I am not Ready for The Future of the Game Industry
How prepared am I for the ways AI might reshape the game industry in the next few years? As a game developer and programmer, I’ve been thinking a lot about my place in a future where AI could fundamentally change how games are made — and who gets to make them. This isn’t about predicting the future (I have no idea what’s coming), but rather exploring the trends already in motion and asking some tough “what if” questions. What if AI makes game development faster, cheaper, and more accessible beyond anything we can imagine now? What if it creates entirely new roles while rendering others obsolete? And most importantly, what can I do to adapt and thrive in this uncertain landscape? In this post, I’ll dive into the potential impact of AI on game development, share lessons from past disruptions, and explore how we can prepare for the future — whether it’s exciting, terrifying, or a bit of both.
In the world of AI, progress happens at breakneck speed. The last four years have seen advances that would seem like science fiction to someone from the early 2010s. Given this pace, it’s entirely possible that much of what we do as game developers could be done faster and better by AI in just a few years.
This often leads to awkward jokes with my peers about being replaced by machines. Yet, despite the looming possibility, our day-to-day work hasn’t changed much. If such a clossal shift is on the horizon, shouldn’t it be prompting more drastic changes in how we plan for the future? It feels like even if this outcome feels so inevitable right now, it will still catch us by surprise if it comes.
Before I go further, let me clarify: this isn’t an endorsement of AI, nor is it an argument that AI will replace programmers, animators, or concept artists. I’m not here to debate the morality of training AI on massive datasets, or whether lawmakers, courts, energy constraints, or panicked investors might slow its progress.
This is a thought experiment. Imagine a world where, in three years, you could type a prompt describing a behavior for a virtual actor, and the AI generates flawless code with all the parameters you need. Or even further, imagine AI enabling people with no coding, modeling, or artistic skills to create better games than your team can produce today. What would that mean for us?
In such a world, we’d see an explosion of games. Development would become cheaper, and the variety of games would likely increase. While this sounds fantastic for players, it could be a death sentence for those of us who make games the “old way.” The term creative destruction, coined by the eccentric economist Joseph Schumpeter, perfectly captures this scenario. AI could create better, larger, and more enjoyable games, but it would also render many traditional game development positions obsolete as a trade.
Before we dive deeper into what could happen, let’s look at what’s already happening.
The Economics of Video Games
Gamers often complain that video games are too expensive, while industry experts argue that games are becoming more costly to produce. In real terms (adjusted for inflation), the price of video games has actually been falling. And while production budgets have grown (in both real and nominal terms), it’s becoming cheaper to make games.
Compare today to twenty years ago. Back then, if you wanted to make a game, you probably wrote your own engine (games and engines were alot more integerated and licensing might not have been an option), likely in a challenging language like C++ or C. You had to write your own libraries, port your game to every platform manually, and figure out most of the game development process on your own.
Fast forward to today. Engines like Unreal and Unity provide affordable access to software ecosystems that were once impossibel or insanly expensive to produce on your own. Tools like Blender are completely free. Even if you decide to build your own engine, there are countless free libraries available. Knowledge is also far more accessible. Higher education programs now teach game development skills, from art direction to graphic programming, in ways that weren’t possible even a decade ago.
The result? If someone wanted to recreate a game from 10 years ago, they could do it faster and cheaper today.
This drop in production costs (I will addresse the increase in the production budget later), combined with an expanding market (more players with more disposable income), has allowed the industry to thrive despite falling game prices. The market has grown thanks to the rise of personal computers, cheaper and different consoles, mobile gaming, and a burgeoning middle class in populous countries like China.
The democratization of game development — an industry that can be highly profitable — has led to an explosion of new games. The number of releases on major platforms like Steam keeps increasing every year¹, peaking at the current 18 thoasand new releases in 2024. You will see me refer to Mathew Ball’s state of gaming industry report a couple of times, give it a read if you have the time.
There’s a classic economics concept that shows the relationship between the quantaty of a good and its price. When supply increases without a corresponding rise in demand (supply is greater than demand), prices fall to the lowest point that production costs allow, due to firms trying to out compete each other.
