Connie Bakshi’s digital shamanism
This is an edited transcript of an interview. You can read a condensed version here.
Q
My background is anthropology and I’ve done some work around spirituality and AI. So when I read that you said you’re descended from ancestral shamans of Taiwan, let’s start there — tell me about that.
A
My parents were both born and raised in Taiwan, and descended from the indigenous peoples of Taiwan. And as you may or may not know, Taiwan has a history of colonization and imperialism. It’s something of a badge of honor to be able to say that you’re descended from the indigenous peoples.
Honestly though, growing up, I didn’t really have a lot of interest in my family’s past, because my parents had fully embraced the American dream when they emigrated to the US, and they wanted to make sure that their kids were fully embedded in American culture. So there wasn’t much talk about our ancestry.
It wasn’t until 2021 when I started to do more of a deep dive into my family history. The trigger was my grandmother passing. The thing about our family — and the Taiwanese culture in general — is that it tends to be matriarchal. My grandmother lived with my family throughout my childhood as the unspoken head of household.
So through her, a lot of our cultural practices were inherently maintained. But when she passed, my family felt shattered. I remember thinking, “There’s so much more to our heritage and history than I ever realized.” That was a massive paradigm shift for my family.
As a family, we had to grapple with how she was the remaining connection to our past. With her passing, so much of this was lost. All of these memories and stories that we hadn’t always discussed with her. So we started to get together as a family, and we talked more about her — her childhood, and how she grew up in Taiwan under Chinese and Japanese occupations, for example.
But also her personal ancestry. When I talk about being descended from shamans — her family, before she had converted to Christianity much later in life, they were very much into what we would call in the West, the mystic arts or shamanism, or the Voodoo practices of Taiwan.
If I remember correctly, her own grandmother was actually a “medicine woman”, which in those days, was the polite term for a female shaman. My dad shared stories that my grandmother had passed on to him, about her uncle, another Voodoo practitioner, and doing these incredible feats of what seemed like magic to her.
She was witness to all of this as a child, and through her young adulthood. Because of this, she was initially resistant to converting to Christianity upon marrying into my grandfather’s family, so there were a lot of interesting dichotomies and tensions going on at the nexus of my grandmother’s life, in the context of our full family lineage.
This was a start for me to wrap my head around my origins, but I think there’s much more to be understood about how spirituality existed in my family’s life for centuries.
So you went on to study like biosciences or something, right? Which is completely different.
I was a biomedical engineer. I often say that I’ve lived many lives. My parents are kind of your traditional eastern “tiger” parents, as we would call them. So they would push their children to become doctors, lawyers or engineers. And my dad was set on me becoming an engineer in the hopes that I would become a doctor.
But I was interested in engineering for more philosophical reasons. I grew up in the age of the Gulf War, when there were all these documentaries about veterans coming back from the war with physical trauma, and going through physical and emotional rehabilitation. This got me thinking about prosthetics.
I was especially fascinated by the phantom limb experience — where in the case of amputation, you still feel the pain and sensations in that limb. I was fascinated with the idea of how that might translate through biomedical technologies.
What does it take to actually create the synergies between organic biological systems and more synthetic ones? Is it possible to even have a seamless transition? And what does it mean for the truth and intent of movement when you’re kind of a cyborg? I’ve inherently pulled into my artistic practice with AI.
.
Do you reconcile the bioengineering and the shamanism, or do you see them as completely opposed? Have your views changed?
Absolutely. What’s fascinated me about shamanism and the spirituality of my ancestry is more about planes of reality, planes of existence. In the Western mode of thinking, reality is what exists around us, right? But in my ancestral practices, there was always this understanding that there is more than one plane of existence, all of which are equally real and equally true.. They coincide and they may coexist in the same space, if not actually interact with one another.
When I translate that into my work with AI, which is digitally native, there is a phantom limb translation where you’re thinking, “ Okay, we’re living in what we consider a physical space. We’re working with increasingly physical digital technologies, and yet there’s a divide”. More often than not, there’s a mediating screen between these planes of existence.
But is there? Are they truly separate, or are there moments where they actually coincide and exist together as a reality?
.
So would you say you have become more spiritual over the years? Would you call yourself a spiritual person?
That’s a tough one because I have a tough time with labels, especially as spirituality comes with a lot of connotations, and it’s a colloquialism in and of itself. But I certainly have a more metaphysical view of the world around me.
I respect empirical science and the rationality that have been socially revered for the past several centuries. But I also recognize that there is so much that we don’t know, and I try to leave space for that.
.
