Living in the Simulation

NOTE: This post is part of a select subsection of writing on my blog that I return to occasionally and should be considered a work in progress! If you’re interested in how my thinking has evolved over time, you can always look at the git commits for the post over on GitHub. First published: June, 2021 Last updated: June, 2021

Some very smart people have debated whether or not human life is just a simulation. It’s fun to joke on the internet about how everything we experience in life might all just be a simulation. But underneath every joke is some truth. Even if life is not a simulation itself, we’re still living in simulations.

Per Merriam Webster: Simulation a: the imitative representation of the functioning of one system or process by means of the functioning of another b : examination of a problem often not subject to direct experimentation by means of a simulating device

I thought a bit more about this idea of a simulation and realized that while it may be a bit of a stretch, but I think most video games are all simulations (some accurate, some not so accurate) in some form or another. People play a lot of video games and I think one could argue they already spend a huge amount of their time in simulations. But why? Why do humans choose to play games when they can live real life? To answer that, we can find some clues in a specific genre of games called (drum roll) “simulation games.” These are games that often simulate menial real world tasks such as truck driving, flying and even things like surgery. Many have been wildly successful. Job Simulator, Vacation Simulator, Microsoft Flight Simulator, Bus Driver Simulator, Surgeon Simulator, even SimRefinery (arguably the most obscure of all Sim games). The list goes on and on.

Why do people love simulation games? What is it about virtual simulations of things that is preferable to just…doing the real thing? Shouldn’t they be more fun in real life if they’re accurate portrays of behavior there? Let’s take a look at some of the benefits of simulations as we seek answers to these questions.

Simulations broaden access and increase basic understanding of complex tasks

Young people in America often hear “you can be anything when you grow up” from their parents. Many live in places that they don’t have access to see what that “anything” looks like. But increasingly, they do have access to the internet, which brings with it video games and social media. Social media can be helpful in providing access to anyone on the planet, but if you want to be an astronaut, there are only so many YouTube videos you can watch about space before you realize that you need more hands on experience in some form to become an astronaut. Thankfully, there are games like Kerbal Space Program, which features a realistic orbital space physics system (though is far from a 100% accurate space simulator). I don’t think we’ve seen any actual astronauts who grew up playing Kerbal (though NASA does play it!), but it is clear that simulations like KSP provide access and opportunity to experiment with complex things that could otherwise be inaccessible. Assuming they have a decent computer and the internet, kids can explore doing that thing they’re curious about without the permission or huge up front investment required instead of just watching YouTube videos about it.

Another net effect of simulation games around complex tasks is that we end up with more people interested AND proficient in a given thing. Popular simulation games can broaden the lowest common understanding of a particular topic. For example, a game like SimCity or Cities: Skylines breeds interest in city planning and increases knowledge and fluency of conversation around what city planning actually entails to those that have never experienced it in any form. This effectively increases the “potential” of the market for a particular job.

Simulations lower the cost of failure

Simulation games are commonly built around high risk jobs like flying and medicine and greatly reduce the cost of failure associated with such activities.

One of the original simulation games, Microsoft Flight Simulator, recently celebrated 35 years of development (it is Microsoft’s longest running active software product line!). Flight Simulator is so old it even has its own history page separate from its main Wikipedia page.

In the 80’s the cost of getting access to planes and learning how to fly was pretty prohibitive for the average person. It’s come down since but relatively speaking it’s still expensive. Even if one has access to a plane, the cost of failing at flying is expensive (both in terms of risk to human life and actual dollars of lost equipment).

In Microsoft Flight Simulator (a $59 piece of software), you can log thousands of hours (and failures) behind the cockpit of almost any aircraft on earth before you ever get into the seat of a real plane. This increases safety of real flying and can help users gain confidence that they otherwise wouldn’t have gained logging only real hours in a plane. Again, this dynamic helps to increase the number of people who have an interest in a particular skill and is a good thing. It is not to say it is a pure substitute for flying a real plane, but that paired with flying real planes it can be an effective tool (that is also just fun).

Many simulations are useful as a precursor to or supplement for real world tasks. But not all need to be paired with reality.

In extreme cases, one might even go so far as to say that some FPS games like Arma 3 are “military simulations” - they give us the ability to experience what it’s like to be in the military with relatively low downside (we’re not going to actually die). However, many of the players of these games won’t even shoot a gun in real life and such a simulator lets them experience this without having to put themselves in a situation that is truly life-threatening.

