Theo Triantafyllidis on game engines, simulations, and his Frieze London Artist Award commission

Theo Triantafyllidis works with game engines, and what comes out of them is usually a system rather than an object. He trained as an architect in Athens before moving into media arts, and the habit shows: he builds an environment, writes its rules, drops in bodies that are human, animal, monster, and machine by turns, and then lets the rules do the work. Ecological collapse, social entanglement, and networked desire are the recurring subjects, handled with enough comedy that the dread arrives sideways.

Feral Metaverse, the multiplayer game he has been developing since 2023, is the clearest case of that method. Players are assigned a body they did not choose, stripped of the usual ways of talking to one another, and left to work out how to cooperate anyway; getting anywhere means piling into temporary group forms with strangers. It has already existed as a performance with Onassis ONX in New York, as an installation at Ujazdowski Castle in Warsaw, and as a download on Steam, each version less a finished work than a test of what players do with the rules. His simulations have also run at the Whitney, Centre Pompidou-Metz, and the Athens Biennale.

On 14 October, a game took on a new form at Frieze London. Feral Metaverse (Spider), the 2026 Frieze London Artist Award commission by Triantafyllidis, turns the fair’s Focus section into a space of collective play. Co-commissioned and co-produced by Forma and Frieze, and developed in collaboration with Google Arts & Culture for the first time in the award’s history, the work takes the form of a large, soft, spider-like sculpture built in Athens by a team assembled for the project.

But this is no sculpture to simply stand before. To enter the game, visitors must climb, hold on, and rely on one another. Speaking to hube ahead of the fair’s opening, Triantafyllidis reflects on rules, assigned bodies, and the forms of cooperation that emerge when a system cannot be solved alone.

hube: You’ve described your work as building performative systems where natural and synthetic intelligences coexist. How did that framing first take shape for you, and what continues to draw you to the unstable threshold between the two?

Theo Triantafyllidis: I think it developed gradually through working with game engines. At first, I was interested in games because of their expansiveness, the worlds, systems, relationships, and the amount of time someone can spend inside them. But eventually I became more interested in the game engine itself as a medium for building systems.

With the simulations, I became interested in what happens when you create a very simple set of rules and then allow different agents to interact within them. It made me think a lot more about natural systems and how difficult they are to replicate. I don’t necessarily see the natural and synthetic as opposites. I’m interested in the space where you try to construct something that behaves like a living system, while knowing that you can never completely control what will emerge from it.

h: Your characters—human, animal, monster, machine—often appear caught between control and surrender, humour and dread. What does it mean, in your practice, to give form to a body that is simultaneously primal and overstimulated?

TT: I’m interested in the body because it’s the thing that ultimately has to operate inside all these systems. We can construct very sophisticated technological or political structures, but we’re still these strange biological creatures with instincts, desires, fears, and limitations.

The avatar is interesting because it creates another layer between the person and the body. In Feral Metaverse, for example, I wanted to remove some of the agency that games normally give you by assigning you an avatar rather than allowing you to customise it. I was interested in what happens when you’re forced to identify with something you didn’t choose.

Humour is important because I don’t want these things to become too illustrative or didactic. The contradiction between the funny and the disturbing is usually much more interesting to me than trying to make a serious statement about it.

h: Ecological collapse, social entanglement, and networked desire recur as forces that are “too vast to narrate.” How does the logic of play and simulation allow you to approach those scales without collapsing into either didacticism or mere effect?

TT: A system can communicate relationships rather than just events. With a linear narrative, you are usually being told what happened and in what order. With a simulation, you can establish a set of conditions and let the audience see how those conditions produce behaviour.

That’s what I find powerful about games and simulations. You can create a very loose score and give the elements a vocabulary of actions, and then see what emerges. The audience becomes part of the system rather than simply receiving information from it.

I’m always suspicious of trying to control the audience’s interpretation too much. I trust people to come up with their own narratives and takeaways. I like ambiguity, but I also like having rich symbols and rules that people can project onto. Hopefully the work creates a situation where the audience discovers something rather than being told what the discovery is supposed to be.

h: Feral Metaverse (Spider) asks visitors to climb, hold, balance, and help one another. How does the physical act of climbing a soft, slightly threatening sculpture shift the usual contract between artwork and viewer?

