Speaking of ... College of Charleston

Human Creativity in the Age of AI

• Season 3 • Episode 19

Share episode

Send us a text

What happens to human creativity when machines can write, paint, compose and even evoke emotion?On this episode of Speaking of … College of Charleston, we explore the intersection of human creativity and artificial intelligence with Anurag Tiwari, a computer science instructor in the College of Charleston Computing in the Arts program whose background is in computer vision, AI and digital animation. 

Tiwari’s journey, from developing video analytics for surveillance to creating interactive installations at places like the Kennedy Space Center, reflects a career at the intersection of technology and art. Now teaching at the College, he helps students navigate a world where AI tools are reshaping creative industries.

Resources from this episode:


Welcome to speaking of College of Charleston. I'm your host, Amy Stockwell, and today we're exploring what happens to cre, human creativity in the age of artificial intelligence and large language models. From writing and design to animation and programming. AI tools like Chat, GPT. Image generators are transforming the creative process, but are they enhancing human creativity or replacing it?

Joining me to answer that question among others is honor Ari, a computer science instructor at the College of Charleston's Computing in the arts program, whose background is in computer vision, AI, and digital animation. Together we'll unpack how the rise of AI and LLMs. Is challenging traditional ideas of mastery, sparking debate across industry and pushing us to rethink what it means to be truly creative in a world where [00:01:00] machines can generate art code and even emotion.

So thank you for coming into the studio today, at the end of a long day of teaching. Um, happy to be here. We appreciate you being here. Your, um, your work bridges, technology and creativity. Can you tell us a bit about your journey into computer science and how you became involved in the computing and the arts program here at the college?

Right. Um, my computer science journey began on a very far side of the, uh, far technical side of of development where I started off with computer vision and developing video analytics for surveillance purposes, like counting people in a video or. Checking for traffic violations, uh, things like that. Huh.

And that eventually brought me into overlaying data on top of videos. So, which is augmented reality, right? Your ar vr stuff. Um, [00:02:00] which translated into game development 'cause ar vr mostly, uh, is seen in the world, in, in forms of interactive games that people play on their phones. Um, at that time I was, uh.

Working towards my masters at University of Southern California in California. Um, and that kind of pushed me towards exploring the creative side of computing more and more, um, with jobs and internships, kind of, um, taking up all of my schedule. I eventually. Burnt out, um, as most people do. Mm-hmm. Um, and, and at that point I was looking for a change in schedule, if you will, and that brought me to Charleston, uh, where a local company called Moondog Animation Studio, uh, wanted to develop interactive games and installations in places like the St.

Louis Aquarium, Kennedy Space Center. [00:03:00] So I, I jumped, uh, as quickly as I could on that wagon. And since 2018 I've been here in Charleston. But then the next problem, the pandemic happened. Right. So the dreaded pandemic. Yes. Um, and at that time, the then head of the company, and I think he's still the head of the company, Ben Davis, got in touch with me to find out if I could teach courses.

Um, at the computer science department teaching Java and Python. So that kind of, that is kind of my full journey. Yeah, like a few sentences. Um, that's how I got to, um, being an instructor at the computer science department. And then eventually I discovered the CA program, which like is what precisely what I was looking for in being a developer where I could combine my.

Uh, I mean, people say it's combining my artistic [00:04:00] skills, which I believe I have none of. Mm-hmm. But, uh, yeah, no, the CDA program at the computer science department essentially combines the computing aspect of development with concentration in areas like visual arts, music theater. Digital arts and now game development, right?

So unlike traditional cs, uh, paths that you might have in a bachelor's degree program, um, the CDA program mainly focuses on equipping students with, uh, technical fluency with using computers nowadays. And the best way of getting about developing. A tool that might assist in them achieving their creative goals, whether it be theater, digital media, or games, uh, video games.

Um, so yeah, that, that's kind of like the culmination of the best [00:05:00] parts of what it means to be human, rigorous technical work, and then combine that to achieve kind of like a creative. Goal or ambition you might have, right. That you want to see through. Right. The CA program basically enables students to have the technical know-how to achieve those creative artistic goals.

