The weird, wonderful AI art of Niceaunties

A conversation about the "Auntieverse"

My guest this episode was Niceaunties, the pseudonym of a Singaporean-based AI artist that uses her cultural heritage and childhood experiences growing up with 11 aunties, plus parents and grandparents, as inspiration for an imaged reality she created called the Auntieverse, short for Auntie Universe. (Find her on Instagram or Twitter/X).

I spoke with nice aunties while she was exhibiting her work at the Zona Maco festival in Mexico City in partnership with the gallery Patricia Conde. This was part of a group show sponsored by Fellowship AI, a collective that helps support AI artists. She also recently completed an online solo show with the Fellowship that included more than 1,000 still images of her own work that she curated, many of them selling through Fellowship’s online platform Daily.xyz. We spoke about her inspiration, the AI tools she uses and how her artistic process has changed over time, and about the criticism of AI art from traditional artists. I had a great time speaking with her and I think you’ll enjoy our conversation.

Nice aunties. Welcome to the podcast. How are you doing?

Hey, James, thank you for having me. I'm doing great.

Yeah, thanks for being on. I know you're in Mexico City right now for a solo show, I believe, of your AI art. Do you want to talk a little bit about that and how it's going?

It's actually a group show with Fellowship and the gallery Patricia Conde, which is a local gallery. So there's a partnership with Fellowship, and it's curated by Alejandro Cartagena. So basically, eight Fellowship artists have got our work in prints at Zona Maco, which is the largest art fair in Latin America.

Wow, that's incredible. Talk a little bit about the Fellowship. For those who might not be familiar with it.

Fellowship is a online gallery focused on NFTs. They started out as collectors, actually collectors of photography. So they wanted to collect great photography in the form of NFTs. And then they started to gather and curate the best artists for post photography on Web3, and then gradually moved to AI videos. So I am part of their AI videos program called Daily.xyz. But recently I'm starting to go into physical prints as well with them, which is what we are exhibiting in Mexico City right now.

How many artists are associated with fellowship now, roughly?

Well, it's growing, so now it's between 60 to 70 artists.

Okay. And it's only video. There's no still images associated with the Fellowship program.

They have done two previous drops called Post Perspectives, which is about still images. So if you check out Fellowship.xyz, you can see those past exhibitions and drops. So they were mainly doing still images, actually, until August last year, when they started their video program, because AI video is very new in the market right now.

The work you're exhibiting in Mexico City are still images. But as a digital artist, a lot of your work so far has been short videos. What are your thoughts on the artistic process for those two different mediums?

That's a great question. I think it's about the length of time you take to perceive the artwork, because my process with making videos is that I always start with the image. So I make an image, and then I animate it, and then a sequence of this footage comes together to become a video. So it's about a big narrative. So each animated image, you only perceive it for a second or 2 seconds, and then it's sort of strung together to form a big story. So the experience is very different. And then there's also sound editing and voiceovers and music. And as for images, people can take as long as they want to study the image, right. So you sort of have to fit the entire narrative into one image. So therefore the details, the layering, the composition, there's a lot more care and attention paid to it, I would say.

And most of your work, at least as I've seen it online previously, has been more around the video than the images. I think you've done a few images here and there, but it sounds like this most recent show was more than 100 still images—

A thousand.

Oh my gosh, 1,000 images. So that's quite the transition. How long did it take you, by the way, to create 1,000 images?

Well, I know you said that you've seen my AI video works, but actually when I started AI, it was in January last year. Between January to July, all I did was make images. And by then I would have done about 40,000 images.

Oh my lord. Okay.

And then from July, it's like the program Runway ML released their text-to-video and also image-to-video programs. So I started experimenting with this new release. That's how I began my AI video creations. So in the meanwhile, I'm still making images because you have to make the image before you animate it, that's my process anyway. So my project — the 1,000 images drop a few days ago — was basically my accumulation of what I've done for the whole year because it's a world building project. We actually called it the Auntieverse, like the Aunty Universe. So it's everything that constitute this world. The cities, the social life, the fashion, the food, the beauty, everything.

That's incredible. So these thousand images were built up over time as you've been working in the background while releasing videos. You didn't create 1,000 images just for this show.