TT: I think that hesitation is actually one of the important mechanics of the installation. In a gallery, people are trained not to touch things. When you put a giant soft creature in the space that looks potentially climbable, there is this moment where people have to figure out what the rules are. Then they see someone else climbing it, or they realise that they need another person to help them, and gradually the behaviour changes.

I wanted the physical installation to reflect the intentions of the game. In the game, you have to collaborate with other players to get anywhere. In the installation, you might literally need someone to help you climb, balance, or operate part of the structure.

h: You speak of the installation as sitting between artwork, social experiment, and interaction. What happens, for you, when the work begins to depend on the awkward, imperfect cooperation of strangers?

TT: I think it’s partly a reaction to the way games are usually structured. Competition and goals are fundamental to games. If you remove all danger or obstacles, people can lose interest very quickly. I wanted to keep those ingredients, but redirect them.

Instead of the players competing against each other, the players have to work together against the game itself. The game becomes the enemy, and I’m almost an extension of that enemy as the person designing the system.

That creates this strange situation where the players have to develop their own language and code of conduct. In the online version, they don’t even have conventional communication tools. They can scream and draw, and they have to figure out how to coordinate through that. I’m interested in the moment when people realise that the system cannot be solved individually. They have to start paying attention to each other.

h: The game’s objective is to “embrace the feral” and unlock dystopian worlds without traditional communication methods. What is being reclaimed in the word “feral” here, and what is being risked?

TT: There is definitely a danger there. I don’t think “feral” means that everyone suddenly becomes kind or that online communities naturally become utopian. There is a lot of toxicity and aggression in online spaces, and I was very conscious of that when developing the work.

Actually, that was one of the reasons for making the world feel desolate and potentially dangerous. If I presented a very clean, utopian world and told players that they should cooperate and be nice to each other, I felt there would be an immediate impulse to troll it and do the opposite.

I wanted to anticipate that behaviour rather than ignore it. The visual language suggests violence and danger, but the mechanics slowly require kindness and cooperation. I’m interested in that contradiction—whether you can create the conditions for a different kind of behaviour without simply instructing people to behave differently.

h: An earlier version of Feral Metaverse is already playable on Steam. How did the transition from screen-based multiplayer to a climbable, shared physical structure change the emotional temperature of the work?

TT: There’s a lot of agency in contemporary games, especially around avatar creation. You can customise every detail of your body, your clothes, your appearance, and increasingly those choices are connected to skins and monetisation.

I wanted to twist that expectation. In Feral Metaverse, the game assigns you a random avatar that you cannot change. It’s a very arbitrary identity, which is actually closer to how identity works in the physical world. You don’t choose the body you’re born into.

I was curious about what happens psychologically when the game takes that choice away. Do you start to identify with the avatar anyway? Do you feel disconnected from it? Do you start treating it differently because it wasn’t your choice?

Eventually, by giving different animals and humanoid characters different abilities, I want to introduce another level of asymmetry and see how that changes the social dynamics between players.

h: Working with Google Arts & Culture and Forma, you’ve assembled a team in Athens to push both the game and the sculptural form further than before. What does it feel like to develop something this ambitious with people who both believe in the vision and are prepared to challenge it?

TT: They are surprisingly different experiences. Online, I wanted to create this very anonymous and primal situation. Players are randomly assigned avatars, it’s difficult to recognise your friends, and communication is intentionally restricted. The idea is that a language gradually emerges through the need to collaborate.

In the exhibition, that anonymity disappears. You’re suddenly looking another person in the eye. You can talk to them outside the game, physically help them, and see their body trying to climb the sculpture. The collaboration becomes much more immediate and physical.

I think putting these versions next to each other makes the difference between digital and physical sociality quite visible. The same basic idea produces completely different behaviours depending on the environment. That’s something I find very useful for developing the game further.

h: Frieze London places the work in the heart of the fair’s Focus section—a commercial and social ecosystem with its own codes. How do you imagine the piece negotiating, or disrupting, that particular atmosphere?