And it sounds like a wide variety of artistic. Yes, yes. So, like you say, music, music, music, um, Dr. Bill Manus, uh, is the head of the CDA program. He personally, I think, teaches three or four of the music and computing courses that are in the CA program for the music, uh, major. Hmm. So, yeah, no music, digital media, theater, um, game development.

If you're someone like me, um, any creative outlet that students. Kind of are looking for, which must make for an interesting group of [00:06:00] students because Oh, yes. They're not all coming from the same technical background. Yeah, yeah, yeah. No. So it is a challenge as an instructor, which is why I think you have such amazing and such knowledgeable, uh, professors like Dr.

Manus, um, teaching some of the most important courses that the students will do. Throughout their college life. Mm-hmm. I believe because you're not just like doing memorization and regurgitating it onto an exam paper. You are working on actual projects from possibly clients that are local to Charleston.

Hmm. That's, that's a, that's a great segue 'cause um, I, well I wanted you to talk about some specific projects just to give people a better idea to be able to visualize. Can you talk about some specific projects that some of the students have done on campus? So, uh, last I checked the projects list. Um, there were two students that, [00:07:00] um, went to GDC Po, I think last year.

Mm-hmm. On the game developer conference, so. They worked on a project that they then showcased, I believe, with Professor Sarah Schumann. Mm. We've had her on a previous episode, so I can direct people that, yeah. Early on in our, in our podcast journey, she, she came on. Right. Um, yeah, no, she has been instrumental in, in teaching or helping students achieve animation and game related.

Artistic goals along which, along with teaching them the technical fluency required for getting to those goals. Mm-hmm. Um, so you have game projects that have been worked on for GDC. Um, not too knowledgeable on the music side of things, but some of the theater, um, related projects with video recording, audio recording mm-hmm.

And just theater presentation in terms of interactive digital media [00:08:00] was presented in. I think it's called, uh, the Bele Studio. Oh, right, right. Uh, the the local guy. Yeah. Yeah. He's a big artist. Mm-hmm. Um, that I, I don't know whether it opened recently, but he has a big studio in North Charleston. Mm-hmm.

Um, I think one of the projects was featured there, um, as part of his. Uh, cryptocurrency and, uh, crypto art. You Oh, very cool. If so, there are projects going on all the time at the computer science department and most of the time it's going to be something local that becomes something, um, nationwide.

Mm-hmm. Like you have your game development conference where you have people coming all the way from the other side of the world to present their. Um, or pitch their game and our students are getting to participate in that. Yes. Wow. So, yeah, the CDA program is, I feel the perfect, and it's one of the first in the country to push students more towards [00:09:00] projects.

Mm-hmm. And taking a project from start to finish rather than just learn about theory, theory, theory, theory. Right. Like actual, literal. Hands-on learning project. Yes, yes. Yeah. So, which is stuff they can put in their portfolio and take with them as they go out into the Oh yes. Working world. No, as far as I understand, the job market nowadays, especially with AI tools now being used rampantly in every industry that we look at, um, employers and companies are no longer satisfied with just, you know.

Oh, you have a master's degree. You have a bachelor's degree. Students now kind of get chosen better into employment opportunities if they have more projects. Huh? No longer are employers. I mean, sure. Your grades are important. It basically tells your employer, Hey, you are able to stick to a. [00:10:00] Schedule for six months, right?

Seven months, two years. But it's the projects that you have at the bottom of the resume that shows them, hey, the student was able to participate and take part in a project from start to finish. Mm-hmm. And see it through presented to the world. So they are used to the rigor of debugging or solving problems on the fly in a very restricted schedule.

Yeah. Because these classes are. Barely six months in length. Yeah. So if students can finish a game, for example, in that time, that is magnificent. Yeah. Right. So I think these projects being the core part of the syllabus at the CA, uh, program courses, benefits the students. Um, and how they get hired nowadays.

That makes a lot of sense. Yeah. That seems entirely logical. And, um. Because then the employers have a great example to look at to see what the students' capabilities are, [00:11:00] right? Yeah. So just, um, to move a little bit away from teaching and students into the larger world of AI and, and creativity. You and I talked previously about films specifically, um, the film's, hereditary and Avengers Endgame.