No, but many of them I did develop it. So there are many ongoing themes that I've been exploring. So I have to remake an image, like, three to five times as I keep regenerating on the same concept. Because AI technology has also improved from early last year to now. So I also wanted to improve on the quality of the images. So yes, many of the 1,000 images were recreated, but some of them were actually from eons ago, from January or April last year.

Most of your work is pretty vibrant with color, but I did see a few black and white images that you had posted online. The ones I saw in particular were of the aunties with the ginormous hair, which I loved. How do you think about your creative process in terms of making an image or a video that's vibrant with color versus something that's black and white and a little bit more muted in color, but still tells a story.

So my inspiration comes from everywhere. About the black and white images, last year, I picked up a book. It's called “Natural Enemies of Books: A Messy History of Women in Printing and Typography.” And I was instantly drawn to this book, and I read it. And the disturbing thing is, it's about women in the 1920s in the printmaking business. And there was a book historian who classified women along with the other enemies of books like: damp, dust, dirt, bookworms, callous readers, borrowers, book stealers, book ghouls, et cetera.

So I was so emotionally affected by what I read that I used AI to make a series of images of women with ginormous hair in these black and white 1920s scenarios in the printmaking business, just going about doing their book binding and work in the publishing houses. So big hair because of big personalities and presence when they're actually doing all the work, but not credited, but instead sort of insulted by book historians. That's how it came about, these black and white images, and they happened to form one chapter of my 1,000 images show. So that's the background.

That’s fascinating. As you described your thought process and the themes of that chapter of your work, it really reminded me of some of the other themes you've woven throughout the Auntieverse. Tell us about the Auntieverse project and how it came about and what the goal of the project is.

So the Auntieverse is a world-building art project about auntie culture. So auntie culture is a prevalent phenomena in southeast Asia and generally in asian communities about a set of behaviors that will cause you to be labeled an “auntie.” So “auntie” does not necessarily mean blood relative. It could be an older woman or anybody, men, women, or everybody in between who exhibits auntie behavior, which generally means being old fashioned, giving you unsolicited comments, very naggy and generally deemed negative. Not in a good light. Yeah. So the Auntieverse is my attempt to portray these aunties in a very endearing, lighthearted way. They give unwanted comments because they care, right. And then they call you fat, but at the same time, they give you lots of food to feed you out of love and care. So I wanted to show that side of the story.

Yeah. And another theme I've heard you talk about in other interviews is that your work is a kind of commentary on ingrained repression. So aunties can feel repressed by society and it stifles their dreams. But your work imagines a kind of whimsical, vibrant world where aunties can fulfill their dreams and their desires and just be weird and wild.

Yeah. I mean, it's not just society. It's also family and culture, like generations of Chinese upbringing and then grandparents passing it down to parents and then to my aunties.

So I remember from personal experience, one auntie wanting to go to church and having western religion is deemed to be a betrayal to Chinese families. It seems to be a common sentiment at that period in time. And I remember my auntie wanting to go so badly she had a mental breakdown. That's one example. And then I have another auntie who wanted to go to Japan and spend a lot of time there, but couldn't because of what her parents thought and also a lot of opinions from the family.

And then my grandmother, my maternal grandmother, she was bedridden for 20 years. She had dementia. Prior to that, she spent her entire life looking after her eight children. She had no sort of career of her own. It was all about other people and not herself. I just felt like there's so much more to what they wanted to do and what they could have been. And through this project, I therefore imagine an alternate reality for people to freely express themselves and do whatever they want.

I'm sure your aunties have seen some of your work. Has your art changed your relationship with your aunties at all?

No, not really. I mean, they are very interesting, open people. Well, my mother and my family, they saw my work on Facebook, which is where friends and families are, and they don't quite understand it. I've received comments like, where did you find all these old people to model for you? They thought they were real photographs, and beyond that, they were just like, “Okay, it's nice.” So it hasn't changed. It's still the same.

But rather, I've heard stories and reactions from aunties outside of my family, which are really interesting and encouraging, like people wanting to pick up AI after they saw my work. Just yesterday at the Zona Maco fair, an auntie who is a traditional artist saw my prints, and she was so fascinated, she wanted to go learn more about AI.

That's interesting. Has that been a common reaction to your art, that when people see it, they want to start adopting these new AI tools and use them for their own art?

