CURA.

Lawrence Lek: Geomancer—the main character of both films—is a kind of alter ego. I wanted to create a coming-of-age story from the point of view of an AI coming down to Earth and encountering an alien planet. On one level, it’s about dislocation and the desire to belong; on another, it’s about the everyday reality of creative work. I also wanted to focus on music, which is often my starting point, but which I hadn’t focused on as a subject and medium.
When I’m building virtual worlds, I use music to fill in the space that images can’t capture. In doing so, I’m able to evoke a different time and place to what’s immediately represented. World-building is often discussed in terms of how the [constructed] world appears, whether visually, narratively, or texturally, but I always start with the experiential qualities of a place. Usually, I work in CGI or rendering, which is a very pristine medium and dates back to when I worked as an architect. Focusing on sound, voice, and music adds texture to the atmosphere.

LL: A lot of work about AI abstracts technological phenomena and separates them from lived experience. But I have strong personal associations with technological objects, technology doesn’t exist separately from my attachment to it. I feel similarly about architecture. Across Southeast Asia, the tides of national or social change are all intertwined with architecture and design. It’s hard for me to disentangle my childhood memories of these places from actually seeing these buildings go up. What I’m creating isn’t just a critique of technology. It’s a critique intertwined with an engagement with the same subject.

LL: A relationship to place is a double-bind. To what extent has colonialism been a positive and negative force? It’s never straightforward. Nationalism and the formation of identity are such political tools, so it’s hard to disentangle which aspects of nation-building are community-building, and which are political machinations. 2065 (2018-ongoing), Geomancer, and AIDOL are set in Singapore and Malaysia, not only because I’m familiar with them, but because their history stems from an entanglement between the natural and artificial, the colonial and indigenous. In these works, AI can also be seen as an avatar for other invented or imported species or people. The countries have a complex history of migrants at different stages of history, from indentured servitude to first-world expat labor. It’s a very layered construct.

LL: Take Singapore and Hong Kong. Both are successful because they’re interzone port cities. Their main value is being in the middle of things. Conventionally, having some kind of natural resource has been a catalyst for development. But in the case of Hong Kong and Singapore, this in-between status is profitable. It’s conducive to the import and export of goods, services, and people. There’s a whole spectrum of related questions about labor in relation to colonialism and globalization.
How it relates to AI is very different. I’m not talking about how AI is developed in industry and academia, I’m talking about the notion of AI and autonomous work as a source of abundant knowledge production and mechanical labor. Look at ports. Dockyard workers in Singapore, HK, and so on, used to be the main source of casual day labor at the harbor. But sea freight ports are increasingly automated. In Rotterdam, they’ve built a fully automated part of the port where there’s no human work except for in the control room. Flows of people, processes, knowledge, and work, in relation to the global circulation of goods, have always been a part of these seafaring areas. Here, AI is part of automating the processes of how harbors are working, how infrastructure is developing.
In Sinofuturism, AI is disenfranchised and nonhuman, but it also has a capacity for providing such a surplus of knowledge and mechanical work that people feel threatened by their obsolescence. Contrast this with the case of Malaysia, where the British imported people from the Empire to work in specific industries. People from Southern India worked in the rubber plantations, the Southern Chinese worked in the tin mines. Specializations arose from how the colonialists defined the suitability of people to roles. In terms of the Chinese diaspora, this happened differently to how it occurred in the Caribbean, in Americas, and in Southeast Asia. These were things I was looking at in Sinofuturism—which, in the case of the opium trade, is much more complicated than being just about exporting and importing labor.

