When miHoYo released Fish Lantern, a Chinese New Year short for Genshin Impact, I found myself thinking less about the video itself and more about the studio behind it. I have played Genshin for 573 days. That is long enough to feel as if I have watched miHoYo grow in public, patch by patch, region by region. I had already been turning over some ideas about games; the short just made me want to write them down.
I am not an industry analyst. These are notes from a player who likes technology and games. Some of them may be naive. That is fine.
Why pay attention to a big company?
People often ask why ordinary players should care about what a large company does. I think a good company can have something like a personality and a soul. Apple is studied not only because it is big, but because people want to understand how it became what it is. miHoYo interests me in the same way.
Tencent is obviously a much larger business, and if I were interested in business models I would probably study it. But I am more interested in computer technology and games. miHoYo sits closer to those interests. Also, after playing Genshin for so long, I have a player’s attachment to the studio’s growth. It feels a bit like watching a character level up in a game I happen to be living in.
‘We have never been a game company’
The line I kept coming back to appears in miHoYo’s employee handbook: the company does not define itself as a game company, but as a technology company. Before I read that, several things about miHoYo puzzled me. Once I saw the line, they started to make sense.
The first puzzle was the emotional force of the characters and stories. Why were these anime-style characters so magnetic? Why did the plot get under my skin? The second was the slogan: Tech Otakus Save the World. What does that have to do with making games? The third was the sheer density of detail: elaborate story sequences, music albums, books inside the game. I do not read many books in real life, yet here I was reading lore in a game.
The handbook answers these questions in three ways.
An IP meant to last decades, not one game
A strong IP needs original content with depth, long-term support, and room for derivatives. miHoYo’s combined Honkai and Genshin universe is already absurdly large. Its characters inspire obsession; its stories leave players emotionally wrecked. The content also has layers. As a science student who has not read much literature, I sometimes miss what the story is doing. Genshin has even taken on a cultural role, carrying Chinese culture to a global audience.
Honkai has received content support for nearly a decade. Honkai Gakuen 2, an ancient game by current standards, is still running. That is evidence of long-term commitment. The derivative side is equally extensive: merchandise, collaborations, and events across the Honkai and Genshin lines. miHoYo understands how to use an IP. It also understands how to charge for it, which is a separate talent.
A technology company, a technology company, a technology company
miHoYo’s founders, Cai Haoyu, Liu Wei, and Luo Yuhao, all came from computer science or electrical engineering at Shanghai Jiao Tong University. The core of the company is still technical people, and you can see it in the games.
The pursuit of technical polish is visible. Honkai Impact 3rd began as what felt like a rough 3D action game from a third-tier studio. By its later stages, it felt like a solid 3D action game from a competent studio. Genshin shows an even clearer curve: from the plain, detail-light fields of Mondstadt to Liyue’s organic use of Chinese elements; from Inazuma’s improved puzzle design to Enkanomiya’s leap in map design and interaction; from the Chasm’s atmosphere and layout to Sumeru’s peak in story and map quality. Every major update has been an attempt to break the previous ceiling.
The improvements inside Honkai Impact 3rd, the jump from Honkai Impact 3rd to Genshin, and the improvements inside Genshin all point to rapid technical growth. The creators are chasing technical excellence. That is why calling miHoYo a technology company is not just a slogan. It has also invested in nuclear fusion research, which is a strange thing for a game studio to do unless it means the technology-company label.
miHoYo does not treat a game as just a game
This part of the handbook stayed with me. The meaning of a game for miHoYo is threefold. First, it is a carrier for content and IP: players meet characters, learn the story, and enter the world through the game, doing things that film and literature also do. Second, a game still has to be fun as a game. Third, it is a commercial product, but miHoYo wants payment to come from attachment to the IP and affection for characters, not only from the desire for bigger numbers.
That is a high standard, and it explains a lot about the company’s choices.
What should a game company be?
Here is where my view differs from the handbook’s framing. miHoYo spends a lot of effort proving it is not a game company. I find myself thinking: is that not exactly what a game company should do?
The phrase ‘video games are the ninth art’ is not widely accepted in China, at least outside younger generations. That is understandable. The domestic game industry grew in a distorted way, and the effects are still with us.
Compare games with the eight traditional arts.
- Painting: games need art, textures, and visual composition.
- Sculpture: games need 3D modeling, which is close to sculpture in a virtual space.
- Architecture: games need believable and functional architecture that fits the world.
- Music: games need music to set mood and bind scenes together.
- Poetry and literature: games need story, theme, and expression.
- Dance: games may need choreography and animated movement.
- Theater: games may need dramatic staging and performance.
- Film: games use light, camera, and composition, except the image is computed.
I am not saying a game is the sum of all eight arts. No game is top-tier in every dimension. But good games tend to be excellent in a few. miHoYo’s games are not the best in any single category, but they are decent across many and improving. Ubisoft’s Assassin’s Creed series has recreated landmark architecture at near 1:1 fidelity. GTA V uses its story to draw three characters and American society with startling vividness. Hollow Knight is remarkable for level design, art, and music.
The point is that a game is an organic combination of traditional arts, and it can do things those arts cannot. In a film, the camera path is fixed by the cinematographer and the story unfolds on the film’s clock. In a game, you are the protagonist. You control the camera. The story unfolds at your pace.
The ninth art has its own material
What makes games different from the eight arts is not only interactivity. At the lowest level, it is computer technology. Technology has driven games forward, and games have pushed technology in return.
I recently bought Valve’s classic Half-Life and Portal games. After playing them, I realized I had now experienced games from the 1990s, 2000s, 2010s, and 2020s. That is what led to the thought above.