Simulations are vehicles for flow states and mastery

What about more…mundane simulations? Why do gamers love something like Bus Simulator or Power Wash Simulator? I would venture to say that one reason these simulation games are popular is because relative to something like Kerbal Space Program where the learning curve is high, they are relatively easy to learn or understand from the start. As a result, these games can be easier to achieve a flow state in and therefore to get better at. This can provide a user with a sense of mastery over a task that may be lacking in their real world circumstances. There are several examples of people in the real world making games out of their tasks - take, for example, a factory worker making a game out of his/her repetitive task as a way to get better and continue improvement (Amazon does this, and yes it’s somewhat dystopian). Simulation games play on this even further but add measurement and progression where the game itself can help engineer these flow states precisely. There is probably an entire category of software that could take cues from this genre of games to help users be more effective, creative, and fulfilled.

Simulations are a canvas for creativity and fun

Simulation games often revolve around something that is inherently predictable or rigid in some way. That predictability is somewhat of a constraint, but it also provides a canvas for creativity. Job Simulator makes a game out of menial and boring tasks (in theory) but uses these task constraints to come up with creative and fun mechanisms to keep the task itself fresh. Even Kerbal Space Program has some constraints and limitations that provide provide the user some canvas on which to experiment - knowing what is a constant makes experimentation easier. This translates into real world creativity: knowing you’re going to get up every day and write for an hour makes it easier to not worry about what you “have” to write - you can just let your writing flow and be free of expectation. This becomes much more possible with the correct set of constraints. And simulations, with their rigid rules, can provide those constraints as grounds for experimentation.

So where are simulations headed in the future and why are they even more exciting now?

So what’s the point of writing about all this? As we continue to push what our definition of a “game” is, I think we’ll an increasing number of games that incorporate these aspects which will become more than just toys.

Simulations will be use to improve human performance. The Statespace team has built an FPS aim training simulation called AimLab using cutting edge cognitive science and AI under the hood. The game effectively simulates other FPS games and their weapons, physics, and more and makes it easy for players to “practice” in the context of those other games. AimLab incorporates a lot of the design principles that make simulation games fun and represents a real opportunity to advance human performance in the brain. When combined with hardware, simulations like Zwift also act as simulations designed to help humans maximize their performance physically.

Simulations will increasingly leverage human input for difficult or complex computational problems. Given humans’ love of games and simulations, more games could be used to outsource elements of complex computation and rely on human input to solve complex problems (these are calledhuman based computation games). This could have useful applications in robotics or in complex economic systems. This type of behavior exists today - Facebook has effectively built an “attention simulation game” used to capture human attention and uses human inputs from the game to power their advertising engine. While some might debate whether this is predatory behavior on behalf of Facebook (there are certainly elements of the product that lack empathy for the user) the user is getting something (content!) in exchange for the data Facebook collects.

With the increasing fidelity of graphics and input devices (via VR, etc), simulations are poised for a host of other interesting uses. One potential implementation of such a human based computation game might be in training real world AI models and robots with human input. This tweet shows an augmented workforce technology, but it would probably be more fun for the human if it were a game. A simulation game that uses human VR input could thus be used to train robots at scale. Several companies have been exploring these use cases but no one has built a mass market game that leverages human input at scale. I suspect we will see one at some point, especially as immersive computing takes off.

Ninantic is effectively building a human powered computation game via their AR platform. Pokemon Go can be thought of as a “Capture Pokemon!” simulation of sorts. The game is based on a fairly simple, repeatable (constrained) premise, and the complexity in the game revolves around the real world locations one visits. Ninantic is likely capturing a host of important mapping data from users who play these games and can leverage this in a bunch of interesting ways in the future.

Simulations built around virtual economies will inform real economies. In 2007, CCP Games (the studio that makes EVE Online) hired Dr. Eyjólfur Guðmundsson to study the in-game economy of EVE Online. In 2012, Valve hired Yanis Varoufakis, a Greek economic theorist (who later became the Greek Minister of Finance) to study the CounterStrike skins market. There is no denying that even in their current basic form, game economies embody some of the qualities of real world economies. In 2021, Web 3.0 and crypto networks are fertile grounds for experimentation - automated market making and smart contracts are already being deployed into economic models in games. These coming in-game economies will be more complex simulations of real world economies than ever and there will be much we can learn from them to advance our understanding of real-world economics.

There are probably a host of other things that simulations can do for us in the future (or that they are doing for us already). I will keep adding them here but for now (in case it isn’t obvious) let me just say that I am very excited by all of this and if you are building a game/simulation or have thoughts about any of this I’d love to talk with you!