As examples of human storytelling with an emotional impact, um, that technology can't achieve. Right. Can you talk about this a little bit more and explain why in, in filming, despite all the advances of ai, um, that AI can't replicate, like that kind of emotional impact? Right. No, I, I love talking about her and, uh, Avengers end game as like the two flip sides of.

One, just pure emotion and catharsis in Ry. Whereas other side you have the extravagant CG that is showcased in Avengers end game. And anytime like [00:12:00] convers me and my, uh, good friend Buen was a filmmaker in India, um, kind of tried to argue, hey. Films no longer like, are just pure stories. They have become this unnecessary, flamboyant portrayal of technology nowadays.

Um, that that's his stance on things, being a filmmaker and seeing the decline of the quality of films. Uh, but I, I kind of argue that even though Avengers end game is this extravagant, flamboyant, CG. Artistic masterpiece. Right? Um, it's the human emotions of the characters that the viewers have stuck with for over, what, six years since the first Avengers movie came out.

It, it's what makes those last 15 minutes of the Avengers end game movie [00:13:00] so painful? Mm. It's the human emotions that. Makes people go watch the movie again and again and again. Mm-hmm. It's what puts people in cinema halls or Netflix, whatever, um, platform they watch these movies on. So you have a movie like that compared to something like Hereditary where there's no fancy cg, it's just terrifying human grief and.

Breakdown of the traditional family structure, if you will. Mm-hmm. Right. Um, and just the pure terror of the things that unfold in the story. I don't want to give too much away for people. Yeah. Don't give the ending away. Yeah. Right. Um, just the terrifying way the things unfold in the story, the way the actors portray those characters, those little idiosyncrasies [00:14:00] that.

Raise that emotion of terror in the viewers is what makes that movie such an amazing piece of art. Mm-hmm. Right. It, it's nothing similar to Avengers End game yet. It is on par in terms of bringing about the emotions that we kind of associate with the joy of watching films. Right, right. So, um, yeah, no, you have on one side.

AI and technological masterpiece on the other end. No CG, if you will. I think few cable work and basic set stuff. Mm-hmm. Um, in Heritage, both of them achieving the same kind of goal and those emotions that it pulls out of the viewers, so exactly. Um, while you can use these AI tools and these [00:15:00] technological masterpieces nowadays, it's ultimately going to boil down to lived human experiences that allow these actors to portray these characters and invoke those emotions in the viewers.

That, that's my take on it, at least. Right, because you and I both want to argue for the value of human input right. In, in the rise of, of ai. Right. But people who are creative, we still wanna feel like we have something to offer. Right. I absolutely, I, I agree with your argument a hundred percent. Is there, is there an example of an a of an incredible AI film like Avengers?

But that doesn't have the human emotion, the human impact. So it, I mean, where do I start? Really? Like, there's a lot of 'em. I mean the, I don't wanna bash on the latest Jurassic Park too much. Oh, yeah, yeah. Even with the [00:16:00] amount of money, the, the stellar cast that they have. Yeah. The story doesn't push a good, um.

I wouldn't say a message, but there's nothing for the movie to fall back on. For people that have watched too many of these monsters and CG animation all day, at least to my eyes, it's just tiring. It's same old, same old, right? If the story is, is not good enough for me to not want to change the channel, I'll just stop playing it on my screen.

Why would I spend money going to the theater to watch it again? The animation, the CG is absolutely beautiful. And then nothing I say will take away right from the actual production of, of that movie or the people that worked so hard towards it. But the [00:17:00] final story doesn't resonate with people who have watched million different action movies with monsters in them, right?

And for right now, ai. Can't do the storytelling part of it. Right. I mean, that's what I want us to believe is that not yet. I, I mean I can go on this topic for a while. Mm-hmm. But if I try to, um, synthesize it into a few words, at least tell me and my friend like to talk about it, is you can't teach these tools.