Well, I've heard quite a bit of stories about older women or older men who saw my art and wanted to try out AI. They were inspired but on Instagram, social media generally, the comments have been mixed. So you have, say, 60% of people who are very positive and supportive and think it’s incredible. And you have 20% of people thinking the artist must have taken some drugs or mushrooms. And then you also have the rest of the people who are anti AI. A mixed reaction.

Your art is very — don't know — I describe it as weird, but in a good way. And I should tell you, by the way, I know we communicated a little bit before this interview, but I haven't told you this. I showed one of your photos to my mom or one of your videos to my mom because she's into kind of, I guess, offbeat art that's a little weird and surreal. And she's an artist herself and exhibits some of her work in local galleries where she lives in Texas. But, yeah, she loved your art. She said it's so cool with, I think she had, like, six exclamation points, and she's obviously an older woman. So, yeah, she definitely connected with it and thought it was pretty interesting.

Oh, that's awesome. Thank you for sharing that.

Yeah, of course. I also wanted to ask, was this idea of the Auntieverse a natural fit and something you knew you wanted to pursue as soon as you started exploring AI tools, or was it something that evolved over time?

Well, close. In the start, when I started my Instagram, I knew I wanted to do auntie culture. So that's why it's called Niceaunties. But the structure back then was very different. I wanted to do a club. It's called Club 33 in an alternate universe where aunties, who are these interdimensional beings, would gather at the club to tell stories to share their adventures. That's why the first few posts in my instagram are about tardigrades, because it's about the aunties visiting their microscopic world and encountering all these tardigrades and then coming back to the club to tell the story. So each post is supposed to be about what the auntie shared at the club. But then as I started to make more and more artwork, it made sense that it’s of a much bigger world-building project, especially when cities came into the picture and you have more spaces and interiors involved.

For listeners that aren't aware, tardigrades are these microscopic creatures that are known for being very robust. I think they can go into hibernation with no food or water for ten years or something and then be reanimated. They're also pretty interesting looking. There are pictures you can find online where scientists took photos with a microscope. And tardigrades are kind of cute. They kind of look like tiny bears.

Yeah, I love tardigrades. They're like, so cool, so cool looking.

Let's talk a little bit more about your creative process and artistic background. For starters, when did you first discover AI tools and start playing around with them?

In the end of 2022, I started to see on Instagram some very interesting looking images. They were just fascinating and they looked very finished and polished. And I was wondering, “Oh my God, where did these images come from?” And I saw the hashtag “MidJourney,” and of course I didn't know what it was at the time. I started to do a bit of research during my end-of-year holidays, and on 1 January, I went to the website and seriously look at it and I signed up and didn't stop. Since then everyday I’ve been creating artwork using AI, it's just amazing. It's just like you key in some words and then you get some visuals.

And do you have an artistic background outside of these AI images that you think, I don't know, drew you to AI art or helped you succeed with AI art?

Well, when I was little, I used to doodle a lot and drew a lot, just like all children do, I believe. And I used to make up stories about everyday life. But afterwards I went into architecture. So I've been in the architecture industry for about 20 years, and I wouldn't say that I was an artist in a traditional way. I only started creating art, as you see it, when I started using AI one year ago.

So it wasn't until a year ago that you started learning about digital art and AI, and also more traditional tools like video editing and sound design?

Yes. That's the great thing about AI. I think it lowers the barrier to learning, and because it's so fast in generating output, you can actually squeeze a lot into the same time frame. So, say, traditionally, if somebody can make an artwork in a month, with AI you can make hundreds of images in a day, and through that process, you iterate and you learn. So I would say the learning process is compressed and it has become very efficient.

How much do you think prompting matters for generating good AI art? If you go online, there are a lot of prompting guides, not just for text to image tools, but also for large language models like ChatGPT, and there's a lot of discussion about how to optimize prompts. And some people, I think, have prompts they kind of consider their secret sauce. So do you think prompting matters that much? And how much do you have to play around with prompts to get the images that you want?

Well, asking that question is the same as asking how important do you think communication is with your team if you want to get your team to do something in a traditional office. So prompting is basically the language you use to communicate with the program, the machine. So to get a good prompt you need to experiment and iterate a lot. So basically I do that for every new image. I would change the prompt until it looks right. There are some very basic prompt structures you can find online. I always start with that and then start to switch words around, move them around, you get different results and say, for one video, I will have at least 25 to 30 prompts to get to what you see.

