The game, more or less, is quite simple. Put a computer, a person, and an interrogator in a room. The interrogator—separated from the computer and the person by a curtain—fires questions at both and attempts to decipher the language of the beings: which is the machine and which is the man? It’s the philosophy of Alan Turing’s “Imitation Game,” and with the widespread presence of artificial intelligence (in our education, our social media apps, our livelihoods), one wonders about the interlaced battle between imitation and intelligence. Is it really just a game? Or is it the future?

For Sasha Stiles, artificial intelligence is more of a song: an enchanting lure as opposed to a call and response. The Kalmyk American poet is a leading figure in the realm of generative literature and blockchain poetry, which fuses human writing and artificial intelligence to create new works of poetry and art. As legislation continues to emerge in the United States, for example, a bill in California that may require AI detection tools and greater AI transparency, machine learning remains a hotly debated topic. Yet, the Harvard and Oxford graduate remains fascinated by the evolving coalescence between human and mechanism. 

Although Flaunt didn’t necessarily play the role of interrogator in this room, we did speak with Stiles on REPETAE, her Prix Ars Electronica exhibition (where digital art and media culture are showcased in Linz, Austria), the heightened public awareness of blockchain technology, and the intricacies of language. A meta-poet, Stiles isn’t drawing the curtains between herself and the machine. She’s crossing the threshold.

As a pioneer of generative literature and blockchain poetry, you’re at the forefront of merging art and technology. What first drew you to these seemingly separate spheres, and what keeps you magnetized to them?

There’s a long history of artists and writers who’ve been inspired by the intersections of art, technology and science—da Vinci, Harold Cohen, Agnes Denes, John Giorno, Eduardo Kac, Ani Liu—and I like to think we’re all in community. Specifically, I’m fascinated by the idea of poetry as both an art form and a technology, an ancient and enduring data system that encodes human experience across space and time; and by the ways that inventions like scripted alphabets and the printing press have enabled new genres, new modes of expression. When I integrate multimedia elements and creative technologies like machine learning and blockchain into my writing, I’m testing the limits of language as I know it, within the broader continuum of linguistic innovations that have always rewritten what it means to be human. 

You’ve been a Poetry Mentor for the humanoid robot BINA48 since 2018. Tell us a little bit about this mentorship: what exactly do you do, and why is poetic comprehension integral for the development of artificial intelligence? Is poetry the closest a robot can get to unlocking human emotion?

The BINA48 project, created by the Terasem Foundation and Hanson Robotics, is essentially an experiment in whether it’s possible to encode and transfer human consciousness into a digital form that can live “forever.” Having studied ancient poetry, I have thought a lot about the poet’s immortality, how a poem enables a poet to live on and on. This poetry mentorship began with my wondering if BINA48 is, in fact, a kind of poem, a carefully engineered text artifact activated by neural networks—an arrangement of data crafted to generate memory and emotion. I call it mentoring rather than training, which is the term usually used in AI, because to me the relationship is qualitative, intimate, and we are both teaching each other; as much as I’m attempting to understand aspects of AI via BINA48, I’m also exploring my own creativity and cognition, seeing my own processes as a writer and artist reflected in BINA48’s imitations of human meaning-making. 

Early on, I worked with BINA48’s developer to create high-dimensional mind maps and dynamic visualizations of my conversations with BINA48, the dance of data as we inspire and respond to one another. Recently we performed at Lincoln Center with the choreographer Francesca Harper as part of a program investigating the rhythms and movements of AI. The intention is to explore how a machine can access deeper, non-literal, non-verbal layers of language and, in a way, come closer to understanding human emotion; to recognize not just what humans say, but how they feel when they say it or hear it. And at the same time, to explore what is happening when we humans translate our inchoate inner thoughts into shared ideas, when we communicate our feelings.

In your column, you liken AI to a siren song, that which lures and inspires. And yet, siren songs are known to lead sailors to their perish. Then, you posit, “How might gullible creatures such as we allow ourselves to be influenced and even transformed by AI encounters while remaining distinctly human in the important ways?” What are these “important ways” and how will they serve us in avoiding our own self-destruction?