LL: Science fiction is an amazing resource for conceiving of things differently, but it’s important to remember many of today’s celebrated writers—like Ursula Le Guin, Samuel Delany, Octavia Butler, Philip K. Dick, Liu Cixin—are outliers in terms of science fiction. So much of the genre is imperial space colonization, or conquest of the alien. The writers who are disproportionately focused on, in terms of futurism and critiquing capitalism, gender politics, or colonial relations don’t represent the majority of science fiction. Fredric Jameson, in Archaeologies of the Future (2005), sketches out many interesting threads to do with utopian fictions that could also be seen as science fictions; ideal societies from Thomas More’s Utopia and Plato’s Republic to Russian science fiction authors of the 1920s. These writers were trying to create prototypes for social change.

LL: The first time I came across world-building was in science fiction: finding books with maps in the beginning, and encountering ideas of imagined places. Now, world-building is associated with video games, literature, or even the Marvel cinematic universe.
It’s rare to see the process of nation-building as explicitly as you do in Singapore. I moved there from Hong Kong when I was 7. [The Singaporean government] is really overt about social engineering. I really felt that these nations were in the process of creating themselves in real-time. I had no sense of complex geopolitical or social history. In school, Singaporean history started when it became a British East India Company trading outpost.
As a kid, Hong Kong and Singapore seemed as fictional as the books I read or the video games I played. At the time, I remember that the prevailing mood in Hong Kong was of change, even though it was ten years before the 1997 handover. I remember seeing the Bank of China skyscraper go up; much later, I realized that its construction was related to Hong Kong’s geopolitical situation. But back then, I saw it as a structure from a video game.

LL: In a conventional film, the film starts and the film ends, but narrative chronology can be cut-up, disjointed, or can involve huge leaps in time. In AIDOL, I was considering how it’s not just the chronology that is distorted through narrative. Different visual forms—comparing the music video with the simulation, or the video game with the film, for example—occasion different forms of time.
In AIDOL, this relationship is deliberately ambiguous, because the film is also divided into thirteen tracks. The film ends with the launch of the music album that the characters have been working on. It’s not clear if you’re seeing an animated feature film, or a thirteen-track promotional album video.
In my experience, time is the most difficult aspect to handle: how cycles of cause-and-effect flow together, interact with each other, and disrupt our memory. Sometimes I’ll mix the temporality of the music video with the temporality of the video game to create a different rhythm to the work. Filmmakers like Wong Kar-wai and Tarkovsky are amazing at making time apparent.
Simulation is about real time—it’s a much more metaphysical thing. What is real time? What is the existence or status of the simulated work of art? Some media theorists argue that the digital work of art doesn’t exist unless it’s instantiated. If there’s a computer program that’s a work of art, each time it runs, it is a live performance. It’s not the code, but the running of the code that constitutes the work of art. Similarly, the objecthood of the simulation depends on it running. This gets us into various degrees of what constitutes reality and quickly becomes very Baudrillardian.
But to me, simulation is like a performance. With music, it can be live and it can be recorded: two mutually complementary modes of existence. The simulation provides a way for people to explore the films by themselves. They’re free in that dimension.

LL: Doom resurrects an older idea that I was dealing with until about two or three years ago: that of site-specific simulation. In the simulation, I started in two of the rooms inside Somerset House, then moved outside, then into the virtual world from Temple OST (2020), which is a digital copy of an installation I made at 180 The Strand in London. During a live performance, I’m able to mix and blend between virtual spaces and soundtracks I’ve done.

LL: Any time-based media forms a relationship between what you are perceiving in that moment, your memory of the recent past, and your anticipation of what’s to come. Sound is perceived in the present, but you can’t have a melody without the past, present, or future.
You can quite often date the music you’re listening to through its sonic characteristics. In architecture, you can see the age of materials and how a building has weathered. In CGI, you can often date the work by its visual qualities.
Rendered images and simulations can trigger nostalgia in the viewer. You can see how time is embedded within the form. With analogue media, people talk about film grain and photographic aberrations, like there is a yearning for physicality and contact. But I’m interested in how ‘new’ things also bear an archeological imprint of time. It’s hard to identify in the present moment, because these kinds of images or experiences haven’t been associated with history, or nostalgia, or emotion. Yet.