Curating archives was a theme that I also explored in a conversation with Tamara Shogaolu about her piece ORYZA: HEALING GROUND that premiered at DocLab 2024. She was at SXSW showing the interactive AR story game ANOUSCHKA exploring the memories of her grandmother through various cultural artifacts (see Shogaolu interview with her about ANOUSCHKA from its world premiere at DocLab 2023). ORYZA wasn’t showing at SXSW, but I had a chance to catch up with Shogaolu to unpack how she’s working with archives to document cultural artifacts enslaved people, and then expand it into speculative art with a consistent aesthetic using a custom-built AI system that is centered on Black histories. Shogaolu details her experiences of instances where large-language model AI systems have erased Black people, and how she’s had to hack existing AI systems in order to get the outputs she wants. But she’s also training entirely new models using materials from archives that are not included within existing models. We explore the paradox of desiring cultural representation in these AI systems while also avoiding the negative side effects of extractive cultural appropriation, and this new trend of training custom AI models as a form of artistic expression leading of future dreaming to unique mash-up of speculative futures and historical archives.
This is a listener-supported podcast through the Voices of VR Patreon.
Music: Fatality
Podcast: Play in new window | Download
Rough Transcript
[00:00:05.458] Kent Bye: The Voices of VR Podcast. Hello, my name is Kent Bye, and welcome to the Voices of VR Podcast. It's a podcast that looks at the structures and forms of immersive storytelling and the future of spatial computing. You can support the podcast at patreon.com slash voicesofvr. So continuing my coverage of South by Southwest with different immersive stories that were there this year, Today's episode, I'm going to be talking to Tamara Shogolu, who had a piece called Anushka, which I covered previously at DocLab 2023, as well as a piece at DocLab 2024 that I didn't get a chance to talk to her about. And that's actually what we're going to be talking mostly about today. It's a piece called Arisa that has this kind of mixed reality installation that was at DocLab 2024, but also augmented reality components. But also she curated this whole archive of material and then used AI to create this speculative, imaginative mashup of historical archives, but in a way that has a speculative future aspect to it. So it's Kind of like this interesting new genre that I'm starting to see where artists are starting to cultivate and curate historical records to start to fill the gaps for stuff that AI isn't covering in their existing models. So in the process of creating her piece, Arisa, Tamara has trouble getting these AI models to accurately represent black people. She had this experience of black people being erased from these different experiences that she's trying to create. She's specifically looking at black farmers in the South, but also these enslaved people from West African cultures and the different ways that they were bringing over their indigenous knowledge of farming into the context of the American South during the time of slavery. So yeah, lots of really interesting things that we're covering there. At South by Southwest, there's a piece called Anushka, which is an augmented reality piece that we talked briefly about. But I'd recommend checking out my previous interview that I did with Tamara back in Amsterdam and IthadocLab 2023, where it won the Digital Storytelling Award back then. And so it was showing there at South by Southwest. And that's also a really great experience where It's an augmented reality story where you start to go into the memories of this character's grandmother. And it's like you're taking an adventure to learn more about your ancestors' memories. So that's Anoushka. I talked more about that in a previous conversation, and we cover a little bit of an update. But in this piece, we're going to be really diving into Arisa and the ways that she has to kind of hack these artificial intelligence systems in order to really get what she wants out of them. So we're covering all that and more on today's episode of the Wishes of VR podcast. So this interview with Tamara happened on Monday, March 10th, 2025 at Southwest Southwest in Austin, Texas. So with that, let's go ahead and dive right in.
[00:02:54.147] Tamara Shogaolu: All right. Hi, my name is Tamara Chagallou, and I'm an XR creator, a multidisciplinary artist, I guess, and filmmaker. And I utilize technology to tell stories. Usually my stories take place between virtual and digital spaces and connecting those through story.
[00:03:11.211] Kent Bye: Great. Maybe you could give a bit more context as to your background and your journey into the space.
[00:03:15.256] Tamara Shogaolu: Yeah, so I guess my journey into space started maybe 15 years ago or so when I was in film school. I went to film school at USC and there was an interactive media program there and I started taking some classes and then learned some basic coding to do interactive experiences and I was really blown away by the the impact that this could have in really simple ways. And I started wanting to explore how to tell stories across different platforms. So after I finished school, I worked on a VR piece called Another Dream, which took me about five years to complete. And I made it on the HTC Vibe. And there was an aspect of co-creation with it, where the story tells the story of a lesbian couple from Egypt who got asylum in the Netherlands. And they were gamers, so they wanted gaming to be a part of their narrative. And I utilized VR to create an immersive documentary animated experience that connects you through them, but you have to interact and engage to move forward with that, yeah.
[00:04:16.021] Kent Bye: Nice. And so I know you've had a number of different projects over the years that I've had a chance to see and sit down and chat with you. And so you're actually here at South by Southwest in 2025 with the project that I first saw back at its premiere at IFA DocLab 2023, which is Anushka. So maybe we can just start there briefly, and then we'll go into the piece that you had this year at DocLab. But yeah, maybe take me back to Anushka and just give a broad context for what this project is about.