Two very important things that make humans, humans. One is pain. Struggle, pain, whatever that we wish to call it. And the other being, um, ambition. Hmm, right. Many great Greek Roman philosophers have this usual cycle that you have a, a period of peace, you have a period of exploration, you have a period of [00:18:00] innovation, then you have jealousy, then you have war, and then the cycle again, repeats to peace.

We are right now in the peaceful slash innovation period of our like time. Um, and AI has now come to help us. We have to rise and move past, uh, using AI just for our daily work and improve our own. Understanding of the world, if you will. Mm-hmm. To not keep falling back into the cycle. Oh, okay. Everyone has AI now and let's just start making better and better AI doing the same exact thing.

Then we stick the cycle, then it becomes about jealousy. We go back into war, and then usually if the war cannot turn into peace, usually leads to extinction of a civilization destruction. Right. You have the Roman civilization, Egyptian civilization. All of these falling, right? So I digress. With ai, our goal [00:19:00] shouldn't be to just offload work onto these tools, right?

But rise above it if you will. Um, yes, definitely use the AI tools that you have available, but they will never be able to replicate your own creative input that is purely human because the simplest reason you cannot teach ai. Human lived experiences, right? It, it becomes incredibly difficult, if not impossible for a tool to extrapolate on these issues.

And we have seen Chad GPT and um, clawed and Gemini kind of turn to sarcasm and try to be funny if you try to give it a very cathartic or very tragic incident and ask you to make a decision on it. It. Mm-hmm. There are like experiments that researchers suggest, like the Mary's room experiment or [00:20:00] the Chinese room experiment, where they argue what does it truly mean to understand something right now, these L LMS can tell me about that experiment.

Uh, so the, let's talk about the Chinese room experiment. The idea is you place someone inside a room and they have a dictionary of every. Uh, Mandarin character on like a book in front of them and the English translation. And for Mandarin being the complex language, it is, you have a combination of the letters that form words and phrases.

Mm-hmm. So this person has access to a massive, massive dictionary that's indexed. The experiment is a native Mandarin speaker and a native Mandarin writer will write a Mandarin message on a piece of paper and drop it into the room. The person then waits for an answer from the room. The person or the technician inside the room does not, is not a native Mandarin speaker, does not understand Mandarin [00:21:00] whatsoever.

But he can go through the dictionary, um, and interpret what it means in English, and then write the answer back in Mandarin by converting the answer from English to Mandarin. Mm-hmm. And this experiment has been done all over the world. For decades now, and every time the technician inside the room is able to fool at least 92% of the time last I checked, um, the technician inside the room is able to fool the person outside into thinking that the technician understands, uh, or is a native Mandarin speaker.

So you have this argument that even though he's able to. Pass the touring test, which is usually the test given to machines to find out whether they can fool a human into thinking they're human. Are they really understanding the data that's been given to them? These, these large language models, they have access to the history [00:22:00] of human life.

They have access to history of English every other language, but do they really understand it or are they just doing a formulaic. Processing, like the technician in the Chinese room experiment is doing. Right? Right. He's so that, that's the experiment. It's an argument of do you really understand what's happening versus do you comprehend what's happening?

Understanding is a very human-centric, um, phrases, if you will, uh, don't know how other to put it, but it, it boils down to things being. Amorous that's very nature oriented. Or I morphic that's very human oriented. All these things, emotion, understanding of a topic are very, I morphous things. You can't really offload these things onto software or hardware or ai, whatever we have nowadays, regardless of how [00:23:00] powerful they get, you can't achieve proper consciousness.

Thank God. I say thank God. Yeah, that's that's fascinating. And it's also really encouraging for me, um, which it kind of takes us to what we wanted to talk about also is I like that, I like that way of thinking. What you're essentially saying as someone who is teaching the next generation of um, students, is that instead of dumbing us down, that AI.

Is essentially raising the bar. Right. So, so, but it's up to us to, to rise to the communication. Yeah. So how do you, how do you advise or mentor your students? their journey through the college and going into the working world. I have a blueprint, um, and I've been trying to figure this out since the first semester that I joined College of Charleston.