TT: That’s one of the biggest practical problems with making this kind of work. The online gaming audience has a very different expectation. If they download a game, they probably expect to spend a significant amount of time with it. In an exhibition, people might give you five minutes, especially if there are many other works around.

I’ve been trying to make the work operate at different levels. Someone can walk in and immediately understand something through the physical installation, or have a short interaction with the game, but there should also be enough complexity for someone who stays longer. The multiplayer aspect makes that especially difficult because you also need other people to be there. In a gallery, the social situation is unpredictable. Sometimes there are enough people and something really interesting happens; sometimes there aren’t.

I’m not sure there is a perfect solution to that contradiction. But I like designing around it rather than pretending that the two audiences are the same.

h: This year’s Frieze London Artist Award specifically invited experimentation with advanced technologies, including AI, as a collaborative tool. How did you choose to let AI enter the process—what did you ask of it, and what did you refuse to outsource?

TT: I think of generative AI less as a tool for making content and more as a strange new material, shaped from huge amounts of human culture compressed into something fluid and constantly changing. I wanted to explore what this material could do, asking it to generate images, textures, and possibilities, while treating its biases and limitations as part of the process. I didn’t want to outsource the artistic decisions themselves. Like a game engine, AI comes with inherited assumptions and commercial histories. I’m interested in working from within those systems, understanding their biases, and pushing against them.

h: The award supports an artist at a pivotal moment. Looking at the trajectory from the Whitney, Centre Pompidou, and Venice to this commission, where do you feel the work is currently pivoting?

TT: As the projects have become bigger, I’ve had to move from being extremely hands-on with every part of the process towards managing a studio and working with people who have expertise that I don’t necessarily have.

That’s been challenging because I’m used to understanding every part of the process. But there’s also something really exciting about putting trust in other people’s intuition and personal drive. A game is already a collective medium, because there are so many different people involved in making one. I think the same is true of the artwork.

I’m trying to apply some of the same principles that I use when designing the systems themselves. You establish a framework, but you don’t completely determine what happens inside it. Collaboration can work in a similar way. You need a shared direction, but you also need enough openness for other people to influence what the work becomes.

h: Your characters search for a tender, hormonal core in an overstimulated world. Is tenderness still possible inside systems designed for speed, extraction, and constant stimulation—and if so, under what conditions?

TT: I think I’m interested in creating the conditions for tenderness rather than making a work that simply represents tenderness. That’s why collaboration is so important to me.

If I make a very beautiful utopian environment and tell people that this is a peaceful world, it doesn’t necessarily produce peaceful behaviour. I’m more interested in creating obstacles that force people to rely on each other. Something as simple as needing another person to help you climb can create a very immediate form of trust.

I don’t know if I would call that resistance in a political sense. I’m wary of making the work too explicit in that way. But I do think there is something interesting about creating spaces where the dominant behaviours of competition, optimisation, and individual performance are temporarily suspended.

Maybe tenderness is more interesting when it emerges unexpectedly. When it isn’t the stated goal, but something people discover they need in order to get through the system together.

h: Once the fair closes and the sculpture is dismantled, what remains of Feral Metaverse (Spider) for you—an archived game, a set of memories of strangers climbing together, or something that continues to mutate elsewhere?

TT: I think it’s a combination of all of those things. The sculpture is temporary, but the game can continue, and the experiences that happen around the sculpture become part of the development of the work.

I’m very interested in iteration. Each version of Feral Metaverse is partly an experiment in seeing what people actually do with the system. I can design the rules and anticipate certain behaviours, but players always find ways of doing things that I didn’t expect. Those discoveries feed back into the next version.

I don’t really think of the Spider as a final object. It’s one manifestation of a larger system. The sculpture will eventually disappear, but the things that happened through it—the strange interactions, the ways people collaborated, the ways they misunderstood or broke the rules—can continue into the next iteration.

In that sense, the work is less like an object that gets completed and more like an open system that keeps mutating. Hopefully the archive is not just a record of what happened, but material for whatever the work becomes next.