[00:04:39.912] Tamara Shogaolu: Yeah, so I started working on Anoushka in 2020 and I would say I'm still working on Anoushka to a certain extent. And it first started as a location-based experience and the pandemic happened and I had been working with Niantic on another project and started playing around with Lightship. And I thought, oh, maybe this could be actually an interesting storytelling platform. So we started adapting Anoushka for Lightship and working with it. And it has three levels, and it basically tells the story of a young girl who, when her grandmother passes away, she's kind of dealing with the grief and ends up traveling into this time portal and then traveling through her grandmother's memories. in order to kind of reconnect with herself and find herself. And I had lost my grandmother, and my grandmother was a really big storyteller, and I think I always just wanted to do something where, like, you're a girl who gets to go on adventures with your grandma. And that tied with sort of some of the archival research and history that I had learned about this particular neighborhood in the Netherlands. inspired the world building aspect of it and the story and we premiered it at IDFA and since then we've been trying to improve it and it had quite a bit of success in the Netherlands and ended up touring and the story takes place in the Netherlands so I think there was a moment where I wasn't sure if international audiences were going to connect with it And then we made a Dutch version as well and made some upgrades to the experience. And now we're here sharing it with the international audience for the first time in the US. So it's been really interesting to see how people connect to it because we thought maybe it would be too niche a world, but people seem to get it, which has been really cool, to be honest. Yeah.
[00:06:20.854] Kent Bye: Yeah, I saw it. And I'm an international audience when I saw it. And I got it. So yeah, a really beautiful puzzle game augmented reality piece. We had a chance to sit down and unpack a lot of it a couple of years ago. And it went on to win the Digital Storytelling Award at IFA Doc Lab. And now it's here in a different spatial context. Because when I saw it, it was like in a whole garden that you could go and explore a big, large space. So maybe just talk briefly around how did you manage to convert it into what was a wide vast space into much more constrained space.
[00:06:52.577] Tamara Shogaolu: Yeah, so it's designed to be experienced outside. And I think that's the optimal version of it, really. But for this, there's a function that we use to like recenter yourself in it. And so we knew that we had a smaller space. So now we're like having people use the option of recentering themselves. And it's been interesting seeing it in an indoor context as well. because we hope people can play it at home. I mean, I started working on it during the pandemic, so I really wanted people to like go outside and move. And I think that's what I wanted to do. So I I enjoyed during the pandemic being able to test things and be outside and working instead of always being in front of my computer and indoors. But yeah.
[00:07:31.414] Kent Bye: Okay, well, I'd like to move to Ariza, the piece that you had at IFA DocLab this past year. And maybe before we start to dive into the project specifically, I'd love to get this general sense of some of the prior work where you started to do like these physical installations that were leading up to what you were presenting at IFA DocLab this year. So take me back to when did you start to actually make physical installations as a part of your art that was having a virtual component or a digital component with this kind of like sculpture based objects that could then be a target for augmentation?
[00:08:06.156] Tamara Shogaolu: Yeah, I mean, I think my first foray into the immersive world was through sort of physical. I mean, I learned how to make websites when I was younger, so I think I was interested in that aspect. But I remember I made a really simple installation when I was in film school that was basically with the Xbox just like tracking and visuals changing as you did it. And now looking back, I remember I thought it was the coolest thing ever. It was really basic, but it really impacted me seeing how people, you know, they were just like, oh, wow, the color changes because I did something or like the shape changes because I moved my arm up or down. And it created sort of connection with each other. So like multiple people could experience something at the same time. And when I was working with VR, I realized that you have this really sort of like deep and intimate experience with it. But I wanted people to connect with one another as well. So I think that started making me be curious about augmented reality. And when I was working on Unresolved, that was sort of the first time that I started doing sculptural work that had like hidden levels, layers in it. And I think experimenting with memorials in the sense of like how they occupy physical space but then also the sort of hidden stories that you could explore adding another level of information and content or experience for audiences connecting it to something maybe aesthetic or physical that they see that could maybe prompt curiosity to want to learn more. And since then, I think it's just become a part of the process. But I think even with VR, now that I think about it, you always want to think about onboarding and offboarding people and how the installation matters. Like you don't just put on a headset. And I think thinking in the way of like onboarding also led to me creating sort of like physical aspects, because even with Another Dream, we made like a video art piece that you would experience before you went into the installation. And every festival has a different setup, so I remember just having to adapt. And I was like, maybe I need to make this physical thing part of the experience itself instead of having to just have it be an add-on that you're tailoring as you go festival to festival.
[00:10:17.385] Kent Bye: OK. And so maybe you could talk a bit about the origins point of the project that you just showed at DocLab 2024, which was Ryza. So where did you begin with this project?
