Francis Tseng

Towards just futures through simulation

December 3, 2020 / Interview by Alex Westfall | Photography by Kira Simon-Kennedy

Simulations offer us a way to understand the world we live in. Sometimes, they allow us to critique the present; other times, they let us imagine alternate futures. Straddling design, games, activism, and research, Francis Tseng‘s work does both. In 2016, Francis made The Founder, a web-based game that ushers players through the founding of an imagined Silicon Valley startup. Accented by a cheery color palette and narrative design steeped in satire, players are led through a multi-level decision-making process that poignantly mirrors capitalism’s far-reaching presence in our lives. 

Francis has taught courses on game design at the New School, worked on interactive media projects for The New Inquiry, and now researches at the Jain Family Institute

We spoke with Francis about what a game’s ‘God view’ can reveal about real-world power dynamics, the possibilities and terrors of machine learning, and the entangling of critique and imagination.


When did you know you wanted to have a creative practice in your life?

Growing up, I always wanted to play video games. I have distinct memories of my dad growing up playing the early Civilization games as well as Dark Castle. But my parents also wouldn’t let me get a console. It was that kind of household.

Many people I grew up with had Nintendos, and we had a family PC, so I could get some games for that. But gaming didn’t become a really big part of my childhood until a bit later than most people in my generation. The most creative outlets growing up were things like Legos—so many building worlds and making up my own stories. That became one of the main ways that I related to the world. Over time, I was drawn to projects that would let me exercise that.

Which games and artists did you absorb growing up?

This was when AOL chat rooms were big. As you surf through those, you encounter a lot of people who do things like role-playing. I would hop in and lurk and see what was going on, though I was never a very active participant.

At some point, I did start playing video games, and that became a very big influence, things like Final Fantasy VII (the remake came out recently, so there’s been a lot of nostalgia around that for me this past year!), or Age of Empires. I have a distinct memory of my dad playing it himself, and there was a teacher parent-meeting at school, and he tried to pitch to the teacher to use this game to teach history. Games had this secondary educational effect in that way.

The office map in "The Founder," Francis' dystopian business simulator game.

It also became clear that games are a really important way of forming people’s ideas of how the world works. It was not merely recreation, but people build these understandings and mental models about how history, nations, militaries, and economics work. That made it clear that these more systemic understandings are important. As I learn more about history and how inaccurate these games can be, I realized that games became a way for both reinforcing these hegemonic ideas of how these systems work and how they should work and can work.

In Civilization, you could pick different governments. It seemed like those were the only possible ones that could ever exist. If you chose, say, anarchism, it would be that the government was overthrown… it was always a bad state of things, and you had to quickly restore order. That helped me understand video games—and media in general—as sites that inform our ideas, and they become deeply embedded, especially at a young age.

Design for imagined technologies in "The Founder."

How has studying cognitive neuroscience in undergrad impacted your current practice? 

I was a really bad student in school. I originally applied as an economics major, but I was also interested in design and general behavior. Economics has a very baked-in set of ideas on how people behave or what drives people’s decisions. Design was becoming more flexible in understanding that people behave differently in different contexts. There are broad values in different cultures, which means that one understanding of an economy doesn’t necessarily apply to another place or people. 

Cognitive neuroscience felt like a way to explore that without getting bogged down by these other disciplines set on assumptions on behavior. I’ll be honest. I don’t think I got that out of the program. But that desire to understand behavior and all its different forms it’s persisted through.


In what ways do you think about the relationship between critique and imagination with respect to games?

Games and simulations are pretty synonymous. For that category of interactive media, there are two kinds of ways that they can be valuable. I’ve found the process of developing games and simulations to be the most useful—especially really detailed simulations—because you have to be explicit about how whatever system you’re simulating works. You can’t make any hand-waving assumptions. Once you start to dig into the material, you realize how much you don’t know. So the process of developing a simulation is similar in that respect. That’s one side of it.