The easier it becomes to make games, the more people will make them. The more games there are, the fiercer the competition for players’ limited entertainment time. And it’s not just competition with other games — it’s competition with movies, TV shows, books, and even Instagram Reels.
One way to stand out in this crowded market is to lower prices. This leads to what I call toilet paper economics — or crisps economics, if you prefer. It’s a market condition where goods become so commodified and easily produced that companies are forced to slash profit margins per unite to near zero, relying on massive sales volumes to make money.
In toilet paper economics, success hinges on two things: driving down production costs (mastering the supply chain) and convincing as many people as possible to buy your product. This not only increases marketing costs but also raises production costs, as companies often expand their offerings and appeal of the product with a shotgun approach to appeal to the largest possible audience.
To bring down production costs, companies might manufacture parts of their product in India, assemble them in Cambodia, secure tax breaks in Singapore, and fund R&D with grants from the EU. Competing in this kind of mass production requires significant upfront capital — it’s not a game for the average Joe.
This dynamic partly explains the economics of AAA games. For digital goods profit margin per unit is less relevant since goods can be duplicated at almost no cost. But the baseline price of games has been pushed so low that most playtime is spent on free-to-play titles like Counter-Strike, Fortnite, and Roblox. Meanwhile, AAA budgets have ballooned to hundreds of millions of dollars. Video game companies and their supply chains are as international as fast-fashion giants like Primark.
Take Skull and Bones as an example. Ubisoft received funding from Singapore² for the project, while Ubisoft Berlin, which helped finish the project³, got support from the state of Berlin-Brandenburg⁴. In today’s industry, even a healthy studio might shut down because Japanese investors — backed by insurance companies and fueled by Saudi sovereign funds — overpromised on post-pandemic player engagement.
All in all, working in toilet paper economics can feel a bit soulless and often drives my friends in the sector crazy. Sometimes I think it might be better to work in the actual toilet paper industry — at least you don’t get hate emails because your toilet paper isn’t “sexy” enough.
At the other extreme end of the supply-price curve are luxury products. Here, companies focus on a tiny fraction of people for whom the product is worth a premium. They sell fewer units but enjoy much higher profit margins per unit. Many indie game companies that have thrived for years find themselves in this space. This is also where most of my focus has been, both with Grizzly Games and Realities.io.
The key to luxury is creating some form of scarcity. For example, wine from a specific region of France is expensive because only a limited amount of grapes can be produced each year, given the unique soil, sunlight, and bacteria. In video games, indie teams create scarcity in different ways. It might be a unique combination of talent within the team, enabling them to create experiences no one else can replicate.
Or it could be the team’s small size and low overhead, allowing them to target niche markets that aren’t financially viable for larger studios. Whether the source of scarcity is innovation, luck, talent or god level marketing, successful products in this sector are unique and have no issue standing out of the crowd without arm racing over price. If you loved Hollow Knight, your best bet is to wait for Silksong.
If AAA toilet paper economics kills your soul, luxury indie economics gives you digestive problems. It’s incredibly stressful and not super sustainable. It could be sustainable if you are super lucky, and not stressful if you are ignorant of just how large luck contributed to your success! Consider wine as a luxury product again. If you own that piece of land, your source of scarcity is guaranteed (barring some huge change like global warming or a pan-European socialist revolution with land reform). But if your source of scarcity is the exceptional skill of your team, then for every new title, they need to remain at the top of their game. Well into their 30s, 40s, and 50s, they must outcompete 20-year-olds who have more time and a better feel for the pulse of pop culture.
There is a lot more to be said here, but the article is long as is. If you have a unique game, you can place yourself anywhere on that graph. Lower the price point, which will cause more sales. But unique games are usually unqiue because they focus and are targeted to very specific audience. So by lowering the price you pull in more people to whom the game is not all that unique. To appeal to them you need to add more things to it, and before you know it you went from making a specific game for a specific group of people to making Fortnite. It used to be that there was more going on in the middle of the graph, but that area has been getting pushed more and more to left and right.