And a few of the students asked me the same question. I have a blueprint for what I have been [00:24:00] doing and trying to figure out the space. Um, my first suggestion is build your own thing. Don't just. Rely on the coursework that you're doing in college. Yes. It can be challenging to take time out of coursework to do your own thing, but we all know as students in an AM bachelor's degree program, we have time.

We just don't manage it well. Right, right. So, um, my first suggestion would be build your own thing. Build independent projects. Don't even need to finish a project just. Conceptualize a project that you're passionate about and see, reach out to, like instructors or seniors to see how well you can realize those concepts using the courses that you might be taking.

So build your own thing as you plan your path through. Uh, something like the CDA program degrees. Because they have a lot of [00:25:00] electives based on what you want to do or where you want to see, uh, your career path, be at the end of the degree program, work towards that from day one. So if you wanna make a game, conceptualize it, talk to your professors, see what courses you should take.

You want to maybe make an animated short for YouTube again. Put it down. Go talk to your advisor, see what courses they recommend, see what you need to learn, what tools you need to know, and then just go and try to achieve it. Even just having that as part of your resume tells your employers that, Hey, this person again, as I mentioned, is able to plan and see something from start to finish.

Maybe the project itself isn't finished yet, but you have conceptualized and made a framework that you know will probably work. [00:26:00] Mm-hmm. So if they see that you've taken the initiative and you're passion, the curiosity, the innovation. Yeah. Right. If they see that you are passionate about something, that's how you get hired at these companies.

Mm-hmm. Because, yeah, old school just interviews, technical interviews and resumes are not cutting it, as I said. Because of how innovative humans have been, AI has been becoming so popular. And with that, the entry level of getting into jobs, that bar has kind of been raised, as you mentioned. Mm-hmm. So we have to do things beyond just the regular coursework and just the shiny new degree you get mm-hmm.

At the end of the four or five years. Right. So that, that's my main kind of. Directive to the students. Build your own thing. Make sure you have enough passion to possibly see it through by the time you graduate. Um, and [00:27:00] portfolio over resume every single day of the week. I love that. Um, build your portfolio more than your resume.

Resume is just going to show your employer, okay? He has a degree. He probably worked on this project. A portfolio will. Actually detail the struggles you had, how you decided to solve them. These highlight your problem solving skills. Mm-hmm. Which is what the employer is looking for, whether you are a right fit in solving or helping the company solve the problems that they might encounter.

Mm-hmm. So, p and I think being able to articulate too, like you say, like not every project you're gonna have time to finish it. Right. Or even not every project is gonna. Meet the goals that you set. Right. But that, that you were able to, to, for students to be able to articulate why it worked, why it didn't work, why it was finished on time, why it wasn't finished on time.

That is crucial. Yeah. That is crucial. And employers want to see more of that. Mm-hmm. [00:28:00] Um, with AI tools now just being used for resume parsing, if you have a portfolio that's attached to your application, um. The employer will see it. Yeah. Um, at least they don't use AI tools to do parsing of websites or, um, an AR portfolio that you might have as a webpage on your phone.

Um. So yeah. Portfolio over resume and build your own thing that you're passionate about are the two main directives I would give. Yeah. To any student, um, walking into a computer science or computing in the arts degree program. Yeah. That's, that's wonderful advice. So hopefully the students are listening so that they can hear that.

Great advice. We'll make sure it, it worked for me. Yeah. It worked for me. Right. Moving from California to here. You the proof. Yeah. Yeah. I, I, I literally burnt out. Uh, in the West Coast. Mm-hmm. And I just up and left [00:29:00] and because I had a detailed in enough portfolio. Companies like Moon Dog noticed. Yeah.

They're like, Hey, can work for us. Right. And I see it as just a shift in your journey. It wasn't like an end point necessarily. It was, it just changed the trajectory in what I feel to be a much more happier life. Exactly. Exactly. And we're glad that you're here. So our students are fortunate that, that you ended up in South Carolina all the way.

Um, so what, just to, to leave people with some hope about. Because I think, I do think there's a lot of doom and gloom reporting out there about AI taking over the music industry and the visual arts industry and all of those things. Um, what, let's end on something hopeful. What gives you hope about the future of human creativity in as AI continues to evolve?

