An evolution led by a bubble economy, a production line of the data annotation line veiled under a mysterious algorithm, the annotators’ body and soul is trapped under the cloud of data…Aiming to go beyond the typical focus on AI technology, this film explores a broader relationship between technology, body, labor, geography, and infrastructure, showcasing an evolution on the cloud.
In a poetic narration, the film gradually brings out the individuality and everyday life of data annotators — repetitive works leading to a bodily detachment from consciousness, into a dreamlike multidimensional spacetime, away from the cruel reality and anxiety about the future, floating in an empty office.
The dual-screen setup of this film enables two storylines, one depicts the reality of data annotation: workers on the production line and their repetitive job, bringing their reality into the cloud of boundless data. The other storyline sets in a dystopian AI-dominated world in the near future, exploring the unclear fate of annotators as AI takes over the people who taught it. Empty factories, abandoned gadgets and name tags, faded slogans on the wall — these are the imagination of a post-human society, as well as a twisted reality under the ever-accelerating technologies. The past and the future are in a vortex, time seems to be in a vacuum.
Xu Zhengyue,born in 1989 in Heilongjiang, graduated from the Central Academy of Fine Arts. Currently lives and works in Beijing. Xu Zhengyue’s work uses video, animation, and photography as the main media.
Her works focus on different historical stages and changes in social reality, staring at collective and individual emotions and situations, through on-site research, media, and theoretical research, she uses poetic video language to express artistic thinking and explores the hyperlink relationship that penetrates the appearance in reality.
Xu Zhengyue’s animation work Who Touched the Feather, won the Award for Experimental Innovation and the Chiyong Zhiren Award of the 12th Animation Academy Award of Beijing Film Academy in 2012. In the same year, Who Touched the Feather won the ASIFA China Award in the 7th China (Beijing) International Student Animation Festival. Her video work Pedestrian was collected by CAFA Art Museum and K11 Art Institute. Xu Zhengyue’s works have been exhibited at K11 Art Mall in Shenyang (2018), TaiKoo Li Sanlitun in Beijing (2017-2018), Hanshan Art Museum in Suzhou (2017), Epoch Art Museum in Wenzhou (2016), Art Gallery of CAFA (2016).
Researcher Yang Beichen (YBC) in conversation with Artist Xu Zhengyue (XZY)
YBC: The first thing to point out is that this film unleashes a new perspective on the issue of artificial intelligence. Perhaps this perspective can be summarized as how to give visibility to AI. Can you elaborate on the ways in which you planned the ß and the “split screens”?
XZY: When I researched the archival materials and images on this topic, I realized that I needed to sort out the various parallel threads (body and labor, cloud, geology, technological infrastructure) in a densely packed system of connections through the technical veil of AI; when I went into the field to conduct research and film, the creator’s sense of the physical body and perceptions were amplified to respond to the changing space and surrounding objects; hence I had to adopt different angles in building the content for the “lens”. I tried to reorganize the multiple structures of the film’s content from the complex materials, and the “split-screen” was also intertwined in a not entirely logical way when the structure was deduced. In the form of a double screen (combined screen) in search of the non-linear intersection of time and space and the semantic connections generated by juxtaposition, whose formal properties seem to lend themselves to constructing a sense of multi-dimensional chaos in the material and digital world.
YBC: The invisibility of the algorithm and the data makes its presence extremely difficult to locate in the moving image. However, the film seems to try to offer a solution by creating a “split” image, especially in the double screen section, where the constant juxtaposition and “collide” between the figurative and the abstract, the real and the digital continue to happen. Can you talk about how you have developed this “iconography”, and what problems you encountered in its implementation?
XZY: I wanted to use the medium’s inherent characteristics to organize the different contents, rather than adopting the essayistic format as a precursor, which requires working with images that both observe and strip away appearances, searching for invisible meanings imbedded in the non-fictional images, and establishing correlations among them. Heterogeneous images “collide” with each other in non-linear time, splitting as a possibility for reorganization. While full of different representations, images with internal coherence produces meaning in motion, mobilizing a hyperlink and sense of flow in the double screen.
The ineffable nature of the image that seems to provoke the formless within forms, and leaving room and air for expression and viewing, are some of the more critical and complex issues encountered when dealing with moving images.
YBC: A large part of the film presents the process of machine learning. Training the AI is essentially training in “viewing,” building an epistemological framework for the AI through the observation of large amounts of digital material – where “double” viewing emerges in the film; in other words, we are watching how the AI looks. Can you tell us if you had new reflections and opinions about this kind of data-based machine learning during the shooting and production of the film?
