Alice Bucknell

In conversation with Sarah Johanna Theurer


Alice Bucknell’s work is research driven and perhaps best described as world building. World building is an active, ontological process of imagination. It’s a craft—and Alice Bucknell might call it a social technology.

SJT: How would you describe the worlds you are building? And how do they differ from traditional forms like science fiction thought experiments?

AB: I think it might be helpful to first distinguish worlding vs. worldbuilding practices. These terms are thrown around interchangeably to signify the same process, but for me, they are very different practices. Worldbuilding stems from a long history of creating narrative environments in which to tell a story. It’s closely associated with speculative fiction and its sub-genres: sci-fi, cli-fi, horror, fantasy, solarpunk, etc. These are typically closed systems: created top-down, often by a singular author, and they don’t do so well with change. As for worlding, I take a page from Donna Haraway as well as contemporary artists like Ian Cheng who see the process as much more collaborative, non-hierarchical, and open-ended. It’s less about creating any singular “one” world and more about an agile system of nested worlds that is constantly redesigning itself. Worlding often has ecological undertones, as it often deals with a multispecies cohort of co-creators, machine intelligence included. It’s also important to note that worlding as a tactic is increasingly enabled by technological advances such as real-time rendering processes and ongoing co-evolutions with AI.

SJT: The narratives that you develop are often modeled on real-life scenarios involving human and nonhuman entities.

AB: Both worlding and worldbuilding offer a really beautiful opportunity to imagine alternative narratives to the present and a scaffold for actualizing some of these systems in the future. So it’s especially important that we pay attention to the kinds of worlds being envisioned, the perspectives they draw on, and the voices that they are articulated by. Worlding is really helpful here as it does not default to any one perspective. Worlding looks both backwards and forwards in time in conceiving the kinds of worlds we actually want to inhabit. In my work I try to challenge the idea that humans are the only species capable of storytelling or engineering a world. I critically look at the impact of anthropocentrism upon the health and future survival of the planet we occupy along with many millions of lifeforms that have been here before us and hopefully will be here long after us.

SJT: Contemporary worldbuilding increasingly relies on computer-generated imagery, visual effects, and game engines. With the virtual image, simulation has become central to twenty-first-century representation in science as much as pop culture…

AB: It’s true that virtual environments have become the dominant aesthetic territory for worlding/worldbuilding practices today. Using real-time rendering technologies that enable the world to evolve as the player navigates it, destabilizes the closed-loop system of more conventional narrative environments. Introducing AI programs into the development of these worlds allows a sort of openness or porosity into how they might evolve and mutate over time, something that neither the player nor the game designer has total control over. In my work, I want to loosen the grip of humanity’s tighthold on storytelling, and create a space for other nonhuman voices and narrators to flourish. I am thinking for example of AI, not as an inert tool, but as an alive and evolving collaborator. It can also challenge so many western ideas of linear time, universally progressive capitalism and binary logic systems that have got us into this mess in the first place, even down to the idea of the “one” world.

SJT: For example, your work The Martian Word for World is Mother critically explores current narratives surrounding space exploration and the possibilities of human habitation of Mars. How do you understand the relationship of prediction and speculation here?

AB: Prediction tends to place us in a linear-track record for understanding (and profiting from) an imagined “singular” future, while speculation is for me a much more generative and generous approach, in that it explores multiple possible futures. In my work, the ratio between present mirror and future speculation varies from project to project. I often start from current issues that pitch into the future. For example, The Martian Word for World is Mother addresses the fantasy of terraforming Mars into a ‘Planet B’ for human habitation. The work is a three-channel video installation offering three different visions of human life on Mars; these scenarios are drawn from interviews I conducted with experts, from planetary habitability astronomers to space law researchers. This research allows me to build up the “realism” of the work, in order to distort it with the speculative elements I then introduce. I also often use real-world data like in Cones of Uncertainty. This video is narrated by a fictional weather prediction AI and looks at how the language surrounding extreme weather events evolved in parallel with hurricane forecasting technologies. It uses a custom language model built by the New Real Institute at the University of Edinburgh and predictive climate data modeling different possible “scenarios” from the Copernicus Climate Projections. The work challenges the idea that predictive data science and climate modeling offer us a clear looking glass into the future. The more projections we aggregate from increasingly larger pools of data, the less accurate these predictions actually become. This relates to Timothy Clark’s idea of “derangements of scale,” a struggle between human temporalities and the deep time of geological consciousness. Predictive modeling is our attempt to stretch our temporal intelligence and dilate our reading of this time-space system we call reality, but paradoxically it often closes down the possibility of a more nuanced and non-singular vision of the future.