AI Could be More of the Same, A lot More
While I’m not entirely sure what AI will ultimately be capable of, it’s clear that what AI can do right now will turbocharge the dynamics already at play in the gaming industry. On one hand, AI will make game production even cheaper, leading to more games, lower price per unit, and an even more saturated market. On the other hand, for handcrafted indie luxury games, even a smaller percentage of exceptional artists will be able to stand out. The pressure on studios that wish to maintain their position will intensify.
A creator’s unique perspective on life could still serve as a source of scarcity in such an economy, but I’m not sure how much that will be worth. Will a deeply personal vision be enough to compete in a world flooded with AI-generated content? Remember we are operating on the “what if” thought experiment that these generated content are actually as good as whatever you are creating. And how many of us truely have such unique perspectives?
Lately I’ve spent a lot of time thinking about what the future might look like for our team. How do we continue doing what we love in the world that’s coming? Frankly, I’m not sure I find the prospects on either side of the curve very appealing. The AAA toilet paper economics feel soul-crushing, and the indie luxury route seems increasingly stressful. At this point, moving to the actual toilet paper industry is starting to look more and more tempting!
But this is just how AI might amplify existing trends. What about entirely new changes? What can we learn from past rounds of creative destruction to understand how this one might unfold?
While there’s no proof that AI won’t eliminate all jobs, history suggests it might not. Take the cotton-spinning machine, invented in 1760. It might have destroyed 5,200 jobs out of the 7,900 people working in cotton textiles at the time. But by 1787 — just 27 years later — a parliamentary inquiry showed that the number of people employed in cotton spinning and weaving had skyrocketed from 7,900 to 320,000. That’s an increase of 4,400 percent⁵.
Similarly, the automobile revolution decimated the carriage industry, which employed an estimated 368,000 people⁶. Yet, by the 1920s, the automobile industry was directly employing 1 million people, with many more jobs created indirectly in related sectors like road construction, gas stations, and tools manufacturing⁷. Other examples of this pattern include the printing press, the computer revolution, and the rise of e-commerce.
Could AI be different and end all jobs? Maybe. But historically, there’s no proof that this will happen.
Another interesting lesson from history is that when new jobs are created, the established players rarely survive or adapt to the new environment — or at least, it takes them decades to catch up. Only a small fraction of silent film stars successfully transitioned to “talkies” (a handful out of hundreds)⁸. Kodak and digital photography, Blockbuster and streaming, and Nokia and smartphones are just a few examples of this phenomenon. Some of these shifts are still unfolding as I write. More recent cases include Intel, which over-relied on CPUs and ignored specialized AI chips, and Samsung, which lost a major market segment in the memory industry by being slow to adopt emerging technologies.
There are many reasons why this happens, and most of them tie into systematic biases embedded in how we think and process information. Established players often have too much invested in their current infrastructure (the sunk cost fallacy). They’re too accustomed to doing things the way they’ve always done them (after all, people are creatures of habit). The emotional impact of losing their advantageous position makes them dismiss the likelihood of disruption (probability neglect). The sheer scale of the disruption means they ignore what they can’t comprehend (scope insensitivity). And because the new changes contradict their mental model of how the world works, they resist accepting them at face value (normalcy bias and cognitive dissonance).
On the other hand, newcomers don’t have these problems. They have no mental models of how things “should” or “shouldn’t” be. True innovation often happens in areas that people consider impossible, and those who are neck-deep in established workflows are usually the last to discover it.
What to Do?
How do you proceed with all this information? It depends on who you are. In our “what if” scenario, some jobs will cease to exist, or openings will become so rare that, unless you’re in the top 0.00001%, you’re probably not going to get one. You can still hire a chauffeur or a charioteer today, but only a tiny fraction of people work in those roles compared to the early 1900s.
Accepting on a personal level that this might happen could be the first step. When I’ve discussed this with friends, the conversation often turns to how things should or shouldn’t be morally. My perspective is more along the lines of: Just because you make all the right choices, work hard, and are talented at something doesn’t mean the destiny owes you success and happiness. Sometimes, shit happens. Shouting at AI is about as productive as screaming at a hurricane or an earthquake. Market-driven liberal democracies are unlikely to pass up the productivity gains this technology is already bringing.
