Four Characters in Search of a Random Exit—Alice, Bob, Carol and David

Viktor Timofeev

media [4]

[1 / 4]

Viktor Timofeev, Bob Doesn't Ever See David, 2022

[2 / 4]

Movement scores for Four Characters

[3 / 4]

Movement scores for Four Characters

[4 / 4]

Movement scores for Four Characters







Four Characters in Search of a Random Exit, 2017, was first exhibited at Kim? Contemporary Art Centre in Riga, Latvia as part of my Stairway to Melon solo exhibition. For RTRU, I revisited the piece, distilling some of its components in order to reveal what lies “under the hood.”

Writing about a past work can be challenging by default—one has to come to terms with decisions made by your previous self. Writing about this piece in particular, with its myriad of intertwining rules and, maybe even more confusing, its exceptions to those rules, made me question whether opening this vault would simply bury me deeper into my own logical hole. On top of that, I found numerous technical fixes and updates to be made now having five years worth of hindsight. But instead of meddling with the past, I opted to re-examine Four Characters using a speculative lens, trying to both respect its origins and discover the work anew. Remembering the conceptual and biographical context that informed it also proved fruitful, and I am now convinced that if I had to face the same creative decisions again, I would not change a thing. The resulting “supplementary” materials—diagrams, sketches and video annotations, grew beyond the scope of the original work and took on a life of their own.

Four Characters in Search of a Random Exit (excerpt), 2017-2021. Custom software. *This is an excerpted screen-recording of a single-channel split-screen version of Four Characters in which the landscape format is divided into two equal portrait orientations.

Four Characters

Four Characters is a simulation displayed on two video channels, which consists of software-driven choreography that can run for an infinite duration of time. The piece is performed by four characters, all of whom are clones of a zombie-like asset purchased on the Unity 3D digital marketplace. One video channel displays a third-person view of each character, while the other channel displays their first-person perspectives. These views show two divergent worlds side by side: elements that are present in one view are missing from the other. All of the characters share the same virtual time-space, yet are not aware of each other and only respond to their own subjective realities. Shaped by conditions and stimuli tailored to each character, these subjective realities cause the characters to behave in specific ways: they move at different paces, move in various directions, or perform distinctive actions. The two channels focus on one character for a minute at a time before moving on to the next character, and so on.

Each of the characters’ realities has a timeline, starting with a common baseline setting (e.g. a solipsistic uninterrupted open space with a blue sky) that then evolves into a unique, exaggerated state the longer that the characters are left alone to play out the simulation. The pace at which these timelines ebb and flow is slightly different with each iteration—the programmed choreography leaves room for randomness and chance, allowing the simulation to operate as a small, interdependent ecosystem.

Virtual Virtual Reality

In order to shed some light on how the piece came to be the way it is, I want to share what was happening in my life at the time. Four Characters is a direct outcome of Physical Capacity (2017), a VR game I made which turns its players into bug-like quadrupeds. The game coerces players to crawl around in pursuit of randomly generated bugs while maneuvering around labyrinthine virtual walls and a system of color-coded prompts. As the audience watches participants wear VR headsets and crawl on the ground while playing the game, the audience is simultaneously unable to see what the participants are responding to in game-space. For the audience, the participants become “philosophical zombies,” a projection of actual decision-making by an observer outside of a shared, empirical reality. This observation was illuminating—not knowing the cause, but seeing the effect made me feel like I was witnessing a hidden logic whose inner workings were obscured, and like I was increasingly aware of a consciousness that I was distinctly outside of.

Idling animation seen from first and third person views, different lighting. *From an exterior, third-person perspective, we see the character idling under an omnipresent spotlight. In their first-person perspective, we see them looking at their long shadow cast by the setting sun.

Watching someone play an invisible game begs the question—am I playing one as well? And if so, is the game I’m playing similar to someone else’s? Is it possible for these invisible, inner spaces we occupy to look alike? How can we be certain either way? After all, the inability to read the mind of another in order to make sense of their behavior is an accepted component of everyday life. This very dynamic in Physical Capacity serves as a reminder that each and every person is guided by mental processes not visible to anyone else. It would eventually also become the foundation for the diverging first and third person views of Four Characters.

Fellow Zombies

Early in the development of the project, I stumbled on a few things that fueled my budding ideas and shape my zombie character. It’s entirely possible I misread these works and only took from them what I thought was relevant. Either way, these proved to be absolutely crucial to my working process, so I want to share a few of them (and a few others that I didn’t get a chance to discuss here are itemized in the Research module).

