Glazed Ceramic, Silk Fiber, Formica Dining Set, Video, Computer Vision, Cloud Simulation, Touchdesigner
In English the word “cloud” is frequently used to refer to an amalgamation of matter - water droplets, small dust particles, as well as data. Unlike naturally occurring clouds, a data cloud cannot appear out of thin air. It is well established by STS scholars that data is a verb whose execution is always supported by visible or invisible human labor and different ways of knowing. Human decisions and interpretations is involved every step of the way in data production, whether in the use of tools, choice of classification, and methodologies we use to come up with analytics. New technologies and infrastructure such as networked smart objects in domestic and public built environment are being established as standards to purposefully and systematically generate data that prioritize machine aided calculation, extraction and optimization, instead of archiving critical aspects of human lives for human interpretation or knowledge production. This data is designed to support machines’ way of knowing which is coded at their cores by individuals who wish to reproduce very specific social orders through wide adoptions of automated processes. In When Data is Capital, Sadowski argued that “Fulfilling the data imperative involves more than just passively collecting data; it means actively creating data. This entails the (total) datafication and surveillance of people, places, processes, things and relationships among them. In a way, we are all becoming part of the cloud.
Teach The Machine Something About Cloud (2024) is a multimedia installation first installed at SOIL Artist Run Gallery in Seattle, Washington. Upon entering, viewers encounter a scene. A vintage looking yellow dining table with shiny aluminum detailing paired with two matching chairs on both sides. On the surface of the table are nearly one-hundred ceramic cubes of one-inch size. The cubes each has five walls, and is filled with heavenly soft blue silk fibers, like fog. Behind the table, a suspended yellow window frame frames a generative cloud animation. A webcam sits on top of the window. Its green LED light on, indicating that the mechanical eye is watching.
The cloud animation is constantly evolving, vaguely responsive to the cubes on the table. To create this cloud simulation, I employed an image-to-image StableDiffusion generative AI model guided by text prompts. The webcam image is analyzed by computer vision algorithms for borders, shapes and colors, which then drives the movements of an animated particle system to prompt the final cloud simulation output. This internal logic only visible to the machine is not accessible to the viewers. Despite viewers religious attempts to solicit feedback, the cloud simulation does not purposefully produce any instant gratification of recognizable validation.
How to make sense of this scene, then? Viewers will find a tablet kiosk on the wall on which a corporate-style onboarding video is looped. The video outlines four steps which an individual can take to "Teach the Machine Something About Cloud". Step one, generate random patterns. Viewers will randomly pick up two ceramic cubes from the table and compare their weights. The one lighter in weight will be flipped upside down, hiding or revealing the blue silk fibers inside. Viewers will repeat this 10 times to generate a random pattern on the table. Step two, fill the grid. Viewers will observe the generated pattern and try to fill out a 9x10 grid on paper accordingly. Step three, data prep. Viewers will translate the grid from previous step into machine readable binary codes - 0 and 1. Step four, data entry. Viewers are asked to think of a cloud lesson to teach the machine, and pair up the handwritten natural language with the binary code they just generated. After completing the four steps, they have now voluntarily produced data and contributed to an absurd dataset -- one that means nothing but carries the affect of becoming anything and everything.
Teach The Machine Something About Cloud }mocks the over-datafication of the world, technocapitalism’s appropriation and reconfiguration of the affects of "cloud", as well as the subjective, unquantifiable and impossible-to-unify nature of human labor involved in data production that is always situated in different corporeal bodies. There are three extra cubes that cannot be neatly placed in the given grid. Comparing the weight between any random two ceramic cubes is a task that is surprisingly difficult to accomplish without the help of a scale. Viewers participating in the work are often perplexed: How accurate does it have to be? Does my work and care for the task even matter, if the goal is to generate randomness and more ambiguity? What does the work mean by “cloud” and what should I teach the machine about it?
The affect of natural clouds influences the aesthetic production of the techno-cloud, shaping our embodied sensory experiences. In turn, the “Cloud” as an information infrastructure reshapes how we position ourselves within natural, technological, and cultural domains. By inviting participants to “teach the machine something about cloud,” this work highlights the cloud's multifacetedness and its intricate role in shaping human experience. Below is a selection of "cloud lessons" authored by participants:
“In the future, clouds will not be permeable.
Planes will not be able to fly through them.”
(Brooke, Stacy, 2024)
Planes will not be able to fly through them.”
(Brooke, Stacy, 2024)
“Designer clouds are some of the most geometrical entities
that exist in the world, as the simulation ran out of RAM,
complex clouds had to be sacrificed in favor of trees.”
(Laura L. C, 2024)
that exist in the world, as the simulation ran out of RAM,
complex clouds had to be sacrificed in favor of trees.”
(Laura L. C, 2024)
“When you sneeze
a piece of what you're feeling goes into a cloud.
And then it rains and evaporates that feeling forever.”
(Lucy, 2024)
a piece of what you're feeling goes into a cloud.
And then it rains and evaporates that feeling forever.”
(Lucy, 2024)