[00:10:28.231] Tamara Shogaolu: So that's a really good question. Basically, it started when I was working on Unresolved, which was a project about 162 racially motivated unresolved cases. And in that process, I realized a lot of the cases and people who were murdered They were farmers and they were killed because they like own land and these sort of issues around land justice that were really present at the time. And then I started reading more about different experiences of black people and their relationship to land. And that was something that started in the research that I was doing with Unresolved. So I started digging deeper. I traveled around the South for probably like two or three months collecting oral histories from black farmers. And in the process, I sort of learned about this hidden history of farming. And I found these journals of early Dutch explorer who had written about these rice techniques that existed in West Africa and that West Africa had been called the rice coast before it was the gold coast or the slave coast. And that was because like the knowledge that was connected to growing the rice was also what was then used to build the plantations in the U.S. And I started reflecting on how when I learned about slavery, I was just kind of taught that it was like a physical labor thing. And then I started reflecting on the technical labor and that, you know, to domesticate rice and create rice techniques over 3,500 years, people often think about China when it comes to rice techniques and it was happening in parallel. but those scientific and agricultural contributions had kind of been erased. And I wanted to connect to that further and create a deeper connection to the land. And also, I think what interested me was that, you know, people being targeted for like indigenous knowledge of working the land, but then also there's a separation of the land that happened because of the pain of having or enduring slavery where working the land became something that in a lot of black communities people didn't want to do because that felt of the past and sort of like what that forced separation means. And at the same time, I had a friend who was part of this new movement called the Black to the Land movement. And they had been telling me a lot about how the land had been part of like a restorative sort of healing for them and their experience. And I just remember being so moved when I heard that, that I felt like there was more to explore there. So I just started talking with tons of people who wrote books about farming and different sort of like forms of land stewards and exploring that connection. And that really started the piece. And I think from then it started evolving. And originally I wanted to continue this idea of telling stories, utilizing or reimagining quilts as storytelling platforms. So with Unresolved, that was the first sort of piece that I did where I designed a sort of digital quilt in a physical space that you can then explore further. And with this project, I saw it as a continuation of that work where quilts could be other ways of archival, in particular in relationship to how African-Americans would utilize quilts to tell stories of families and pass it generation to generation. So I was like, maybe an archive can exist in a physical sense in that way.
[00:13:45.529] Kent Bye: Yeah, and there's a lot of components to this project where there's a physical installation. In that physical installation, you have a number of different videos. In the videos, you have documentary footage with some of these interviews and conversations, but also a lot of generative AI techniques that are poetically imagining new forms of culture with artists. AI trained on some of these patterns to create new patterns. And you can get into the specifics of that Afrofuturist part, maybe. I don't know how you describe it. But then there is also an augmented reality app that also has augmentation to this installation. And you have the audio soundtrack to everything that is going on. And then you have a bunch of written text that you wrote to explain even more context to some of the stories that you don't have the space or bandwidth to really dig into some of those other stories. So with all these different components to this project, which ones of these did you begin and how did they start to play off of each other? Just trying to get a sense of how it all evolved.
[00:14:43.396] Tamara Shogaolu: I'm trying to think, too. That's a really good question. I think it started with the documentary, to be honest, because I took the camera with me when I was doing research and I knew I wanted to continue in the work of Quilts' archives and continue to explore that. But I didn't quite know in what full form it would take. So I went out and I started shooting and In the process, I realized that there was just so much information and like so many layers to it that I actually decided that it needed to be a film as well. So I started developing a documentary. And while I was working on it, because it took me four years of research of working on this project, I saw what was happening with AI in particular with like the literal erasure of black people. I was playing around with it, you know, and I would like input, try and do style transfer. And in style transfer, I would literally turn black people into white people, like completely erase people. And then I realized I was like, I'm trying to hack these things, you know, like this tech. And I was just playing around with so many different AI platforms. And there was like some really funny things. Like I did a test I wanted to generate an image of, I put like a prompt in, I can't remember, I think it was Runway, and it was like black woman twerking. And when I put that in, it told me that that was against policy. And then I put white woman twerking. And then it generated something. And then I was like, it's interesting how in the language and the policy, one thing is being sexualized and like deemed inappropriate and one isn't right. And how that bias becomes part of even the way you interact with the technology. And the more we interact with it, the more the way our brains change. Right. Because you use different software and you start learning how to talk to it. So you're like, OK, this model has that weakness. So I'm going to hack it in this way. And that one does and has a different one. And then I wanted to kind of like step away from that. And I was like, how can I utilize this and maybe train a model that creates a connection back to the land? And I wanted because of the quilting theme to create a literal thread. And in the research that I was doing, I'm really interested in symbols. so i started just collecting like you know i had already started this work with unresolved with like the meaning of different motifs that were set up and then tracing the history of them and many of them go back to west africa and there were simple things that i found out with the researchers that we were working with as well that like for example there were symbols that existed in West Africa, but also were found in African-American quilts, but then were also found in fences that welders did in South America. So everywhere that the African diaspora existed, these symbols were kind of in the, you know, like you see them in buildings, you see them in places, and I didn't know that they had traces back to Africa. So I wanted to train a model that could create because, you know, they try to find the similarity between places. I was like, I can utilize AI to build visual language that unites the diaspora and creates like a sort of new visual thread or connection. So I saw this project as like a time travel portal in a way. Like so if you see, you know, the sculpture kind of looks like a portal. And with AR I wanted, now we've built like a fully AR version that you enter it as a portal and like enter this other space that you're like inside the deconstructed quilt. And you get to like interact with all of these symbols that are basically like uniting the diaspora. So all the layers kind of feed into one another where I utilized a lot of generative AI tools by like hacking them, but they're also all based on actual archival. So like each image I partnered with the Dutch National Archives and used a lot of their archives to then sort of deconstruct things like maps and imagine them in different ways. And there was a lot of like art that I did by hand or myself and then would input into the models and like trying to play with how to hack around things.