And I remember reading online somewhere that it can take you up to 30 hours to create a video. Is that right?

Yeah. The quickest to get a video, half a day at least. It depends on the complexity and the length of the video. So I've done anything from a few hours to a month. The longest one was for a music video because there was a client, and then you have to follow their specific instructions, so that typically takes longer.

What was it like working with a client and helping them generate AI art? Were there things you had to kind of teach them or expectations you had to set with the client?

Yeah, definitely. Especially with early AI video models. There are some unexpected elements in it, like the strange morphing or weird creations, like fingers. Fingers and limbs are a very typical anomaly that you see in AI art. And then you have to tell the clients about this artifacts. And then my clients, some of them were quite understanding. They accept it as part of AI, so they really look forward to it. While other clients, I have gotten comments like, “Oh, that's so weird. Can you make the two eyes of this woman blink the same? What's wrong with the left eye?” Something like that.

So they kind of have a sense that they want an AI video, but they don't necessarily know what an AI video means or what kinds of artifacts are typical in AI videos, it sounds like.

Yes, that's right. So I do have to explain to them. And then, yeah, they're quite open so far.

If there were no limitations on the capabilities of AI tools, what kind of projects would you be undertaking?

Wow, I think I'll probably be making full length movies by now if it's unlimited and follows specifically what I want. Wow, imagine that. That's not a future too far from now. Maybe one day you get a device where you just plug your brain into the computer, you get straight visuals.

Well, Elon Musk's company, Neuralink, just implanted its first chip in a human brain for human trials. So you never know. It could happen one day.

Do you have a plot and idea of what you would make a full length AI film about? Is it also the Auntieverse, or do you have other ideas?

Yeah, for sure. Because for the past year, I've been building the bones of this world. So now I think I'm ready to start filling it with stories and narratives. Yeah, I've tried to do that for my past few videos, actually. Like, “We are good” and a “Nail spa.”

Would the film be like a comedy or a drama or horror?

Well, maybe. Think about it as a tv series. So I'll have, like, short episodes. So, yes, they will cover a wide range of subjects. It could be all of the above. Comedy and drama and horror, maybe, and mystery. It could be anything. Just not, like, to limit myself.

Nice. Well, I hope that happens someday and I'll definitely watch it.

How has your approach to AI art changed over the past year? Do you have new tools that you're using or new processes you've started to employ?

Yeah, in the beginning, I used mainly MidJourney, and then when AI video came out, I used Runway and Pica labs. And then in the past six months, I have been using DALL-E 3 a lot because it adheres to the prompt, very much so. And there's a limit to the amount of words you can put in, so it forces you to be very succinct in your prompting. And then there's a tool called Magnific AI, which is incredible. It's like a magic upscaler. So I use that to improve the quality of my image.

Yeah, Magnific is pretty incredible. I've seen online people upload pixelated still images from 1990s video games like Tomb Raider and Magnific will upscale them to fully rendered, high-definition characters. It's pretty incredible.

Yeah. Mind blowing.

It is mind blowing. I found, in my experience, that DALL-E 3 is better at adhering to prompts when there is text in the image. So, for example, if you want an AI generated image with a newspaper headline or something like that, DALL-E 3 seems to do a better job than MidJourney. Is that your experience as well?

Yes and no. Yes, you're right. Like, DALL-E 3 could do very good text, and you need to re-roll it a lot. And recently, with MidJourney Version 6, they have incorporated text generations. I have not personally tried it, but many of my friends did, and it looked pretty good. So let's go try it after this call.

Nice. Okay, I'll have to give it a try. You mentioned earlier that sometimes in your Instagram comments, traditional artists and even non-artists will get defensive about AI art. Some people consider it stealing, right. Because artists’ work is used to train AI models, oftentimes without their consent. What's your point of view on that?

This is, like, such a big question, because if you work in the creative industry, you would know that for every single art project or creation process, we look for inspiration, and they can come from anywhere. And people constantly look at reference images from the Internet. Right? So we are influenced by everything that has come before us.

