Poetry was humanity’s first language technology. AI is next
Sasha Styles turns GPT-2 experiments into a self-writing poem in a Museum of Modern Art installation – and a new way to think about text-generating AI optimization
A photo of a layered, animated screen from A Living Poem, where lines of text unfold and overlap in real time as part of Sasha Styles’ AI-assisted poetry performance.
Poetry and artificial intelligence may seem like opposites—one deeply human; The second is cold and mechanical. Sasha Styles sees them as an expression of that same impulse. The Kalmyk-American poet argues that poetry, “one of our most ancient and enduring technologies,” invented a system of meter and rhyme to store important information. She considers AI as its natural successor.
Styles’ path to AI began with literature, not code. But science was never far away: His parents are documentary filmmakers who worked with Carl Sagan on the original universe series, and she grew up traveling with scientists and philosophers as they interviewed them. She came of age with the Internet and realized how it shaped the way she thought and wrote. When she encountered the technology underpinning modern AI in 2019, she didn’t just want to write about it – she wanted to write with it.
scientific American spoke to Styles about why language may be the defining medium of the AI moment.
On supporting science journalism
If you enjoyed this article, consider supporting our award-winning journalism Subscribing By purchasing a subscription, you are helping ensure a future of impactful stories about the discoveries and ideas shaping our world today.
(An edited transcript of the interview follows.)
How did you create art at the intersection of poetry and AI?
In 2017 I read about Transformer-based architectures that drive natural language processing, and something clicked. That was the moment I realized I didn’t just want to write around the idea — I wanted to understand what it felt like to write using those models, using AI as a tool, as a collaborator, and as a medium.
I started doing research and looking at who was using these early versions, who was doing interesting things with the language. This was long before ChatGPT. I was reading the work of people like (data poet, artist and creative technologist) Ross Goodwin and (computer programmer, poet and game designer) Alison Parrish, who were building poetry bots on Twitter. And I tried to figure out how someone with a humanities background could take baby steps into that world.
In this moment of A Living Poem, the generative text installation displays a grid of evolving poetic gestures that explore reflection, action, and the status of language as a living system.
What did those early sessions look like?
I started working with GPT-2 in 2019, taking lines of my own poetry and inserting them to see what would happen if I asked a language model to take an idea I had and run with it.
One of the lines from one of the first poems I wrote was “Are you ready for the future?” Came from writing. Repeatedly varying parameters in the same system to see how the output will change. It was not intended to be poetry – merely research. But I found it really interesting and I compiled 30 of those hundreds of outputs into a small poetry cycle. The results ranged from the very sublime and beautiful to the very misogynistic or obscene – really looking at the spectrum of what you were capable of outputting in that moment.
How did you go from feeding lines to building something trained on your voice into a simple model?
I took a manuscript that I had worked on quite a bit at the time – 200 pages of poems – and put it into a dataset to create a fine-tuned version of GPT-2.
So I had a system that actually had knowledge of my own writing – not just (knowledge of) the canonical poetry already in the archives but also knowledge of my style, my vernacular, my thematic areas of attraction. I can use lines of my own poetry as input, knowing that the system knows how I’ve already written that poem.
That process eventually proceeded a living poem At the Museum of Modern Art in New York City. How does that piece work?
I think of a living poem As a living language system – an hour-long script for an unchanging, constantly evolving poem in which code sketches, datasets, accelerated architectures and human influence come together to execute a real-time loop of poetry, visuals and voice. It’s essentially an environment in which I can think about language and let language think for me while making that process tangible.
I have long been attracted to metapoetics and generative or automated writing in its many forms. One of my earliest encounters with computational text art was house of dust (1967), by the (late) Fluxus poet Alison Knowles – an early computer-generated poem. And contemporary language artists like (Jenny) Holzer, (Edward) Ruscha and (Barbara) Kruger have been formative for me.
a living poem Implicit in this lineage are the technological and cultural conditions of the 20th century – mass media, broadcast culture, industrial printing, early computing – that shaped modern text art, which in turn shaped me. At the same time, it is a space where I can experiment with new modes of expression emerging from the techno-cultural conditions of the 21st century: language as a living, productive field where meaning is created at speed and scale through repetition, plausibility, multiplicity, networked imagination.
In this scene from a lively poem, stark white text against a dark blue field announces a line of poetry.
You have described poetry itself as a technique. What do you mean by this?
Many people think poetry and technology are contradictory, but I find them relevant. Poetry is not just an art form or decorative language. Humans invented poetic language before the alphabet was written because we needed a means to store, preserve, and transmit information through generations. We invented meter, rhythm, and rhyme so that we could remember really important human data. Poetry is one of our most ancient and enduring technologies – a very fundamental data storage system.
Does this change the way you look at AI?
Looking at AI through the lens of poetry is a way of saying that there is something very human about the fundamental impulses behind technologies like AI. I think of poetry as our oldest form of mixed intelligence, a way of intertwining algorithms and emotions, like our new technologies.
If we can accept that these technologies have enabled self-awareness, consciousness, and our ability to articulate inner worlds – then perhaps this is now useful for the conversation around artificial intelligence. Perhaps these tools can take us to new realms of consciousness, just as poetry has enabled us to do for thousands of years.