[00:18:41.144] Kent Bye: Yeah, there was a session by Ingrid Kopp at IFA DocLab R&D Summit where she was trying to have these group discussions around AI. And it was kind of an interesting discussion because there were some people who were basically saying, we should not use AI at all. And then Ingrid, she's based in South Africa, and she runs Electric South. And part of her perspective was there's this kind of paradoxical thing where you don't want the AI technologies to have this colonizing effect of seizing control over culture. But at the same time, you also want to have some representation. So if it is going to be happening and existing, then if parts of your culture are not existing within the context of the data that the models are being trained on, then there is this other erasure component where these colonial impulses are doing double damage in that sense of like it's taking control of the information and data from those sources without full consent, but also not really being in right relationship for all the things that are not included into it. So it's kind of like this weird paradox where both things are happening and harms on both ends. And so just curious to hear some of your thoughts and working with the models, because there is this idea that was also shared, which is that perhaps part of the role of the artist of the future is to create these archives where you're now the curator of this content that then is allowing to have that more fuller representation. And then from there, perhaps have a little bit more agency to see how you want to remix and reimagine that past into some sort of future that doesn't quite exist yet. So it's this very interesting blend between the past and the present and the future where you're creating those archives, but then using that to leverage, but in a way that you are negotiating your own ethical boundaries for how to not exploiting that culture in a way that doesn't feel like it's in right relationship. So anyway, I'd love to hear some of your thoughts on, on that role of artists who are working with AI and all this way.
[00:20:29.887] Tamara Shogaolu: Yeah, I know exactly. I think I was there and I was thinking about it, too, because it's something that I really struggled with when before I decided to start working with it. But I think for me, I don't know. It's just like it was really interesting because not only did I like sometimes try and deconstruct archives or try to imagine you know um things that aren't in the archives and the possibility of like recreating things that maybe like you know there's a certain aspect of my imagination that i'm imagining that this is what the archives are saying to me right but i also sometimes utilize like quotes from interviews that i did or from archives and then would put them in without like telling an image so One of them was just like, I want to build. I remember somebody said, like, I want to build a black utopia where everyone is free and equal. And I put that in and it gave me like 10 women in saris dancing around the tree, like as the image. And I was like. because then you're like you you know how you know the models are trained and then you start wondering like huh like where you know like what does that mean and you kind of start being in conversation with it and i feel like to a certain extent ai is an archive where it's trained on like everything from our society is like a reflection of society right so it's a reflection of like all our biases and everything as well in it so it kind of feels like you start having a conversation with historical record to a certain extent but I wouldn't call my, I like fear calling myself an archivist in a sense where I see myself as a role as an artist to be maybe more of like a medium or like of how I've connected to some of the archives and my interpretation of the information in them and sharing them. Because I feel like it's a slippery slope to like having revisionist history to a certain extent. And that aspect of it scares me a little bit because I wouldn't want to like reimagine history in a way. It was for me, I saw it as like, OK, we did all this research. Here's the real research I want you to see. And this is my interpretation and my dreaming and imagining of what isn't in there, so to speak.
[00:22:40.936] Kent Bye: Yeah, and I referred to that as Afrofuturism because there's this element of using the imagination to imagine these speculative futures, but it is also like leveraging these archival materials. So I don't know how you like self-describe this kind of emerging genre of this realm of taking these archives and then using these generative techniques in your own imagination to kind of create something new.
[00:23:02.926] Tamara Shogaolu: I don't know. I mean, I feel like you'd be better at naming it. You tell me. You see everything, so you know. You have the big picture. Yeah, I don't really know. You know, I was really inspired by Julie Muratu, who does the big paintings and maps and this relationship with the past, but abstracting through it and emotion. And I think I was really moved by her work and how I could utilize that in my work in terms of imagining by deconstructing But I would be afraid of saying that, like, that's why I think I wanted the real text and the real research. And I made sure to cite everything to be there because I didn't want it to seem like I was making up this history. Like, oh, this is now history. I was like, this is the research and this is my interpretation. So I don't really know what it is. I mean, I think I guess it's like abstraction. Somewhere like maybe, or what does it, Rafiq Anadol calls it, future dreaming. Maybe that's it. Maybe it's a future dreaming in a way, but grounded in the past. Yeah.
