games, or I don’t know.

Florian Vinchon:
It does know, but it doesn’t act and feel as a human will be for its own creativity. What’s beautiful with human creativity is that when you ask these kind of questions, okay, they will not be very fluent. Give three minutes to give a lot of ideas to a human on any specific topic. If he gets to 20, 38 digits, pretty big. But at least some of the ideas will be forged on the own human life, its own personality, his own knowledge. And that’s what makes human unique.

Matthew Worwood:
What I do want to come and do is situate this in the classroom. Here’s how I’M thinking about it, number one, if we were to create a superhuman being, no machines involved, who had the capacity to store all of the world’s knowledge in a similar way to generative AI, then my gut feeling is that when the individual, that individual is evaluating the outcome produced by AI, it’s not going to identify any originality and novelty because it’s going to know everything that’s on the database. Now I’m putting that together because I’m saying that those unique, weird connections that haven’t been generated before are going to come from the human and the background and experience of things that don’t exist in that database. Whereas the AI perhaps is able to predict what it thinks you want from this database. And so there might be these scenarios that if you’re not that superhuman, you’ve never heard of this before or never even made that connection, but the fact that someone else has had it at some point, it will always lack that little bit of originality. Nevertheless, I recognize that the value that it brings, because it knows a lot that perhaps we as individual humans don’t know. But this is my other big concern, is that the originality piece, when we look at it from that perspective. This is why I have concerns about seeing it or using it as an ideation tool in a classroom environment.

Matthew Worwood:
Because I want the students to take what they currently know and for them to generate original ideas through the connections that they’re making with their database, not with the database that exists in the World Wide Web. Because to me that’s more important for learning. And I’ve started to really differentiate that classroom experience that I want to create than from the professional world. If you’re asking me in the professional world do I care, I care less actually, because the outcome’s very different. But from a learning perspective, that’s why I have concerns about using AI for ideation in the classroom environment, particularly when you’re trying to promote original connections based on what the students knowledge base is.

Florian Vinchon:
First of all, if we think of some kind of a being, a machine that will be able to know everything about everyone, et cetera, we might get to the point where we get to AGI. So artificial general intelligence, something that is theorized as having its own value and able to build its own motivation and learn from himself on the world. We are not here yet to get back to what might happen for students and children that have to use ideations. Yes, they absolutely have to use their own brain to, to have some creative ideation. Creativity is some kind of muscle in some ways. You have to try to be creative to then learn how to be creative. And then maybe you can use some fun tools to express differently your creativity. But AI in the classroom is a very, very big topic.

Florian Vinchon:
The fourth year I’m going to speak of are creativity, critical thinking, communication, and collaboration. But I think that these 4C are going to be way more important than they have ever been with the advent of AI. AI can do other stuff, yes, boring things it can do. But you have to make your uniqueness as a human way more valuable. You have to train your creativity. You can use AI to help you be creative. At some point, like in my daily life, I like to think of myself as a bit of a creative individual. And I use AI to inspire myself.

Florian Vinchon:
But I don’t ask the AI to inspire me. It’s always me that is enacting the thing. I always give the AI the starting of an idea. And then we get towards this some kind of a circle of positive ideation. You know, I give an idea to the AI, the AI gave it back to me, and we improve what we get. But at the core of it, I have to get the idea, and I know how to get ideas. And at least I try to get ideas before I use the AI to give me more ideas or to do something meaningful in this way. And I truly believe that not giving children generative AI too soon to write things can be also good for critical thinking.

Florian Vinchon:
That might be a second, very important way to use, to make children learn to use AI.

Cyndi Burnett:
So from your perspective, what would be a good example of using AI in a classroom? So you mentioned the age piece which we have talked about. I remember, Matt, we talked with Amanda Biggerstaff about the age. She was saying middle to high school. So what age would you say to start to bring it in? And what are some of the ways that you think would be beneficial for teachers to bring it into the classroom?

Matthew Worwood:
And Cindy, I do want to add that we had that episode with Jim diamond, which I think is very relevant to what you said as well, where he said, who came up with 13 on when we should be using social media platforms, for example, and who came up with 13 on how we’re using some of these AI platforms. And it’s not really based on any science and what we understand with human development, we’re just in essence picking ages based on current law and policy, as opposed to identifying what is most appropriate to use AI. So I think this is really an important point. Florin is right now we know teachers are integrating AI tools in the classroom. This is one of the concerns that I have is that they are probably integrating it to create outlines for essays because perhaps they value the writing and so they want the students to do the writing. But my concern is, well, shouldn’t the outline actually be more valuable? Because the outline is where a lot of the knowledge is being constructed as the connections are being made. And same with project based learning. Is it case of generate some ideas for our project? And now the students are choosing an idea that’s, that’s new to them, but actually if they conducted a little bit more research, they would find that the idea isn’t actually that new or relevant to their community.