This is a difficult one. Yeah. No, come on. Don't say that one. 'cause I don't have notes for this. Yeah. Uh, two is. Uh, [00:30:00] we are always going, it, it's very human to be afraid of things we don't really understand. And, and let me be honest, even the smartest AI researchers don't really understand what's going on internally in the, in these AI models.

Um, that being said, we've always been afraid of things we don't understand whether it's the invention of the printing press that scared. Traditional artists that hand wrote dictionaries and journals for kings. Great example. Um, whether it's the invention of the printer that that took away from the traditional printing press, internet took away from your traditional printers to now things being digital.

It is very human to be afraid, but, Again, the reason why I feel we haven't hit that, at least I don't believe we will hit the point where we will have actual true simulated human [00:31:00] consciousness in AI models is again, because you cannot replicate human lived experiences. Um, and what gives me hope in, in the kind of market we see nowadays for new, fresh graduates.

Going into the world with these AI tools being available is that there's still value for genuine human creativity. And then, uh, innovation. And we see this everywhere. Whether it's like these small boutique furniture stores. We have fur, we've had furniture for eons. Now what seems like, but there's still, and these are Ikea furniture that are our factory made and whatnot.

There's still place for those small, little. Handcrafted chairs that people like, there's still meaning and value to handwritten letters that people write each other to friends, even though we have Facebook, Instagram, whatnot, email. Mm-hmm. So [00:32:00] my, my, um, money is still on human lived experiences and what it truly defines us as humans, um, in that.

Regardless of how automated the world becomes, they'll still be place and value. So you can most likely earn money and probably earn money doing things that are innately human centric. That is creativity. Making a bold that looks like a dinosaur head. No AI model is going to come up with something like that.

Mm-hmm. You have a person that is a carpenter that really likes dinosaurs, so they made a bowl that looks like a dinosaur's head. These little things, these little sparks of creativity will persist throughout. Um, humankind struggle against automation and we have seen it's been happening since 18. [00:33:00] The 18 hundreds when the printing press came out, right.

So you're right. Uh, we will struggle and we have to rise to the occasion. 'cause anytime these technologies come out, the bar gets raised a little bit and a little bit and a little bit. So we have to rise to that occasion. But still, uh, human lived experiences and what it truly means to be human cannot be replicated by these machines.

And there'll always be value in. Those little nuances that make us human. That's beautiful. I mean, we should end right there and not do it another words, because that's beautiful though. Your words are beautiful. Very inspiring. So, um, I will add to our listeners, I'll add show notes about all of everything that we've talked about, but, but just so, um, listeners know, where can they find more of your work?

Where, where can they find more of your work? So, as I said, I, I kind of. Ran away from the West coast. Mm-hmm. So I'm not gonna mention any of that. Mm-hmm. But I did work at a few game studios, um, [00:34:00] and that, that made games that people might have played. But, um, I'll just highlight things I've done in the East Coast.

So, um, if they walk into St. Louis Aquarium or, uh, USS Yorktown at Patriots Point mm-hmm. Or Kennedy Space Center, they might find some of the things I have worked on as interactive. Um. Presentations or games. Um, I am also currently working with a few students at the computer science department on a few AI based research projects that focus on image and video coloring.

So if you have black and white videos and black and white hand drawn images, how can we. Craft tools that understand what's in the images and then auto color them based on context. The goal is to be able to take old cartoons and auto color them and eventually possibly auto generate audio for [00:35:00] silent movies.

Oh, very cool. Um, that's a great project. So, yeah, they can probably find the things I'm working on, um, if they come by the computer science department. Yeah. Which is in, uh, Harbor, walk East, uh, or they can find it on my link of the college website. I'll put that in the show notes too. Best way to reach me.

Yes. Yes. Well, thank you for coming into the studio and thank you to everybody for listening to this episode of Speaking of College of Charleston for show notes and more episodes. Visit the College of Charleston's official news site, the College today@today.charleston.edu. You can find more episodes on all major podcast platforms.

This episode was produced by Amy Stockwell with recording and sound engineering by Jesse Ks from the Division of Information Technology. Thanks and have a great day.