The other side is the actual experience of playing a game or simulation. Putting you in a decision-making position can be really important. Sometimes, when we talk about decisions that other people make, it’s hard to really understand the context in which that decision was made, and what the competing interests and considerations were, and so on. Video games can help situate someone in that decision making context in a better way. That’s what I had in mind when I made The Founder. It wasn’t meant to say that Silicon Valley people are evil because I don’t think that is useful. That doesn’t provide any path to actually dealing with the issue at hand. The Founder exists more to show that you’re going to adopt these malicious behaviors within a certain context or situation. It’s not about removing a person from power. Still, it’s about effecting a more systemic change that doesn’t put people in that situation where these malicious decisions are rational or necessary in a perverse way.


What was the development process for The Founder like? You note that Sim City inspired you—what did you take from that game, and what aspects did you leave behind? 

I lived in the Bay Area very briefly. When we first moved there, I was interviewing at different companies—talking to founders and trying to get hired–so I saw that aspect of it too. I worked at Ideo, a design firm that has relationships with many of these big tech companies in Silicon Valley. The tech culture of San Francisco was another exposure I had to it—people would only talk about their startups at parties [laughs]. 

When I originally came up with the idea for the game, it was much more benign and simple. I had wanted it to be a mobile game, so the first year or so of development, I was making it in Unity. I had a lot of frustration with Unity as a platform, though, because I don’t think it’s really well-tailored to that kind of game. I was constantly worrying about what platforms it would be on, how people would access the game. I decided that making it a web game would be easier because anyone with a browser could play. I was more comfortable with all the coding and tools for browsers. 

Simulations have this top-down “God” view. I understand that—when you’re dealing with a big system, the most intuitive way to do it is from a system-level view. What I liked about Silicon Valley or startup founders as a topic is that making that game from a “god view” is part of the critique: these founders have a disproportionate amount of power in deciding what happens within these companies. 


Character design for "The Founder."

Part of what draws people to startup culture is this feeling that you’re part of something that’s growing and progressing. That’s why there are these upgrade mechanics and pathways. When you have a character and unlock these skills over time, you’re invested in the character in the same way you’re invested in this company, it becomes part of your identity. You have a much stronger attachment to it. Also, the “god” view works not only on a mechanic level but also as part of the critique I was trying to show.

What was the process like for conceiving the illustrative and aesthetic features of the game?

A lot of the aesthetic, especially the 3D stuff, was mostly practical—low-poly modeling. I have very basic 3D modeling skills, and it was about what I could accomplish. Then the simple color palettes were also easy to work with. I made the employees cone-shaped figures because I had no idea how to make them humanoid, and animating that would have been really tricky. That was why they’re cone-shaped people [laughs]. In terms of the interface, I wanted to make it look really naïve and cheery, and colorful because I thought the contrast with what the actual material of the game was dealing with was really interesting and also reflective of how a lot of Silicon Valley web design works. It’s meant to look very cheery and friendly, but usually, the actual workings behind the scenes are much more insidious.

GIF capture from The Founder.

What constitutes success in The Founder in your eyes? I imagine it might be different for players because each player has their distinct priorities.

When I originally made the game, I didn’t want there to be any win condition. The point was that this is a system that expects endless expansion and growth, so there’s never really an end to the game. You just constantly have to hit that return rate. Then, I felt obligated to have some end in it. I thought it would be interesting to have that ending become this true ending, where you develop an AI that’s supposed to increase the company’s productivity and profitability until it ends up replacing you. I don’t know if that would be a success in the game, but it’s the only ending as far as I remember. The point of the game is no one really wins in that situation.

Minigame design for "The Founder."


I’m also curious about your involvement in The New Inquiry and the different interactive design projects that you worked on there. What led you to decide that certain projects would need an interactive component versus a more traditional, written critique?