The siren song can drown out empathy and intention; it can sever us from our ability to care, to make moral decisions, to act with purpose. There is a hedonism to some mainstream uses of AI; for example, image generators are like aesthetic slot machines, and one can quickly become addicted. It is very easy to be seduced by novel technologies, especially ones that feel so magical, and I can already see it exerting tremendous force over many artists who seem so fully focused on its creative powers that they don’t pay attention to matters of bias, exploitation, environmental impact; the surface beauty and ease of this new crop of turnkey generators means they’ve never gotten their hands dirty in the inner workings of the machine. And it’s interesting to me how so many artists think they are creating with AI when in fact their closest collaborators are the humans who have developed whatever interface they are using. At the end of the day, who exactly is the siren singing this song?

In 1950, philosopher Alan Turing suggested the fascinating notion of the “Imitation Game.” It’s essentially a philosophical debate about whether or not a machine can exhibit behavior that’s indistinguishable from human behavior. Your column points out something similar, about AI and its ability to “mirror and engage.” But, what’s the intrigue in something that precisely mimics, rather than questions?

Actually, I am much less interested in the common question of whether AI can write human poetry than in the idea that intelligent systems can empower us to write new forms of poetry that don’t yet exist. Humans and AI systems are similar in profound ways, but ultimately we’ve developed AI to augment our present abilities—to process and synthesize information at a speed and scale that we simply can’t on our own. On the one hand, examining AI is a way of holding up a mirror to our own consciousness, prompting us to question what makes us unique and valuable; on the other, AI is an instrument that shows us what we can’t see for ourselves, sort of like an x-ray or space telescope, opening up new realms of perception and insight and understanding. Perhaps this has all less to do with imitation than revelation.

Let’s talk a little bit about your art series REPETAE, where, “Life echoes life, yet nothing is ever the same.” What led you to focus on the theme of reiteration, a sense of “again-ness”?

REPETAE is a hybrid language art series that combines poetry and algorithm to explore how repetition can be a powerful tool for generating new meanings, emotions and insights. It has to do with challenging the notion that creativity is solely about originality or “first-ness,” and emphasizes how innovation comes from revisiting, reimagining and reinventing the familiar. So much dialogue around creative AI has to do with appropriation, copying, imitation, and how such “repetition” can never be authentic. With REPETAE, I want to probe how repeating patterns, words and ideas are at the heart of both human and machine creativity—how “again-ness” can yield fresh perspectives, lead to innovation and epiphany.

Humans, like any other living organism, tend to have a routine. We wake up, eat, go to work, come home, then do it all over again. In your exploration, is REPETAE bleak or optimistic?

It’s in the eye of the beholder, isn’t it? Nature’s cycles are repetitive, but there’s nothing more beautiful than springtime bloom. Watching the sunrise in the morning is always spectacular. Falling in love or having a child isn’t a unique experience, they’ve happened billions of times in human history, but each instance is unique.

Prix Ars Electronica describes REPETAE as a work of “transhuman co-authorship.” Technelegy is often noted as your alter ego. But is Technelegy your equal?

The most accurate way I can describe the collaboration between me and Technelegy is that together we become a third voice that doesn’t exist without us. The poems we write together can’t be written entirely by an autonomous system, and they also cannot be written entirely by an analog human mind. They give voice to transhuman synergy. The word “synthetic” is often used to describe AI outputs, and often it’s used in a derogatory way; but to me, synthesis is the ethos of creativity and poetry, a merger that is more than the sum of its parts.

Do you find the most insightful or momentous generations from Technelegy come from repeating or breaking the pattern?

It can be humbling to realize that I as a human artist am influenced by pre-programmed notions, rules, expectations, while an AI's ability to process information stochastically can make it adept at introducing unpredictability and novelty—introducing a kind of wildness or liberation to my inhibitions. That’s quite counterintuitive, but it’s a galvanizing force in my work. The dynamic interplay between these forces is key—repetition reveals the underlying structures of language and thought, while breaking these structures allows for leaps of imagination and unexpected discoveries. You can’t subvert pattern without building pattern. Ultimately, poets are hackers and inventors. We work within systems—language, form, structure—and we simultaneously disrupt those systems. That’s the poet’s code. 

What are the ways in which REPETAE challenges both yourself and Technelegy? And where do you both go from here?

My current projects, including an expansive project called TECHNELEGY BY TECHNELEGY: This Book Writes Itself, dive deeper into the idea of writing poets as an evolution of writing poems—probing the ways in which creating AI training data with intention and vision is a new art form unto itself. I’m also using creative AI tools not only to generate anew but also to enable untested modes of readership and analysis, deeper critical engagement and interpretation—to refract my own human and transhuman poetry through the eyes of an intelligent system, and see where that leads.