And from what I understand about the music industry, they also do sampling. In video creation as well, people take footage that has been made before to create their own new video. So would you call that stealing? You know what I mean? It's something that everybody does already. And collage, the act of collage is taking existing images, photos from newspapers, magazines. There are copyrights of other people, right. But you're putting it together in a new way, and that became your art. So do you call that stealing? I have not heard that kind of conversation before until AI happened.

So I actually think that, firstly it’s a process that allows you to stand on the shoulder of giants. We have never been so connected before with this data that's already there. It's just faster access, more efficient access to these datasets, and you're using it to create things in a more efficient manner. So why is it stealing if it's already existing behavior?

Right. But how would you feel if someone used your work to train an AI model? As you said earlier, you've produced tens of thousands of images and videos. Now, not all of them have been made public, but at this point, you do still have quite a volume of work that can be scraped, that's online. An individual or a company could create the Auntieverse text-to-image tool that's specifically in your style. And then instead of having to spend 30 hours on a video, any person could just come and they could type in a prompt and they could get a video or an image in seconds in your specific Auntieverse style. Wouldn't that bother you? Because that's basically analogous to what many traditional artists and non-artists are objecting to with these AI tools and how they were trained.

Well, it is inevitable, isn't it? When your work is good, people will copy it. I think that's okay. And then I would like to quote Rick Rubin at this point. I heard a podcast of his recently, and I felt like it's very on point. He was asked about AI art as well, and he said he doesn't know much about it, but he understands about these huge datasets. But it's not really about the data, but more about the artist's perspective. So for your work to stand out as an artist and for you to have your own identity, you need to have your own unique perspective on things. That's how you can differentiate yourself from other people. So we all have access to the same resources, but your ideas and concepts can easily be different to other people.

So, yeah, people can copy me or whatever I'm doing or what other people are doing, but do they have their own opinions? You can sort of look at their accounts to get an idea right. Are they sort of going all over the place, or do they have a consistent narrative?

Yeah, that's an interesting point of view. Do you think more traditional artists should be experimenting with AI tools, since it's a kind of new technology?

Well, I think that all creative people should be open to all possibilities and all mediums. So why not try it before deciding that it's bad and you don't like it? I wouldn't say no to anything, being an artist.

Have you had any conversations with traditional artists about this topic?

No, not really. But I've had positive affirmations from traditional artists about my work, not because of the medium, but because of the content that comes through. Yeah. What do you think?

I have mixed feelings. I mean, as an artist, it must feel like your life's work is basically being used against you. Your artistic output is gathered without your consent, and it's used to train AI models that can do in, I don't know, seconds what you spent basically your entire life learning how to do. And that can't seem fair to traditional artists. And many artists may be put out of work because of these AI tools. So basically their work is kind of being used against them in a sense.

On the other hand, I'm generally quite pro innovation. Text-to-image tools and large language models are truly incredible, and I think they can unlock an immense amount of creativity and productivity across the world. And I find AI art in particular to be almost like a new medium. The way that AI art tools generate these strange artifacts we were talking about, and they create these kind of morphing effects. It's a kind of visual imagery I haven't seen before. And I find your work and the work of others who employ AI tools to be genuinely beautiful.

Thank you. I suppose it's the same with everything in life, right? I mean, things are going to keep moving forward and some things are inevitable. So what do we do about it? That's life attitude, I suppose.

We're almost out of time. Tell us about your upcoming plans. Will you continue to make AI videos and still images about the Auntieverse? And do you have any other plans upcoming for your art?

Yeah, for sure. I'm going to continue making AI videos. Stills, not so much, even though making stills is part of my AI video creation process, but probably will not publish them as much.

And then I'm going to have a physical solo show in Berlin in April, at the end of April, during gallery week. So really looking forward to that. And also lots of things coming up in the physical world and hoping to manifest some of my ideas in other art forms. So that's very exciting.

You mentioned earlier that you have a day job working in the architecture industry. Are you going to continue to work in that capacity or are you planning on trying to move full time into AI art and invest your time there?

Well, I would like to think of myself as a multidisciplinary artist. So yeah, I'm still going to be involved in architecture and my time is pretty flexible. So yeah, we just go with the flow, but definitely a lot of energy will be placed in art and AI art.

What a life. That sounds fantastic.

Yeah. Thank you. Really excited, getting well.

Nice aunties. It's been great chatting with you. Thank you so much for joining me.

Thank you, James, it's been a pleasure.