[00:24:08.737] Kent Bye: Yeah, it's sort of like elements of speculative futures and other kind of futurist genres and domains. So, yeah, there have been conversations I've had with black creators who are resistant to the term of Afrofuturism. Specifically, it's a conversation I had around the piece that was at Sundance a number of years ago called Interstitium of Octavia Butler. I didn't know if you were self-identifying with it. It kind of feels like it's in this new emerging genre where it's kind of a blending of all these things. That's maybe why I was asking, because it's unclear as to what this new form of working with AI is. Just to take a step back, because I see a lot of AI slop, let's call it, like this AI that just... is not appealing and it just feels very extractive and not intentional and lowest common denominator. And there's a lot of like music videos that look very polished and it's getting more and more, but in terms of like spiritually and culturally just feels very vapid and hollow and not really that interesting to me most of the time. And so that was the context under which that I was seeing your project. And I was like, Oh, this is actually like quite interesting. You know, the way that you had to have done some serious prompt engineering to fight against these systems because it doesn't seem like this is just what is naturally being output. So what was your process of trying to coax and manipulate these systems to kind of go towards what seemed to be a pretty consistent aesthetic that you had across this entire piece? So yeah.
[00:25:36.522] Tamara Shogaolu: My fear of AI replacing animators like I think went out the window after this process and it became very clear that they're tools and I think I mean I even think back to like you know it always starts there's always like this sort of exclusion of people of color in many spaces with new technology where like Even with makeup, you know, I remember when I was a teenager and they didn't have so many shades of makeup, but now you can find like every single skin color like as something as basic as that, that you like people keep kind of like hacking things. So I felt like, you know, if I had to hack makeup as a teenager, like now I'm hacking AI as an adult to like because it's basically if this is like our record, our public record library that we're interacting with, and saying that I don't exist in that, that didn't sit right with me. And I wanted to really, through the hacking, create a conversation of what aspect of me doesn't exist and what are you seeing? And it was a tedious process, but it was also really just interesting. Because you learn about society while working with it. And now I feel like I see an image and I know exactly what platform software was used for it. Because they each have their own kind of looks. And sometimes you have to mix things. And it was definitely hard to set up my own taste or approach to it. And honestly, even in the end to... unify things like I did work with real life animators as well to like be able to make some cohesive things because I wanted there to be like smoother transitions and things like that. So there's like a base of the work that's AI, but then there's also like a top polishing level that definitely needed to happen. But I think that conversation with the AI that happened in the process of making it is also part of the piece in itself for sure.
[00:27:35.270] Kent Bye: Yeah, and what's interesting about these pieces is that I found that I don't really understand the nature of the AI until I have the conversations with the creators to hear some of the challenges or stories that they had in the process of their creation of the piece. But within the piece itself, it's obscured to something that is a part of your own artistic practice of things that you have to go through in your journey of trying to manipulate and control it. And there was another piece that was at IFA Doc Lab this year. I'm not sure if you had a chance to see it, but it's called Burn From Absence. piece that was next to the wall, and it was based upon reimagined family memories because the artist, Emile, has Vietnamese heritage, and her grandmother had destroyed a lot of the photos of the family in that she wanted to recreate a lot of these family memories, and it was in Vietnam in, like, the 70s, and so every time she tried to create any sort of prompts, it would immediately go into, like, a war context of this... militarized or very stereotypical depictions. And so she had to kind of fight around and find other ways of either using Vietnamese language explicitly or finding other ways of describing things in order to get the aesthetic that she was, that felt like it was more authentic to the villages that she was trying to recreate. And she spoke around that as a challenge that she had to overcome. But as you watch the piece, you would never know that that was a part of the limitation because it's not really explained. And, you know, it's a part of, again, it's part of the artistic process for how it came about. So in the process of creating Arisa, I'm just curious to hear around, like, what was your journey of trying to find what you call these hacks? Like, what are the specifics of, like, how you are hacking AI through this prompt engineering process?
