Surrealists believed in the facility of dreams. Inspired by Freud's theories of the unconscious and dream work, André Breton saw the irrationality of dreams as a creative method that might reveal latest, revolutionary ways of being.
A century later, the meaning of dreaming extends beyond the unconscious to the disembodied processes of machine systems.
Dreams have develop into a metaphor for the way artificial intelligence metabolizes and produces information.Backup slope“.
It isn’t any longer possible to exit AI systems. Each digitized trace will eventually resurface as an artificial output. As AI conjures us from our information, its neural processes exceed our understanding – and reinforce our biases within the service of capitalism.
Data Dreams on the Museum of Contemporary Art brings together artists who make these tensions visible.
AI and the “social dream”
Hito Steyerl argues that generative AI “medium images“: Statistical averages derived from training data sets. These are “social dreams without sleep” that reflect what society pays attention to.
Courtesy of the artist, Andrew Kreps Gallery, New York and Esther Schipper, Berlin/Paris/Seoul, stills © Hito Steyerl / © VG Bild-Kunst, Bonn 2025
Steyerl's video installation Mechanical Kurds shows footage from the Domiz refugee camp in Iraq. She accompanies Kurdish refugee clickworkers as they annotate data for the war systems they monitor.
Sculptural supports within the gallery space reflect the bounding frames that appear on the screen. They are subtle reminders that we’re enmeshed in these infrastructures.
Lynn Hershman Leeson's video works explore the political interests of AI. In Logic Paralyzes the Heart and Cyborgian Rhapsody, their cyborg protagonists move between problems with military surveillance, facial recognition systems, machine hallucinations and the rapid obsolescence of technology.
By staging dialogues between humans and machine doubles, she makes visible the instability of human agency when machines are involved in our social interactions and artistic lives.
Courtesy of the artist; Altman Siegel, San Francisco; and Hoffman Donahue, Los Angeles and New York. © Hotwire Productions LLC
Trevor Paglen's series of AI-generated images, Adversarially Evolved Hallucinations, explores the instability of perception in computer vision.
The images are created using early machine learning algorithms, so-called Goose and are printed using photographic processes.
Paglen suggests that these “hallucinations” lie on the sting of what’s classifiable and visual. This insight into how algorithms “see” shapes our vision.
Image courtesy of the artist and Museum of Contemporary Art Australia © the artist, photo: Hamish McIntosh
Kate Crawford and Vladan Joler's Anatomy of an AI System maps the social and material infrastructures of an Amazon Echo.
Your diagram traces all the life cycle of the machine in addition to disassembled components and raw minerals. The work shows how deeply AI is embedded in ecological and economic systems, competing for the resources that sustain life – only to ultimately be discarded as toxic Electronic waste.
Image courtesy of the artists and the Museum of Contemporary Art Australia © the artists, photo: Hamish McIntosh
Material ecologies
We are invited to think about whether AI could help us understand what intelligence is and what it would constitute generative System.
Courtesy of the artist and Marian Goodman Gallery. Photo: Elisabeth Bernstein
Agnieszka Kurant's sculptures and paintings suggest that intelligence shouldn’t be a purely human concept. She works with biologists to uncover the origins of life through calculations. Their temperature-sensitive, changing copper color fields reply to human emotions expressed on social media.
Her works are wild of their premise and execution. They prompt us to grasp how machines and minerals expand our view of the world beyond us.
Image courtesy of the artist and Museum of Contemporary Art Australia © the artist, photo: Jacquie Manning
Anicka Yi's jellyfish-like organisms are a part of a kinetic sculpture installation and video piece created by an AI model trained on her previous work. These forms tackle a form of speculative biology while visualizing the earliest life forms on Earth.
This is reflected in Angie Abdilla's video installation, which uses machine learning to visualise Big Bang narratives alongside Aboriginal creation stories.
Image courtesy of the artist and Museum of Contemporary Art Australia © the artist, photo: Hamish McIntosh
Dreams of speculative time
This exhibition is broadly concerning the way AI is used to visualise history, predict the long run, and query who has the fitting to achieve this.
Fabian Giraud's meditative video piece, The wild oneuses AI as an algorithmic steward of a collective millennial murals. Artists of all generations will construct “epochs” that teach the evolving system the best way to see.
The current era produced by Giraud marks the primary of those iterations. The video piece shows a scene from 1,000 years ago by which villagers in a forest are seen contemplating the tip of humanity after consuming a psychoactive substance.
Courtesy of the artist
Giraud deftly points to the continued importance of language in training AI systems, with echoes of mythical notions of “humanity.” AI is used as a poetic and speculative device, allowing us to take into consideration stories and the long run while serving as a witness to our evolving humanity.
Christopher Kulendran Thomas' installation was partly filmed within the Tamil homeland, which is now inaccessible because of government restrictions. The work overlays generated footage, deepfakes, and scraped social media text to create a counterfactual political story.
The work is installed between reflective screens and features latest dialogue and latest footage with each performance cycle. The video relies on an algorithm that scans social media for brand spanking new content, creating an unstable and unpredictable reality that reflects our age of misinformation.
Dream again
These works encourage us to take into consideration what happens when our social encounters and even our sense of time are increasingly mediated by AI. What happens to dreaming?
Here, Breton's defense of the dream as recalcitrant, embodied and immune to the logic of productivity appears of latest relevance.
The defense of dreaming could also be more of a political statement than a nostalgic reversal. It suggests a desire to ask how we will use AI to assume otherwise and to dream speculatively and wildly about what is feasible and implausible.
This could possibly be certainly one of the last terrains to withstand algorithmic optimization or slopification.