I guess that manifests in the work.
It does, and in different ways, because it’s an inquiry about the nature of reality. What are the connections between these different planes of existence? When you look at my body of work, you might see different subsets, themes and aesthetics emerging. The questions that I pursue in one project may evolve into the questions for another project — because reality is a complex thing woven together by facets of knowledge, values, and truth
Let’s look at Midnight in the Garden of Good and Evil. It says in your description, “drawing from lore and Orthodoxy across Eastern and Western divides”.
Yes, that one’s part of a larger world called Xenogenesis — a genesis story of a world that exists in a future that’s long past humanity, in which the machine is creating in its own image.
.
So humans are extinct.
That’s the general framework. But there is a residue that remains, in terms of data fragments — pieces of digital DNA and identity that have eroded across hardware and software over time — that this intelligence is attempting to synthesize and reconstitute on its own terms and through its own understanding.
It’s akin to how we exist now. I think about my own family history and the loss of memory, and how it now exists only in fragments. It can only be reconstituted based on accounts and family conversations, and interpreted through the values we choose moving forward.
At the same time, for an AI to kind of create in its own image — I think that’s an area that’s rich for me to explore, because what does it mean? What is an AI’s own image? AI was created through an imitation game in relation to the human, right? Can we imagine that perhaps this AI has incorporated other non-anthropocentric intelligences, to the point where it can now look across this fragmented dataset and try to piece together something that’s more?
We talk about data biasin AI all the time. In a lot of ways, I don’t lean into techno-optimism, but I’d like to speculate about a world in which maybe there’s an AI model or set of models that perhaps do not have that kind of weight distribution in their data.
.
Okay. Shifting gears a little bit, even before that you studied piano, right?
I did. Like the first 20 years of my life. I was a concert pianist, doing competitive circuits from the age of nine.
.
I was talking to someone else trained as a classical pianist, who is now doing AI stuff. And I’m wondering whether it influences how you think about creativity and working with AI. Is AI more of a tool, or more of a collaborator for you? Do you see it as creative itself?
Let’s start with the piano, because I think I had a very sensitive and intuitive relationship with the piano. For those who aren’t familiar with pianos, every single one responds and reacts differently to touch.
When I’m playing, there’s a feedback loop that occurs between you and the instrument where you become one entity. It feels like this interaction of collaborators who are conversing and debating with one another. That’s in the context of trying to understand and interpret a composition. I’m providing my emotional intent and I’m hearing how the piano responds, and then perhaps adjusting my own touch and my own intent to see how the piano’s going to respond in kind, through a very recursive process.
Ultimately, when I go to perform this piece, this feedback loop and these responses between us have been codified into my muscle memory, so that now it’s like a collaborative output, the performance.
But what’s interesting to me is that it doesn’t end there, because there’s a feedback loop with the audience too. The stress and the energy in the performance hall. You’re hearing your music resonate throughout the hall. It’s not perfectly tangible, but you feel the energy of the audience and how they’re responding.
So there are a lot of intuitive cues that you have to feel in the same way. I think it’s the same with AI. Because every single model is different, and there are anomalies and idiosyncrasies — not only to the model itself, but in terms of how I’m interacting with it. Each model fosters a different flavor of collaboration and performance.
Whether I’m training a model and I’m adjusting the distributions and parameters, or I’m working with a generative AI model. And even just in the way I’m using language when interacting with each model. It creates a very, very different response that I’m ultimately sculpting, then fine tuning through a recursive process.
And of course, with the piano, you’re creating something that’s completely intangible, some sort of magic, you know, emotional. Like you say, it’s almost like creating another level of existence or something. I guess the question is, do you see AI doing the same thing? I can sort of see it in your work, but interested to hear what you think.
”World building” has almost become kind of a cliche of a term in the past couple of years, but as artists we’re world builders. Whether I’m playing music or creating art through AI, I’m hoping for that moment of connection, of resonance across a divide. And it’s so ephemeral. It’s not something that can be set in stone, or built brick by brick. In the same way, I think that these planes of existence that are separated, have these moments of coincidence, which are very rare and fleeting, but magical when you evoke them.
.
I guess the main way it manifests in your work through the spatial design, from your training at Art Centre, right?
That’s informed a lot of the way I approach scaled installations. You always want to be conscious of how it lives in the space with the audience. And I think my experience as a spatial designer has informed how I think about sequence, too. And all the sensory cues in a larger space.
At KanvasDUBAI, where Midnight in the Garden debuted, that was a four-channel installation — three massive walls and the floor. Everything kind of flowed together, and I was trying to time movement and undulations. And also choreograph the music in such a way that the audience was drawn into these characters and how they interacted.