Experiencing Yvonne Rainer’s performances of her seminal choreographic works from the 1960s kind of blew my mind. The generative scores kept me searching for recurring patterns and wondering what was coming next. The repetitive, pedestrian movements instantly made the dancers more relatable. As I watched them, I kept thinking—“What prompted this person to move in this particular way at that specific point in time? How did the person on the other side of the room know exactly when to respond to that? What is the balance between improvisation, chance and structure?” I ended up attending several performances executed by the same dancers in order to see how they would differ- and I was pleasantly surprised at the variety of outcomes. This feeling of wonder was compounded by a Twilight Zone episode called “Five Characters in Search of an Exit” from 1961. In it, a group of characters is fatefully brought together for an unknown reason which becomes more apparent as they interact and their backstories are revealed. I connected the episode’s inward-facing narrative with Rainer’s embrace of emergence and was inspired to make my own version using the Zombie asset I bought online.

I used this specific Zombie asset because it was the only character I found with a crawling animation. Crawling was crucial, as it was my bridge to the aforementioned VR game Physical Capacity, which encouraged real life participants to behave like quadrupeds. The character included seventy animations, enough to act as a convincing, non-playable zombie. I chose a small selection from this pool of movements, looking for those that were generic enough to remain ambiguous, interesting, and relatable outside of a zombie game context. These included idling, walking, crawling, running, squatting, climbing and dying.

The characters now needed a stage—I gave each one a unique, allegorical reality to live in and respond to. These were loosely based on themes that I was thinking through in my own life, namely aging, relationships, and family. For example, in the world of character 1 (a.k.a. Alice), the speed at which time passes accelerates the longer Alice is alone. Coupled with a simple rule that Alice is only allowed to move during the day (and is forced to idle at night), Alice ends up covering less and less ground over time. This connects to the principle that perception of time accelerates as one gets older.

Still from Bob Doesn’t Ever See David, 2022. *The annotated video lays out the logic at work behind Four Characters step by step, telling the story of each of the characters from their perspective.

Alice, Bob, Carol, David… Viktor?

It is not lost on me that somewhere along the way, working on Four Characters I started to identify as one of my zombies. For several reasons, my life seemed to resonate with the puppet-like behavior of the characters I was programming. Whatever I was going through in my day to day, at least I could play God here and find cathartic satisfaction in watching my characters try to execute my every whim. I grew attached to them on an emotional level and started to envy their reliable logic. In the same spirit I eventually decided to give the characters proper first names, instead of referring them to 1, 2, 3 and 4. I based the names on the four most common placeholder names used in cryptographic systems and protocols—Alice, Bob, Carol and David. The names ring with a generic banality that struck me as particularly fertile. As a result, the diagrams I was preparing for RTRU started to read less like a development log and more like a silent theater play, complete with stage directions for each of the “actors”.

The supplementary materials include an annotated video recorded straight from the Unity game engine Editor window, with the game scene displayed in isometric projection. This view is loosely analogous to the third-person view in the finished work, but is more bare bones since it lacks many of the visual effects and automated camera movements that were added later. Filmed mostly from above, the video follows each of the characters one at a time, laying out the rules that they are following in realtime. Every movement is annotated and tracked in a forensic style. Their simulated life cycles are also sped up, shedding some light onto the limits of randomness in the process—Alice, Bob, Carol and David never do the same thing twice and are all stuck in loops that eventually blur together.

research [10]

[1 / 10]

Viktor Timofeev, Four Characters (2017 - 2021). Installation view at Metabolic Rift at Kraftwerk in Berlin.  Two-channel custom software, sound. Courtesy of Berlin Atonal.

[2 / 10]
[3 / 10]

Viktor Timofeev, Physical Capacity (2017). Installation Views. Custom VR software, modified hardware,  furniture. Courtesy of the artist. 

[4 / 10]

Samuel Beckett, Quad (1981). Score and performance views and score. 

[5 / 10]

Yvonne Rainer, Score for Trio B: Running from The Mind Is a Muscle, 1966-68; graphite and ink on paper.  The Getty Research Institute.

[6 / 10]

Selected stills from The Twilight Zone, Five Characters in Search of an Exit (1961), Season 3; Episode 14. 

[7 / 10]

Stills from Google’s Deepmind Labs (2016). 

*I was particularly drawn to the use of positive and negative reward systems in order to study A.G.I., and  made my own version of a research paper published by DeepMind. My version fictionalizes the research  experiment and replaces the positive / negative reward scheme with one that is non-binary. You can find  my version of the paper here: 

[8 / 10]

Unity 3D workflow - Animator window, state machine view for Character 1 (Alice).  *Please don’t judge my bad workflow habits.

[9 / 10]

Unity 3D workflow - Visual Studio window, excerpt of C# script for Character 3 (Carol).

[10 / 10]

Unity 3D workflow - Editor window.