[00:29:17.051] Tamara Shogaolu: I mean, let me see if I can think of a good example. I mean, so I... I wanted to like originally because it takes place across like the African diaspora through different time periods and. Every frame is grounded in real archival right so there's like a starting image or map or something that I started with and then either like you know I would utilize like photoshop to. kind of like tweak or design them or like abstract them however or I would input the original archival and then play with prompts in order to like how to change or reimagine what I wanted to say about it and I wanted at first like this aesthetic where black people were like literally the color black and like more in silhouette kind of like I don't know if you remember with Unresolved like on the sculpture there were like these black silhouettes that were there so I wanted to like continue that visual aesthetic and like have them moving But yeah, like black people just didn't exist in some of these tools. And I remember having to write like very dark person, like African, and then that would come with like a certain imagery of like how Africans are perceived. And I could tell because, you know, we worked on this maybe in like a year that we were animating and playing with it, like, the tech was getting better, because they would always, like, give us South Asian people as, like, dark people. So I had to start kind of inputting, like, a black person as a character and, like, describing this as a black person, and then that would become part of the character of the scene as, like, a reference. I was, like, training, in a way, the AI by telling it, like, this is what a black person is. And it was, you know, like, having to explain that was, like... was interesting but that was like the way of doing it because like I remember one time I wrote like black black the color black black like like and it still wouldn't like um so that was like that's like one example um I also like I think It was interesting also with plants, which I guess maybe so like I loved seeing, you know, the Future Botanica project that's here also because, you know, the Oriza comes from the name for rice, right? So it's like starts kind of with telling the story of rice. And we were trying to generate like rice plants to have that be part of it and like seeds. And I just started giving me like really trippy things. And then you just kind of go with it. And then I was like, you know, I'm just going to like reimagine what rice plants could look like. And I gave them like just playing around with it and went like and built some really crazy looking rice plants that ended up looking pretty cool and like then we turned them into models and like now they're in the ar experience but even with like it's just not accurate right so it was like the plants didn't even look like the scientific drawings that we put in there so then you just kind of start breaking it and i would just sometimes like remove elements of it and put other things in or even like just tell it really crazy things. Like I would sometimes describe, like I'm trying to think, like I put in a poem and then there was this poem. I don't know if you've heard the short story of the Flying Africans by any chance, but like Faith Ringwald used to put these Flying Africans in a lot of her artwork, which is based on this drowning that happened at Igbo Landing. I think it was off the coast of Georgia. where a group of enslaved people committed mass suicide when they had taken over a boat and they were about to be captured and they decided to commit mass suicide instead of being enslaved again. And there's this myth in African-American lore that they grew wings and flew back to Africa. And I remember Reading poems about it. It was like present in a lot of books and different stories So I wanted to add those elements in it because it was all about like connecting back to the land and Africa so I put elements of the poem in there just to see what it would give me and It was really interesting because sometimes it would give me stuff that looked like it was like out of Black Panther and then sometimes it would give me like literal birds with like African heads like And it was just like, so, you know, like it was like a weird poetry also, you know, where you're just like, is this how different people could interpret this poem and what that means? So that was like another way of hacking where it was like you could work with it or you could really try and manipulate it. So yeah. Some approaches I just like gave into the poetry of like what it's interpreting and the conversation with it. And then in other instances, I just like really tried to get it to give me what I wanted by creating characters and inputting it and like bit by bit training it and then putting in the image again and then changing it. So it was a mixed process.
[00:34:09.344] Kent Bye: Yeah, it's reminding me of the work of Joy Bolamani, who was featured in the documentary called Coded Bias. Have you seen that? Yeah, she's at MIT, right?
[00:34:17.069] Tamara Shogaolu: Yeah, yeah.
[00:34:18.174] Kent Bye: Yeah, so she was looking at facial recognition and finding that the facial detection algorithms were not detecting her black face. But when she would put on a white mask, it was detecting her face. And so she started writing a lot around the ways in which these data sets were not completely representative and also biased towards working for white folks and then not working for black folks. And so there's this whole documentary that I think To me, when I think around AI, Daniel Lufer, he works at Access Now and was looking at AI Act. And the way he put it to me was that a lot of the moral reasoning of AI ends up being very utilitarian, meaning like what's going to serve the most people. But if you take that approach, then the people who are the most marginalized are then not representative and also continue to be marginalized in a way that is amplifying that marginalization. And then what he's advocating is that you really need to have a human rights approach, which is trying to counteract that by having certain rules that you're able to de-bias those existing cultural biases. It seems like it's, like you said, it's a reflection of culture, but it seems like the default moral reasoning of utilitarianism is amplifying these imbalances. And so whenever I think around AI and people who are really gung-ho around advocating and I always think around but what about all these blind spots or harm is being done but they may not even have any anecdotes or examples of specific things so I come back to conversations that I'm having with artists like this conversation I'm having you with right now that sort of grounds those experiences where you've started to bump up into some of those limitations so I'd love to hear some of your thoughts
[00:35:56.997] Tamara Shogaolu: I mean, do you know Halsey by any chance? So he's another artist, and he had a project that was about creating a Supreme Court justice model. And I think when I heard him talk about it, like if AI started implementing our laws, And it looked at, for example, how many people have been convicted for this crime and created data on like the probability of you being a criminal for X, Y, Z is greater. And then court becomes something where they scan your face. And then because of the probability of because like, let's say more black people have been convicted of these crimes. Like, therefore, that means you're more likely to be a criminal, right? Or like, even if you listen now, I was just watching the news and there was somebody who was an American citizen, but was Latino, was like arrested by ICE because of the probability of what they thought based on bias right so if you have behavior or like data sets that are created from bias and then it's being like trained on that as well like it gives little room for like the human brain to assess the situation which is why judges exist and the sort of automation of things like limits access and stuff and i i've been reading about also like immigration models that are starting to try and use AI to like gauge people and I heard about a case where it was like an interracial couple where there was like a white European woman with a black African man and they were like moving to Europe and the AI model said there was a probability that this was like a fraudulent case so they got denied but then they were able to appeal and meet with a real judge and stuff like that and solve it but Just think about how much of a pain in the butt it is to call the bank with automation nowadays. You can spend a whole day with the bank until you can speak to a human to try and solve something. Or even when they ask you to prove your own identity and then they ask you where your cousin that you never talked to lives and it's associated with you. There's all these random questions that automation has created and if that bias is added and powered by automation, I think that, to me, concerns me. So like, I don't know, to go back to your question in a way, I guess, I think kind of like what Ingrid was saying, like, it's better to not exist at all if that's really an option. But we do exist in the space, just in a negative way. So it's like there needs to be another option for this if this is going to be like building our future systems, you know?