It’s worth thinking about what’s truly important to you. Did you start working in games because you love entertainment and want to be part of it? Good news — there will probably be a lot of new jobs created where you can entertain people in entirely new ways. Or maybe you’re here because you’re deeply passionate about coding, painting, or modeling, in the very specific ways we are doing it right now. In that case, it might be time to think about what other interests you have and how you can provide value beyond this industry.
I mentioned earlier that, looking at history, it’s likely there will be many new jobs in entertainment. If you’re aiming to fill one of those roles, how do you prepare for it? I’ve seen people try to theorize their way into predicting what’s coming and placing all their bets there. From prompt engineer (I still can’t even…), to moving to sections they believe won’t be effected by AI.
But if there’s one thing you can learn from the historical cases I’ve mentioned, it’s that no one knows what’s coming. Or if anyone does, it is pretty much by accident. It wouldn’t be much of a disruption if you or I could think our way through the future. There are marketing jobs in e-commerce that would have been inconceivable to the older generation. Software engineers today solve problems that people 50 years ago didn’t even know existed.Just to prove the point, look at this hilarious graph, showcasing how people thought VR will grow as a market the past few years.
If you can’t theorize your way to an answer in a purely abstract way, then what?
The only way to figure out what’s coming is to experiment with it. It’s time to get empirical. AI is already incredibly useful in how we do game development right now. Jump in, use it, build with it. Try to move past the cognitive biases I mentioned earlier and play around with it like it’s a toy you’ve just discovered. As you tinker with it, you might find that you can do fun, unexpected things that turn out to be valuable. Put those ideas out there and see if others find them valuable too — and just how valuable they are!
Besides that, consider not tunneling too hard on one segment for a while. Spread out a bit. Explore adjacent roles to what you’re doing now. Read a bit about business, finance, art, politics, history and more. All of that might help you gain new perspectives and discover interests you didn’t even know you had!
Diversifying a bit will also help if our “what if” scenario turns out to be wrong. AI could, after all, eliminate the majority of jobs in entertainment. Or it could create new ones that are terribly paid and not all that fun. After all, our historical examples were about the automation of manual tasks, not cognitive ones. The speed at which AI is growing right now is unlike any historical example I’m aware of.
Alternatively, the real toilet paper industry might not be a bad option after all. It survived fancy Japanese bidets; AI is unlikely to disrupt much there! Jokes aside, I realize I haven’t provided much answers. As usual, this post has been less about me having any answers, and more about thinking through a problem and trying to learn. What I have written on the last section is less an opinion about what others should do, and more a plan of what I could do.
References
- State of Video Gaming in 2025 by Mathew Ball: https://www.matthewball.co/all/stateofvideogaming2025
- Singapor and Skull and Bones: https://www.cnbc.com/2024/02/21/singapores-first-major-video-game-title-launches-to-mixed-reviews.html#:~:text=Ubisoft%20Singapore%20received%20a%20grant,title%20from%20the%20city%2Dstate.
- Ubisoft Berlin contributing the Skull and Bone: https://www.mobygames.com/company/47450/ubisoft-berlin/
- Berlin Senate Paying Ubisoft to expand in Berlin: https://www.gamedeveloper.com/business/ubisoft-berlin-granted-1-84m-in-subsidies-from-city-senate-for-expansion
- Economics in One Lesson Page 34
- carriage and horse industry: https://blogs.microsoft.com/today-in-tech/day-horse-lost-job/
- Automobile emplyoment: https://www.richmondfed.org/publications/research/econ_focus/2003/winter/economic_history#:~:text=The%20rapid%20expansion%20of%20the,automobile%20dealerships%20in%20the%20country.
- Grove, Andrew S. Only the Paranoid Survive: How to Exploit the Crisis Points That Challenge Every Company. Currency Doubleday, 1996. (Page 17 in the 1996 edition.)