Because of the scale of it, it was fun to be able to create the little moments of movement that you might not see when you look for them directly, but you might notice them just out of the corner of your eye.
.
Something I read in another interview you gave, you said you’re questioning how AI might help us access and express experiences that go undocumented, or exist in the spaces between our conventional classifications. I think that is referring to the other planes of existence.
When I’m looking at these in-between spaces, this is where I have different approaches or different topics by which I’m trying to poke at the idea — perhaps from a broader context. Let’s consider the example of mythology. In Garden of Good and Evil, I was fascinated with three particular elements: the woman, the serpent and the choice. Maybe when you first hear those terms, your mind might automatically go to the Garden of Eden.
But when I started doing a lot more research, looking across world cultures, especially in Eastern cultures, I was seeing the same exact elements, but very, very different settings and outcomes. For instance, I looked heavily into Nüwa, who is a Chinese serpentine goddess who created the world.
I was cross-referencing that against Manasa in Bangladesh. Then I look at the Abrahamic creation narratives that split the woman from the snake and created this binary of good evil. And the woman becomes the gateway to sin.
So this is where I really wanted to understand this in-between space. What are the synergies that might create? Almost a mythology or story that predates all of them, and AI became a method by which I might create an existential coincidence.
.
I’m not a biblical scholar, but an anthropologist [Mary Catherine Bateson] said something about how in the Christian religion, God created things and things began to be named and separated — this binary, rational classification started there. But then Jesus comes along, and he starts saying, “No, everything is connected, not separate.” He disrupts things. And you see what happened to him!
It’s fascinating to look at the Bible as both an anthropological, cultural and historical document, because you’re seeing these interesting paradigm shifts happen within this document. Granted, it’s gone through many revisions — but that’s a topic for another conversation!
In another of your interviews, you use the term “manifold destiny”. Can you say something about what that means?
Manifold destiny to me is a kind of reaction and response to the manifest destiny politics and social dynamics that have really fostered and created the false (to me) American dream. It’s about ongoing expansion and possession of territory.
Manifold destiny is a reversal of that movement to a kind of implosive introspection. Where can we, deeply within ourselves in an introspective way, have an ever-expanding understanding of who we are as individuals, as a collective, as a species?
And to me, when I started to work with AI that was my idea about it. What could AI be for me? If I’m looking at these multiple planes of existence — let’s say they’re parallel. Could AI be one of those mechanisms by which I could start poking holes between them and creating connectivity? And this connectivity could actually become the substrate of reality.
.
Tell me if you agree with this: I’ve come to believe that the most important thing about AI is not all of the stuff that it’s doing — solving problems, making things. But how it’s causing us to question a lot of things about how we’re human, how we run society and things like that. We have to rethink governance and creativity and all these basic concepts. Would you agree with that?
At the base level, it is forcing us to question a lot of these aspects of our lives. But I’m also seeing a propagation of, and perpetuation of, colonial hierarchies and histories in the usage of and — even in the governance and creation of AI.
I think a lot of the larger conversations around the technology are valid, but I think also what I’m not seeing enough of — with the exception of certain artists who are working with AI very intentionally — is this idea of questioning the interactions and feedback loop with an AI model. What is it? What kind of reactions is it actually getting from the artist or the user? That’s a very different thing.
You know these moments of, “Oh, wait. I hadn’t thought of it that way.” There’s still a certain novelty to the outputs. Sure it’s fun to chase the high fidelity and technological tip of the spear, but is it really helping to evolve our thinking and our own intelligence?
.
Good point. Would you consider AI to be an artist in any way by itself?
There are so many different AI models. There are models that speak to interesting means of creativity. Obviously Botto — really incredible things coming out of that community. I just heard of an agentic AI artist that’s enrolled as an art student — Flynn, by the artist duo Malpractice. Their artwork was just released as part of The Second Guess exhibition that I’m part of. As a student, it’s trained to resist patriarchal politics. A model that creates questions with every interaction. That, I think, is a very creative and artistic framework in and of itself.
I’ve certainly seen AI models, or human-AI collaborative efforts, that are provocative and insightful. Sougwen Chung is one who I think makes really, really beautiful work that evokes, once again, the question of, “How much of this is human? How much is AI?” It can be both, right?
.
I was going to ask something similar to the student thing, which is: AI as an audience. On one level, there’s social media, and all of the content that we’re creating is for AI as an audience, and it’s hoovering up all of that stuff. But could you imagine making work explicitly for AI? You do train your own models sometimes. So in a way, I guess you do sort of make work for AI, in that sense.