SJD: So when you are relying on machine learning programs and game engines to build your worlds, how much does the medium itself speak to the concept of the worlds you built?

AB: In her text Gamer Theory, McKenzie Wark made the claim that we are all gamers now and that reality has been consumed by “gamespace.” Gamers or not, we are locked into what often feels like a slow apocalypse where we don’t know the rules. The rich symbolism of game engines offers an exhilarating space for worlding. Their language is also legible far beyond the art world; the idea that my work can permeate other areas of contemporary culture and reach new audiences is very appealing to me.

SJT: Game engine suites include built-in applications that allow the establishment of unique physics like lighting, textures, and other functions. How do you develop your own vocabulary of forms from this?

AB: That’s a good question—the way the camera tracks through a scene and many of the 3D objects I use are mostly based on pre-produced asset packs. I see elements, like a certain pack of palm trees, for example, that reappear throughout games or works by other artists. It creates an interesting parallel or unintended conversation between practices that I like. It’s really easy to make a landscape in a game engine; it’s really difficult to make a biome. There are two reasons behind this. Firstly, when you think about environments in video games, they often default to a binary: a pre-industrial, utopian, gorgeous rolling meadow with no humans and zero traces of ecological damage—think of something like The Legend of Zelda. On the other hand, it’s a dystopian scene—be it an abandoned city or burned-down forest of a survival game like The Last of Us. It’s very easy to imagine either a total apocalyptic wipeout or an escapist “wilderness”; there is very little attention paid to ecological realism, particularly with the destabilizing impacts of a heating planet. Secondly, there’s a lack of concern for portraying distinct species, whether it’s flora or fauna. If you wanted to create an accurate representation of a particular region, such as a desert landscape in Southern California, you’d be better off heading out to Joshua Tree yourself and 3D scanning in the landscape vs. searching the asset store. This is changing a bit, with organizations like the Harrington Lab at the University of Central Florida scanning entire sample ecosystems and botanically-correct plants that can be used in game world design. That being said, to capture and represent a biome perfectly feels both paranoid and sacrificial; like it’s tapping into the museological desire to document and preserve a specimen, so there’s less of an incentive to keep the species alive. It also has some colonial undertones. I don’t think it’s interesting to over-fetishize representational accuracy when it comes to video game ecology; rather, I like to superimpose. For instance, one of the levels in my latest project uses a 3D scan of the Hoover Dam, but inside this landscape dwells a species of moth and a species of tree that have not (yet) appeared in this landscape—although they are tracking northwards from the California desert due to the climate crisis.

SJT: You are currently working on your first open-world video game titled The Alluvials. Alluvials are young sediments deposited in river beds or flood plains; not fully formed, they are always on the move. The game is an attempt to portray this kind of landscape, based on the actual LA River, but is also about human-designed systems on Earth: political systems, economic structures, systems of belief. Can you speak a little bit about these borderlands of ecology and human-made systems and why you chose the open-world game to represent them?

AB: The Alluvials expands on the idea that the end of an anthropocentric worldview is not the end of the world, but a beginning for many other forms of life. It draws on both future climate projections and historical datasets that detail the condition of water in the LA region, using a combination of language and image-based AI tools to bring this data to life. The script was co-written with a custom language model trained on both historical and near-future “masterplans” for managing the city’s water crisis. It has a very elaborate cosmography that can be explored as an “open world”—one that exists besides the ongoing battle between the environment and economy, resource extraction and preservation. Through the co-construction of new worlds with human, nonhuman, and machine intelligence, the project underscores that nature is an intelligent system, a technology in its own right.

SJT: How did you compose the work formally?

AB: The Alluvials follows the idea of “storied matter” as presented by Serenella Iovino and Serpil Oppermann which sees matter—in a specifically ecological context—endowed with meaning. The project utilizes a variety of technological tools and interventions. The structure of the game expands and becomes looser from level to level, stripping itself of language while opening up a more abstract space for interpretation through a collaging of visual approaches. These range from game engine mods to the drone mapping I did of the LA River’s unpaved areas, transforming these topographic scans/point clouds into 3D assets. As the moving path of the river could not be translated into a 3D asset, the water is now a ghost, an absence, a haunted void in the model. I also use custom Stable Diffusion ‘hallucinations’: AI-generated videos using custom models trained on historical images of the river prior to its channelization in the 1930s with future-speculative renderings of the river and even image stills from my game.

SJT: Each of the seven chapters and game levels has its own sound scored by musician Ken Yama. What is the role of sound in the game?