When I was there, we developed this idea that we ended up calling rhetorical technology or rhetorical software. The idea there was that these software projects could embody a critique in the same way a written one can. Written critiques can have a very powerful impact, but software critiques or rhetorical software can have different kinds of impacts that feel a bit more direct. The primary example of that was with the Bail Bloc project because it’s at once a critique of the bail industry and the prison system, the justice system, and at the same time, it actually has functionality as part of that critique that’s meant to alleviate some of the harm caused by the criminal justice system. Joining direct action with critique was one way to summarize the guiding principles of those projects. 

We were contrasting the gulf between the criminal justice system—that, with cash bail especially, criminalizes poverty—and the cryptocurrency world, where obscene amounts of wealth were (and are) generated through currency more fictitious than the kinds we’re used to… Basically creating wealth by doing nothing. The people who get wealthy off cryptocurrency are those already well-insulated from the criminal justice system. Bail Bloc tries to close that gulf in some way by tapping into the money the cryptocurrency space was generating and redirecting it towards bail, by doing “nothing.” We were also interested in using the computing power of institutions that may have some complicity in the prison industrial complex (for instance, universities with pensions that have investments in private prisons). Cryptocurrencies were trendy in mainstream media at the time, but talking about the bail system wasn’t. Connecting cryptocurrency to bail gave journalists a hook to report on a topic they normally wouldn’t, and get it in front of people who normally wouldn’t seek it out, like on a cryptocurrency news site. The hope is that the project not only exposes people to this critique and changes their minds but also actually bails people out of jail.

"Bail Bloc."

What draws you to the world of AI and machine learning? Especially with regards to your project, Conspiracy Bot?

I became really fascinated with machine learning in 2014. I was especially interested in how natural language processing could be applied to journalism and the news. I had a very hard time keeping up with the news regularly—if you miss stories as they come out, later stories make little sense unless you spend time catching up on them. I thought I could make a system that would keep up with the news for me, and then when I was ready to read it, it would provide a summary. It seemed ripe for some of these technologies in other spaces within journalism—dealing with comment sections, for instance.

Over time, I became very cynical about AI. I still am. These technologies are expensive to develop and deploy. You need large research budgets, a lot of computing power, and so on, and it seems like it’s going to disproportionately benefit people who already have a lot of resources and power. We’ve pretty much seen that to be the case, so that’s one reason I’ve become sour on the topic. There’s a lot of really important work that people are doing in pushing back against that, to trying to come up with ways that are more socially valuable applications of the technology.

Conspiracy Bot was made after Trump was elected. I found it interesting that in the same way we deride conspiracy theorists for having these misinformed and actively harmful understandings of the world, AI, through its own shortcomings, can develop conspiracy theories equally as harmful. The idea there was to use the fact that a lot of these object recognition and facial recognition algorithms have pretty high error rates—to use them as if they didn’t. Those error rates function as a conspiracy, seeing patterns where there aren’t really any.

I haven’t really done much AI work lately because I’m pretty pessimistic about it, and I’m always worried that any work I do would be used in a harmful way.

You mentioned that there were some cases of people who you think use AI technology for a socially-just reason. 

Medical applications in terms of image recognition, identifying tumors early—those seem like reasonably good applications. I’m interested in when people do aikido—flipping the opponent’s momentum against them. A couple of weeks ago, people were using facial recognition technology to identify police officers in streaming footage and stuff and track police officers with, especially violent behaviors. Those are promising applications.

I should also mention a project I worked on with the artist Sean Raspet called, which was a critique of how pharmaceutical R&D works, but also potentially where AI can help challenge that. One of the given reasons for why drugs are so expensive here is that drug development is costly, which it is. It’s so hard to find compounds that actually have a pharmaceutical purpose. There are 10 to the 60th power possible chemicals, which are more atoms than make up all planet earth. It’s an obscenely large number. Essentially, what the pharmaceutical industry does is have to sort through this massive amount of compounds. As you can expect, there’s a lot of research effort into figuring out if AI can help identify promising new compounds.