[00:38:29.642] Kent Bye: Are there any other examples or anecdotes that you have with your encounters with AI that you've directly encountered these levels of bias, either chat GPT or text-based things or other contexts in which that you've directly encountered that type of bias?
[00:38:45.023] Tamara Shogaolu: Yeah, actually, like one of my colleagues once like in the office asked ChatGPT, like our studio is called Auto Auto Pictures. And he asked ChatGPT, like, tell me about Auto Auto Pictures. And it said that it was a studio created by a man named Jan Peter. and like completely erased me from my own studio and work. So I thought that was pretty funny. I was like, well, I don't exist. And I guess Jan-Peter started our studio. So, I mean, at that point I was just like, okay, I hope people don't start using this as like fact, you know, and that's those things where it's like a mix of it's, you know, a modge podge of everything coming together. To find the common denominator, basically, which I think is interesting, but hopefully there's like education that goes along with that, so that people can like know the difference between models or what to trust and how to check things but. yeah I mean i'm excited I don't want to sound like I like all doom and gloom about Ai because I do think it's an interesting tool that has potential but. I am scared if there's no regulation or framing around it or education that goes along with it so that people can reach their own conclusions, I guess.
[00:40:01.027] Kent Bye: Yeah, it's just talking to some artists here. There's a lot of excitement for different ways that they're having very pragmatic uses in the way that it's in their production pipeline from concept art to creating more rapid iterations for creating these more vast immersive worlds. But yeah, I guess I also think about these conversations that I have with folks like yourself and other artists who have more directly and bumped up into some of those limitations. Yeah, I don't know. I feel like I'm kind of split, but I tend to lean more towards some of the skeptical side, but also I use WhisperX to do transcriptions of interviews. But I also don't know if there's systemic bias of mistranslations that have happened because I don't have the time with how many interviews I've done to basically do a cleanup or a correction of all that things that may be inherent to the model. So there's ways in which that I'm simultaneously skeptical of it, but also using it in a way that I find my little pocket of utility that I find really useful for making even the conversations I have a little bit more accessible than they would be if they didn't have those transcripts. So I think Ingrid, at the end of the day, was saying, as artists, it's our responsibility to try to explore the potentials of creativity, but also maintaining a critical eye and maybe even turning towards more policy issues to critique it at a larger systemic level. There's no other leverage points with working with the technology. So that's her own personal path of starting to look at it and look at the policy implications. So, yeah. I don't know if you have any other thoughts on that.
[00:41:27.327] Tamara Shogaolu: No, I mean, I feel exactly the same way. Yeah, I agree with that. Yeah, I hope we get some policy. That would be nice.
[00:41:38.209] Kent Bye: Well, there's a number of other components about your piece. There's like augmented reality components. There's a lot of text that's also written there. So I'd love if you could maybe give a brief overview of how those other components are fitting into this overall immersive experience of Arisa.
[00:41:52.833] Tamara Shogaolu: Yeah, I mean, I think while I was working on it, especially because I was partnering with an archive organization, I learned a lot about the work and role of archives and I think appreciated it in more ways than I really understood before. And I then started looking at how they were colonial tools, like particularly like the Dutch were very good at keeping archives. So like the amount of detail you can find was just like incredible really. But then you realize it's a lot about like ownership, like why did humans start creating archives? And it was to like track land ownership, ownership of people, like, you know, transactions. It was like all kind of like, very extractional so to speak and now i feel like we look at archives as way of serving as memorials or legacy or people utilize them to find their own family histories and i found that really interesting that it went from being like a tool for commerce to being a tool for like connection or or society you know and i always think about that show like finding your roots you know like like where you go into archives so I wanted to kind of explore that and also tying it to like you know AI as I see AI as an archive and a living archive in a way that continues to expand and reach conclusions that we get to interact with so I found that dynamic interesting but I wanted to also be able to like directly connect people to the archives that I worked with and it's really it is text there's a lot of text in that and i something that i had to toy with myself like will people read all of this but i think with a lot of my work i like people to kind of explore at their own pace so i wanted people to be able to like okay now i'm going to read one chapter tomorrow i'll read another and like just work on it bit by bit but i felt like i needed to share all of that information like i just wanted people to have access to it so that they could reach their own conclusions because I didn't want them to just explore, see the film or explore the interactive elements with the voiceover without like really knowing the root of the real world archive and how then it became this abstraction or like dream version of what I saw.
[00:44:08.703] Kent Bye: Yeah, I think I was having some trouble with the triggering of the text, or at least I was trying to bring it up on my browser to go read back where I was staying, but it was kind of disappearing. So I ended up taking screenshots of everything, but then it was a lot of screenshots, and then I just finally got it. Website.
[00:44:25.489] Tamara Shogaolu: There's a website. I could have sent it to you.
[00:44:27.231] Kent Bye: Oh, yeah, that would have been easier. But there is one thing that I just wanted to have you read real quick. Just before we had our conversation, I was going through and reading some of the different texts, and this paragraph really jumped out to me. I'm wondering if you'd be willing to read that.