AB: In The Alluvials, each of the seven chapters is narrated by a different nonhuman character, and each chapter has a different soundtrack that matches up with both the perspective of that character and the game component of the level. The score aims to develop a more dynamic form of engagement with the rhythms of the transformation of the landscape and the different tempos of the storytellers, the dynamics of their movement as they pass through the site. Sound is a super important component to all of my projects and a great example of the expressivity of the nonhuman that is nevertheless very accessible by humans.

SJT: But humans are not central to the world of The Alluvials. However, it is designed as a series of game environments and a game suggests interaction and a sort of agency that can be exercised. But within The Alluvials, this is not always the case. How can we imagine this?

AB: The Alluvials focuses on the slippery interplay between engineered ecosystems, nonhuman intelligence, and speculative financial systems that define the future of Los Angeles and its river. There is no pre-established body or individual to interact with. When you play it, you will navigate the same ecosystem in different bodies. Some bodies may have more of what we would describe as an “agency” over the landscape, while others have very little. Some of the game dynamics work counterintuitively to most of the video games you’ve probably played; for instance, with the final level, the wildfire level, the only way to “beat” it is to lose, over and over again, before the fire shares its story. Moving beyond the project’s game design, a broader challenge that I face quite a bit in my work is the question of how to best represent nonhuman intelligence within these narrative worlds as a human creator. While there is no perfect answer, a strategy I’ve used increasingly is to refuse total legibility in how that language is structured, how it operates, and what it’s saying: so both linguistic and paralinguistic architectures. Withholding pieces of information creates an effect of opacity that can challenge the typical assumption that human intelligence supersedes the intelligence of nonhuman life. It makes the humans in the room have to work that little bit harder to follow the narrative—which is sort of an impossible gambit, anyway.

SJT: I am thinking of Jacques Derrida’s famous lecture where he narrates his experience of shyness over his naked body in the gaze of his cat. There are also more recent experiments in speaking for nonhuman entities, like Henry Hoke’s Open Throat. In your experience, what are the challenges and what insights can we hope to gain?

AB: It starts with the question: What could nonhuman criteria even look like? I studied anthropology. Within the game, we experience what I would understand to be the biggest learning and at the same time the biggest failure of this discipline. We are part of complex assemblages which we cannot fully comprehend nor control without changing them in the process. One level of The Alluvials, the Hoover Dam world I mentioned earlier, sees the collision of two weirded and mutating biomes. Players switch between playing “tree” and “moth” as an embodied exercise in extreme symbiosis at the end of a world. To focus on environmental matter is not to exclude the human, but to decenter it, to reverse the figure/ground relationship that typically holds between individuals and worlds. Maybe Karen Barad’s term of “intra-action” (instead of inter-action) can be useful here. The idea is that we humans are not simply entities opposed to environments outside ourselves.

SJT: The boundaries between the natural and artificial are not so clear anymore. You spoke of The Alluvials as a “digital rewilding initiative.” Could we use technology—in the broadest sense—in a way that serves nature?

AB: Historically, nature has oscillated between being that which is beyond human control and that which is subject to such control. But nature and culture are actually indivisible. But, yes, in a literal way this refers to me hacking into the game engine of GTA V to depave the LA River and increase its biodiversity over the duration of the video component to The Alluvials. On another level, I like to complicate the idea of recursive and predictive models by using them incorrectly to an extent: feeding them contradictory input texts, or taking the output for one model and training another model on that output. I want to resist how these models are designed to operate. I also loved Alenda Y. Chang’s Playing Nature: Ecology in Video Games (2019) which explores games’ pedagogical functions in showing us the links between the natural and artificial spheres. Although the speculative worlds I create may seem extreme, versions of this are already being prototyped in other industries, from speculative real estate scams in Miami to privatized water companies creating synthetic water in California. There is for sure a Ballardian undertone to my work, in pointing out that reality is already so much crazier than anything one could possibly fictionalize.

ALICE BUCKNELL is a North American artist and writer based in London and Los Angeles. Working primarily through game engines and speculative fiction strategies, her work explores interconnections of architecture, ecology, magic, and machine and nonhuman intelligence. In 2021, she established New Mystics, an editorial platform merging magic and technology. She is a Somerset House Studios resident in London and a Supercollider SciArt Fellow in Los Angeles.

SARAH JOHANNA THEURER is a curator focusing on time-based art practices and techno-social entanglements. She works at Haus der Kunst München where she has commissioned numerous performances, spearheaded new productions and co-curated major retrospectives. She sometimes acts as a dramaturg and frequently publishes in catalogues and magazines.


All images:
The Alluvials, 2023 (stills)
Courtesy: the artist