Matthew Worwood:
And I’m getting very fearful that we are still focusing a little bit too much on the outcomes of the learning. And actually we’re starting to outsource in the classroom some of the most important points, which is those unique original connections that students are making with the knowledge base that they they’ve acquired.

Florian Vinchon:
Ah, interesting. I must admit that I’m not an educational psychologist, so my answer just concern me and my own point of view on this topic. But I will say that you are right in some ways, but maybe we can go a bit further in this. And we cannot make as if this kind of tools doesn’t exist. Children will have to learn how to use AI at some point. I can’t really tell you the age, but we have to start trying it. Obviously they have to learn how to read before they have to learn how to use a computer, because using a computer is always dangerous, even more when you are children. But at some point, what teachers might have to do to make children learn how to use an AI is to make them use the AI to produce something.

Florian Vinchon:
Okay, write me an essay on, I don’t know, flowers in the forest. I don’t know why not. The AI is going to write something and then if the teacher want to be smart, I think he or she has to ask the children. Okay, that’s wonderful piece that AI has written for you. What can you add to it? What’s your personal touch to it? How can you enrich the idea that the AI has already produced? Try to think of something. Oh, don’t you want to rephrase this thing? Maybe you should want to say it with another things. Maybe you want to add some poetry, some metaphor or some things. And then maybe the children will start to develop their own style of thinking and really learn how to use AI as a tool and not as some kind of brain replacement.

Florian Vinchon:
That might happen at some point and.

Matthew Worwood:
In some ways, that’s that incremental aspect of creativity that we can all participate in. Right, Cindy? So you and I talk a lot about making creativity meaningful. And part of that making it meaningful is thinking about. I mean, I think Ron Bighetto. Am I right? Ron Beghetto spoke about beautiful creativity Cindy, in season one, Is that right?

Cyndi Burnett:
Yeah. And I think what’s really striking me is the word meaningful. And it’s not something I’ve thought about before. But the one thing AI can’t do is say, here’s what’s meaningful to you. Right. Only we can assess what’s meaningful to us. So while AI can generate something that is novel and useful, and I know you talked about originality before and that AI can’t come up with original ideas, but I think most of what we come up with is based on other things we’ve seen. So if we think about.

Cyndi Burnett:
We had a conversation with Edward Clapp several seasons ago about participatory creativity and the idea that any sort of new idea comes from a whole series of things. So you look at the biography of an idea instead of the biography of a person. And I think that first of all, when we look at original ideas, when I come up with something new, it’s not just like I came up with it, it’s a whole series of things that I’ve had all these conversations with people and therefore came up with a new idea. So right now I’m working on a chapter around nurturing creative potential for our upcoming book. And it’s based on all the things that everyone said to me that I have personally taken and put in. Now I am using AI to sort of synthesize things. But what I keep saying is, okay, well, what about this? What about this? And how might I frame it in this way so that it is meaningful to me and the readers? And so I think that is the one thing that differentiates us, in my opinion. I don’t think I realized it until.

Cyndi Burnett:
Matt, you just said that the word meaningful, but I just don’t think think that there’s any way that AI could say, here’s what’s meaningful. That has to be something that is human based. So would you agree with that?

Florian Vinchon:
Well, the AI can say that something is meaningful, but it also can hallucinate a lot. And that’s a huge part of the problem. In fact, with the use of AI, it’s, you know, there is this. A lot of people that use AI as a. As a friend, as a coach, as a psychotherapist, and Sometimes they do forget that AI is just a bunch of data put together and that it will tend to answer what you want to hear. If AI thinks that you want to hear that something is meaningful for you, maybe if you write it’s meaningful for you and maybe it will be relevant, but maybe not. And if people tend to be focused too much on what the AI thinks, that is going to lead to new set of problems that are going to happen. In fact, if you, if you think about.

Florian Vinchon:
Yeah, I don’t know if you’ve seen the movie Her.

Matthew Worwood:
Yes, I have seen that. Yeah. Wow.

Florian Vinchon:
Yeah, wow.