[00:44:41.556] Tamara Shogaolu: In recent years, African American quilting has made waves in a craft that is often associated with white quilters. African American quilting is a revolutionary act of expression and persevering. Their threads are vehicles of the past and blueprints for connecting communities and envisioning new futures. The heart of quilt making is gathering and remembering. It is black women telling their own stories.
[00:45:05.287] Kent Bye: Yeah, I felt like that was kind of capturing the spirit of the piece, but also this archive that you're gathering. And yeah, I don't know if you have any other reflections on that quote.
[00:45:15.757] Tamara Shogaolu: Yeah, I mean, definitely. And I think, you know, when I worked on Unresolved, that's when I started kind of looking at it this way. And I don't know if you've heard of Gee's Bend, but it's a community in Alabama where I think for hundreds of years there's been like quilting has been the main source. So when you arrive in a community, you see like all these quilts hung everywhere and there's like an association of quilters and quilters. women and multiple generations of black women like go and quilt and this is what that town is known for and they made pretty iconic quilts and it's just been such a part of black sort of like archiving or like family tradition because even when you go back to times when people were enslaved and families were separated because like children were sold and people would tear parts of their own dress or like make scraps to like stitch together sort of way of remembrance and keeping a connection and that's why you have quilts that are passed down generation and generation that each one has like a piece of someone and people didn't have access to photos they didn't have access to like being with each other you know families were separated and this was a way of staying together So, I think that for me, it seemed like that was a form of archive and utilizing technology, I could add another layer to that and kind of imagining what a quilt today is in many ways.
[00:46:38.280] Kent Bye: Awesome. Well, as we start to wrap up, I'd love to hear what you think the ultimate potential of all these immersive and emerging technologies might be and what they might be able to enable.
[00:46:52.225] Tamara Shogaolu: I don't know. I hope they become part of society. I want to live in a world where we walk outside and people just know that there's always an AR experience somewhere around, that they can take a little break and go outside. explore the same way that somebody checks Instagram or checks Facebook that they can actually like engage with art at their own pace you know everyone always talks about democratization of tech and all of that and I'm like you know what does that actually mean and I think that means increasing access and I think now I've just been thinking a lot and with this project you know we partnered with the National Archives as well because like they have a museum and So the piece is going to be somewhere for a couple months, and I'm looking to partner with other museums in the States to have the work exist in a space for longer, to give a chance for more audiences. Because I think festivals are great for launching it, but, you know, you have to pay tickets, you need a badge. And I'm like, when is the artwork going to become part of everyday society? And that's my dream. I don't know if that's what's going to happen, but that's what I'm trying to work towards. And what I... what i hope happens and it's something that i think i'm thinking about more and more in my work like how can we have more impact more reach because i think that'll make our work better too like the more and more people interact with it the better we'll get at designing and making things so that people don't necessarily fear tech and can really engage right is there anything else that's left unsaid do you like to say to the broader immersive community Um, stay strong, stay alive. No, I mean, I don't know. I'm excited. I'm trying to be optimistic to be honest. And like, I think this is like, you know, there's so much to say so much to talk about now. So like, I just hope some good art comes from it. And like, I hope that we use our skills and abilities to hack stuff, to, to do something really, and like connect with one another and, and, and support each other. Yeah.
[00:48:54.904] Kent Bye: Awesome. Well, Tamara, thank you so much for joining me here on the podcast. I love that spirit of the hacking of these systems to kind of help nudge it in ways that we want to direct it towards the things we're trying to do. So just really great to be able to sit down and unpack so many different dimensions of this project and hear more about your journey in creating it. So thanks again for joining me here on the podcast.
[00:49:14.437] Tamara Shogaolu: Thank you. I mean, and you're the ultimate archivist of this space. So I, yeah, I'm excited. Maybe we can turn your stuff into a quilt.
[00:49:22.803] Kent Bye: There we go. Let's do it.
[00:49:25.625] Tamara Shogaolu: Thanks Kat.
[00:49:26.566] Kent Bye: Thanks for listening to this episode of the Voices of VR podcast. And there is a lot that's happening in the world today. And the one place that I find solace is in stories, whether that's a great movie and documentary or immersive storytelling. And I love going to these different conferences and festivals and seeing all the different work and talking to all the different artists and sharing that with the community. Because I think there's just so much to be learned from listening to someone's process to hear about what they want to tell a story about. And even if you don't have a chance to see it, just to have the opportunity to hear about a project that you might have missed or to learn about it. And so this is a part of my own creative process of capturing these stories and sharing it with a larger community. And if you find that valuable and want to sustain this oral history project that I've been doing for the last decade, then please do consider supporting me on Patreon at patreon.com slash voicesofvr. Every amount does indeed help sustain the work that I'm doing here, even if it's just $5 a month. That goes a long way for allowing me to continue to make these trips and to to ensure that I can see as much of the work as I can and to talk to as many of the artists as I can and to share that with the larger community. So you can support the podcast at patreon.com slash voices of VR. Thanks for listening.