We took some of those algorithms and built a system that would continuously propose new ones. A recent court decision has now changed this, but at the time, our understanding was that if you publish those compounds on a website, then they’ll count as prior art. This means that if it does turn out to be a useful compound, a pharmaceutical company can’t patent it, and it remains in the public domain. That was the general idea—the fact that you can use AI for public research and protect the privatization of socially valuable things.



On your blog, you propose a version of Magic: The Gathering that incorporates an infinite set. What do you find limiting about the game’s existing version?

Magic: The Gathering was a big part of my childhood. I still play with my brother. It’s a relatively limited ruleset, but it can be endlessly entertaining. Even with all the variation that does happen, it can feel stale after a while. That’s why they keep making new cards and adding new mechanics—they don’t want the player base to get bored. I like games that are unpredictable in a way and pressure you to think on your feet or strategize on your feet. I like the creativity that comes with coming up with new strategies for a game like Magic. In a way, it’s a lot like programming. Someone wrote a blog post about how you can make a computer out of the rules of Magic: The Gathering. People have done all sorts of really wild algorithmic stuff with Magic, coming up with some of the largest possible numbers.

I thought it’d be interesting to have a meta-layer to that in which new rules and new mechanics and new cards are constantly printed or generated, such that you can never settle on a single strategy because the landscape is always changing. You would have to be really dynamic in your thinking. I enjoy that challenge.


An imagined infinite set for Magic: "The Gathering."

You’ve also explored what you call an ‘asymmetrical game.’ What would it mean to include multiple lenses within a single game?

Usually, when you play a game, it’s really limited to one perspective, in a sense. For instance, with Sim City, you’re usually playing as the city planner. You can go into an individual resident’s view, but you don’t really have any meaningful decision-making power in that context. I thought it would be interesting to show two games that look like they are two separate games, but they are one and part of the same game. 

Imagine in an arcade setting. You have one half of the room in this game that has something to do with finance or trading, and there’s another side of the game that’s more Sims-like, more at the level of a household or individual.

People playing that finance game—they’re having a blast, they’re trying to maximize their rate of return. Their decisions in that game, however, are screwing over the people that are playing this other one, to link these two systems that are presented as isolated and actually have them influence each other directly. The asymmetry there is that the people in finance have, without really realizing that that’s an important part of it, a ton of influence over the Sims-like side of the room. People on the Sims-like side of the room basically have no influence over how the finance people are playing. 

Can you tell us about your current role at the Jain Family Institute? Generally, how you see your research practice influencing your creative one?

At JFI, I have two main roles. One is more technical, building tools, and visualizations for ongoing projects there. Lately, that’s been a lot of mapping visualizations, satellite imagery stuff, which is creative in a sense, not in the same way as designing a game, but it is pretty satisfying. Especially because the topics they work on are important to me, so I’m happy to contribute to that.

The other side is the more open ended-research that I do. I want to understand the basis of production for all the material elements that make up our lives. I’m researching tetra chemicals in plastics, like climate controls. In particular, air conditioning is really interesting, especially with climate change and the role it has there. Then, remediation—how do we take all these damages, degraded landscapes, displaced peoples, all that and turn the clock back on that? What’s the process for reconciliation, remediation, regeneration? This connects to all the speculative future work that I’ve done. I really want to have a solid understanding of all the problems with how we produce our material lives today and what a more ecologically and socially-just version of that would look like in the future. What’s a rough sketch of a path to that? My focus is mostly on the engineering and technological side of it—what technologies compose that world. That’s kind of where my head is at now.

There’s a game that I’ve wanted to make for several years, but I haven’t had a concrete idea. It’s called Fugue, as in repetitions of a theme in music and as a dissociative state. The idea there is to tackle some of what I’m researching now. I’m interested in competing theories of change. Most of the internet arguments we see, at least on the left, are competing theories of change. For instance, people getting angry at others for pursuing electoral routes, or other people who would rather focus on mutual aid or receding into your own community as ways to better lives and a better world. This game is how I want to explore those two sides of the world.