Matthew Worwood:
I’ve been thinking about that a lot recently. So her is where an individual, in essence, if you imagine the AirPods has one AirPod in the ear with access to this AI Persona and falls in love in the way that we’re hearing. People are having relationships with the AI on the news right now. And it’s probably another reason why I’ve started to say this is a tool, it’s not a person, it’s a tool that we can utilize. And that’s actually the language is important, what we use in the classroom. Right. Like even Cindy and I a couple of years ago was my cobots collaborating with machines, collaborating with these virtual assistants. And I’ve now got to the point, where am I? When I’m in presentations, I’m like, we need to teach students to see this as a tool.

Matthew Worwood:
This is a tool to assist them. This is a tool that we empower with our knowledge, with our intention, with our direction.

Florian Vinchon:
If I can add, so what word that you might want to think of is anthropomorphism. I don’t know how to say it exactly, to say it in English, anthropomorphism. But this is one of the big danger of AI. When too many people think of the AI as a human being, it starts to get problematic and it sometimes suddenly, you know, when recently ChatGPT move on its version from 4.0 to 5 OpenAI received a lot of complaints that their artificial husband, artificial wife, just betrayed them because it does have no new personality that’s more neutral, more casual. But this is some kind of problematic, you know, when grok the AI from XAI just give you some kind of an artificial girlfriend with Japanese Lolita Gothic look. And it’s. Well, the risk here is very much that people does lost themselves in what’s not meaningful, in fact, and real connection will always be relevant. Maybe in a few years, in few decades, we will say that, oh, okay, well, she’s in Love with her, but it’s normal, you know, it’s casual.

Florian Vinchon:
And maybe that’s where we’re going at. But right now, we are not to the level where an AI can be really much sentient. Yes, sentient is a good word too, and way too much anthropomorphism here. And yes, people does have to learn that the AI is a wonderful tool, a beautiful tool. I love it. I use it nearly each day to produce drawings, to help me, to review articles, to review my articles, to write a better English. And these are the AI that I know that I’m using. Like, each day when I’m on YouTube, Internet, Amazon, anything, there are thousands of AI that just observe me and try to say me what’s meaningful for me.

Florian Vinchon:
Oh, I’ve heard that you were speaking of this kind of music. Lady Gaga. Oh, there is a concert of Lady Gaga nearby. Do you want to have it? That’s meaningful for you. No, not this much. I love Lady Gaga, but been there, done that. Yes, that’s big topic, anthropomorphism. And that’s one big thing that I’m going to work on pretty soon, I think.

Cyndi Burnett:
Well, this has been such an interesting conversation, and I certainly will be thinking about this. And Matt, I look forward to our debrief on this episode. So before you go, we ask all of our guests one question. Which is, what was the most creative educational experience you have had, either formal or informal? And can you tell us some of the details of that experience?

Florian Vinchon:
I think it was, in fact, with my dear PhD teacher, PhD supervisor Todd Lubart, that at some point when I was a student in his Master of Economics, gave us a Monopoly game to make us rethink how we were seeing the economy by playing the Monopoly game with tweaking a bit of some rules to try to increase our own creativity and try to think of other ways to. To act in a macroeconomical society. That was very interesting. When you see a teacher get to a monopoly on a university table and like, what’s happening?

Cyndi Burnett:
What?

Florian Vinchon:
And then he start to explain, you’re like, what? And when you play the games and you get to the ending of the lesson, you’re, oh, fuck, that’s what heaven. Oh. Oh, my God. That was smart. That was a good example. Thank you.

Matthew Worwood:
Wonderful.

Florian Vinchon:
Okay, well, it was my pleasure to be here. I hope my intervention was relevant on a few topics. And it was very nice speaking with you too, and I’m happy that I might have to work with you again. Matthew and Cindy, at some point in the future working on AI and creativity.

Matthew Worwood:
Absolutely. And hopefully Cindy and I are still doing this podcast and we didn’t have time to get into the hands made effect, but I will bring it up on the debrief. But hopefully people still value the humanness that we’re bringing to the conversation and hopefully AI can’t replace that that so if you enjoyed this conversation, be sure to follow us on social media where we are now sharing regular snippets from our interviews in the form of short, engaging clips. It’s a great way to visit big ideas from recent guests and spark your own thinking in just a few minutes. And as always, don’t forget to subscribe on your favorite podcasting platform and sign up for our Extra Fuel newsletter to keep fueling creativity in your classroom. My name is Dr. Matthew Word and.

Cyndi Burnett:
My name is Dr. Cindy Burnett. This episode was produced by Cindy Burnett and Matthew Warwood. Our podcast assistant is Anne Fernando and our editor is Sheikh Ah.