I’m definitely aware of the algorithmic gaze that we experience with AI and social media. You see this pop up in social media and now as specific agents, in which Twitter accounts have been specially trained. I’m always fascinated when my art comes across these accounts. What they have to say, sure, but more so, why they are responding the way they do, and what, if any, codifications reinforce these responses.
Say something about your practical working process. I think people will want to know what sort of process you follow.
I think you know that every process is different depending on the project, but I definitely use a mix of tools, whether I bring my datasets across different generations of AI models, or incorporate analog methods. The earliest AI model I worked with was actually a custom diffusion model. I trained it, and I wasn’t trying to create artwork in particular. I was just trying to understand the nature of light and shadow, because at the time I wanted to generate my own dataset around light and shadow. It’s actually very critical to the aesthetic or the feel of my work.
I’m always fascinated by spiritual and religious art and narratives. And what always struck me about iconography and religious art across world cultures was that line between light and shadow. It was really defining a binary divide between the exiled and the exalted.
.
Good and evil again.
Yeah. And it’s such a strong binary that’s embedded within us. So when I was working with these diffusion models, I was playing around with textures, trying to get a very specific light quality and ambiguity of surface.
These were my earliest studies, which became one of the main datasets that I would bring into later generation models. And I think that’s why my work is pretty visually consistent — because of that light and shadow interplay that I was able to attain early on.
.
Okay. And the human form came in a bit later I guess.
It did. But I was much more interested in abstractions of the human form. Early on, I wasn’t interested in faces, because even then, if you were to bring any sort of dataset into one of those commercial models, it would default to something very figurative and very specific — and biased. And I was like, “No, no, no”. I want to actually understand the nature of flesh, and the corporeal.
And so that’s how I started. Is there a world in which, does this now manifest as kind of an interesting synergy between the organic and synthetic, or the human and the nonhuman?
Your work, if I’m right, has been primarily digital on the screen. But being trained in spatial experience design, would you want to expand into sculptural or physical works?
The compulsion is definitely there.I kind of dipped my toe in. It was a year or two ago — a collaboration with a poet named Ronnie Angel Pope and theVERSEverse, which is a collective of poets and artists who argue that the poem is a work of art, and they’ve created quite a movement in Web3 in the art world. And they’re dear friends. Approaching AI with Pope’s poetic language, I created sculptures that were a foray into the physical translation of the digital.
.
Looking at your website, you seem to work in different series or themes. Are there ongoing themes that you’re working with?
The themes inevitably cross over and influence one another, but my website presents the major buckets as I see them at the moment. I will say, the nuances and idiosyncrasies of language were central to my work early on and continue to influence my work — particularly in considering the engine of AI, and how language is interpreted and then manifested in that world.
A later grouping is around this world of XENOGENESIS, which I think will take me a long time to structure and develop. Even here, language shapes the development of the artwork and narrative
For example, if I’m thinking about the serpent, the woman, the choice. There’s wordplay going on in the back of my mind — thinking about the mores of myth, the moire effect, thinking about Moray eels, mora of verse — and how these “mores” coincide. That’s how my process works — there are overlaps at different points of development that create a forking path from an initial theme. I keep it all kind of open-ended. I don’t recall the exact quote, but I think it was Anicka Yi who noted that the overarching thesis of an artist is a thesis that might not be fully formed until the end of one’s artistic journey or lifetime, even. I hold by this. All I can do for now is chase the ever-evolving question.
.
You mentioned Sougwen and a couple of other people. Do you have other favorite artists these days, people whose work you’re looking at?
That’s a dangerous question! I can’t play favorites. I have too many dear friends in the space who are all so inspiring and influential on my thinking. Sougwen is certainly one who inspires — thinking about her early work with D.O.U.G. and her approach to using data as material in her latest series. There’s also Gretchen Andrew — very different approach — who considers the algorithmic gaze and re-appropriates it in her FaceTune portraits. I often think about Peter Wu+, who creates and builds provocative worlds around deep curatorial narratives. These are brilliant artists who are harnessing algorithmic and computing tools, yet are able to channel their own artistic voices.
So many others I could name. The community of artists that I’ve built around myself are the people who I lean on to challenge my thinking and process. There’s a bi-directionality in how we support one another conceptually — but even more so as peers and friends. It’s an uncomfortable passion, being an artist, but it’s much more bearable and joyful when you’re surrounded by people who light your fire.