XZY: AI technology is still in its early stages, but we have “naturally” accepted its existence and became one of the vast amounts of data, but the more natural it is, the more we need to reflect on it as the rapidly changing technology brings convenience to life, but also squeezes the physical and mental space in the daily fragments. The film takes a slow gaze at technology; it is also an internal confrontation and reflection on the speed of development.
AI is open to acquiring diverse information and the macroscopic vision to deploy a full range of information in data-based learning. I believe that these characteristics are what human beings should learn and think about in reverse when constructing cognitive systems, i.e., how to internalize the multi-dimensional cognition of the complex world into personal thoughts as much as possible, through personal historical limitations in established paradigms and biases.
YBC: The film seems to emphasize an “outside” perspective, or perhaps we could call it an “earth” viewpoint. These passages, mainly done by drones, stand out in the film, where mountains, forests, large infrastructures, and artificial intelligence, i.e., “nature” and the digital world, are connected and circulated through a channel. Can you talk about the significance and role of drone images in the film?
XZY: Throughout the film, I tried to make the authorial viewpoint as “marginal” as possible, and drone images are one of the ways to reach this periphery.
The drone replaces the author’s subjective point of view, providing the possibility of macroscopic observation of geology and creating a channel between two “cloud” media, such as the juxtaposition of the AI-controlled drone interface in the cloud and the aerial photography of the data center hidden among the Karst mountains. Also, by adopting the “earth” perspective, I want to remind viewers of the importance of existing in the physical world. When using the footage, I also intentionally avoided technically flashy or overly ornamental images instead of using unintentional jams and retaining the camera shake in slow flight as reminders of the camera’s subjective existence.
YBC: The film allows room for the depiction of “people,” including numerous close-ups of their faces; however, on another level, the work of programmers or AI trainers is reminiscent of a labor-intensive factory, and they still seem to be performing some traditional labor in this so-called artificial intelligence era and are therefore subject to exploitation. Can you talk about the role of “people” in your film from this perspective?
XZY: The “human being” as an individual and a group has an essential place for me in my work, and thus I intentionally emphasize the “human being” space when exploring technical issues.
During my research, I went into the production line of digital laborers. After acquiring skills quickly and intensively, the markers were pushed into the data ocean, where capital and technology conspired to train the intelligent machines that would swallow them mechanically. When confronted with a specific environment, there is indeed a sense of pulling disconnect due to the status quo of traditional physical labor exploitation that still exists in the age of intelligence. This is one of the reasons for juxtaposing the long-shot portrait (a state) of gazing at people with the grand technological evolution. The way I deal with music also includes human voice chanting as a kind of call and comfort for “human beings .”In the present time, under the pressure of the rapid development of technology, it is crucial for the “human being” as a community that shares common destiny and an individual to relate to the material world and evoke physical and spiritual perceptions.
Dr. Yang Beichen is a researcher and a curator based in Beijing, China. He is the director of the Macalline Art Center (Beijing), and one of the members of the Thought Council at the Fondazione Prada (Milan, Venice).
Currently teaching at the Central Academy of Drama, his research examines the agency and potentialities of the moving image in the context of contemporary technology and ecology, deploys media archaeology as a radical framework to excavate alternative modernities, and ultimately aims to re-interpretate history and geopolitics from a New Materialist standpoint. His curatorial practices grow out of and attest to his multidisciplinary academic approaches. Notable curatorial projects include “New Metallurgists” (Julia Stoschek Collection, Düsseldorf), “Micro-Era” (Kulturforum, Berlin), the Guangzhou Image Triennial 2021 “The Intermingling Flux” (Guangdong Museum of Art, Guangzhou), “A MOON WRAPPED IN BROWN PAPER” (Prada Rong Zhai, Shanghai) , etc. Between 2019 and 2021, he also led a three-year research project focusing on the art of moving image in China at the NCAF, for which he curated three research-based exhibitions: “Anti-Projection: Media Sculptures in Early Chinese Video Art”, “Embodied Mirror: Performances in Chinese Video Art”, and “Polyphonic Strategies: The Moving Image and its Expanded Field”.
He has also contributed critical essays for the catalogues of the artists such as Cao Fei, Laure Prouvost, Omer Fast and HO Tzu Nyen, etc. His academic monograph Film as Archive will be published soon.