In the 1990s, id Software used raycasting to make pseudo-3D games, lowering the computing cost of 3D and bringing it to personal computers. The company then built id Tech, the first game engine, which made 3D development much easier. It inspired others: Valve built GoldSrc, and Epic Games built Unreal.
In the 2000s, 3D engines matured. Valve released Source, Unreal moved to Unreal Engine 3, and Unity3D appeared. Development became easier and results more realistic.
In the 2010s, 3D game technology advanced quickly. Games from that era can still look convincing today. Unreal Engine 4 arrived and, with Unity3D, became one of the two best-known engines. New techniques entered games: ray tracing, AI, motion capture, VR, and AR.
In the 2020s, games can be nearly photorealistic. Some films now use game engines for visual effects. Game technology is a broad testing ground for new tech.
The relationship runs both ways. CPU and GPU performance improvements enabled 3D games in the 1990s and 2000s, and the popularity of 3D games gave hardware makers a reason to keep pushing performance. By the 2010s, hardware was powerful enough for sharper textures, finer models, and more realistic lighting, with enough headroom for ray tracing and VR. Game demand was a major reason GPU makers invested in ray tracing hardware.
So the most distinctive material of the ninth art is computer technology. There are other distinctive elements, such as level design and human-computer interaction, but technology is the foundation.
Back to the definition
If games are the ninth art, then a game company should treat games as the ninth art. That is my simple definition.
Examples support it. Ubisoft’s Assassin’s Creed team did field research and built a near 1:1 3D model of Notre-Dame; after the fire, it could serve as a reference for reconstruction. Rockstar’s Red Dead Redemption 2 reportedly cost 6 billion yuan and took eight years, producing absurd levels of detail; you can watch NPCs cook. miHoYo’s Genshin draws Liyue’s story, architecture, and landscape from traditional Chinese culture and presents Chinese qualities of patriotism, unity, peace, diligence, and self-reliance. Liyue’s prosperity mirrors China’s current momentum. Genshin is arguably the first game to successfully present Chinese culture and Chinese spirit to a global audience.
Does that mean only AAA games count? No. Literature has philosophical essays and light novels; both can be art. Games have many genres. Whether a casual game is art depends less on genre than on the attitude of the company making it.
Two faces of miHoYo’s attitude
miHoYo’s ultimate goal is not to be a game company, but its current business is games. So it is worth looking at its attitude in both roles.
As a game company
I once watched a two-hour recording of Shanghai game-industry bosses talking about success. I did not know miHoYo before Genshin, so I was curious: how did this studio become famous with one game?
The other bosses talked about the last big trend: first MOBA, then battle royale, and how to imitate the next one. They talked about Honor of Kings, PUBG, and Genshin’s revenue, and how to make a game to compete with them. They talked about DAU and how to keep it high; one person spent a long time on matchmaking algorithms.
The miHoYo founders talked about something else. They said that during the internet boom, many companies did not treat games as content, which led to chaos. They asked how to industrialize game production so that high-quality content could be produced continuously. They said they wanted to make content games, attract users through content, and let DAU follow naturally.
When other companies looked at recent hits and wanted to copy them, miHoYo’s founders said many of those hits were not worth learning from. When other companies looked at high-revenue games and wanted to compete, miHoYo was thinking about raising its industrial standards and sticking with content games. When other companies repeated DAU and data feedback, miHoYo kept talking about content: if the content is good, the numbers will come.
The contrast is stark. miHoYo’s founders look like craftsmen; the other bosses look like merchants. That is not an insult to merchants. It is a different job. But it explains why miHoYo seemed out of place in that room. The boss of Lilith Games seemed to understand miHoYo and expressed a similar attitude. I had not known that company before, but after hearing that, I trusted its games more.
As a technology company
miHoYo’s vision goes beyond games. It calls it the Honkai Divine Realm: a world, or several worlds, with immersion comparable to reality, where a person or a specific group of people would be willing to live. The ultimate vision is to build it.
That goal is enormous. It might require breakthroughs in mathematics and physics for algorithms that can simulate highly detailed virtual worlds. It might require advances in chip manufacturing to provide computing power for high-precision rendering. It might require neuroscience to support true artificial intelligence and the theoretical basis for brain-computer interfaces. None of that fits inside the category of a game company.
The vision is not pure fantasy, at least on paper. miHoYo’s planned route is to use games as a sustainable business to fund the development of a tool platform for industrialized, efficient content creation. On that foundation, it hopes to encourage humanity to build a virtual world more real than reality, resolving the conflict between Earth’s limited resources and unlimited human desire. That is the ultimate goal of Tech Otakus Save the World.
The company also knows where it stands. The handbook says that for now it is focused on efficient content creation, and that this task alone is so large that it will take an unknown number of years.
Whether the goal is reached is not the only point. The pursuit matters. If humans had never looked up at the sky and wanted to reach it, would we have invented aircraft, rockets, and lunar landings?
Is the Honkai Divine Realm possible?
The ideal
When I first learned about the Honkai Divine Realm, I was genuinely shocked. A world more real than reality, a place where the virtual and the real no longer matter, a world people might choose to live in: these ideas hit my worldview hard.
I wondered: if a perfect virtual world existed, a place where I could live without worry, would I choose reality or the virtual? The question quickly became philosophical. Which matters more, the body or the mind? With my current experience, I cannot answer it clearly.
But I can say this: it is a new direction. It sounds absurd. So did heliocentrism, Newton’s laws, relativity, and quantum physics. Each sounded absurd in its time and each changed human society. So whether the Honkai Divine Realm is possible or not, it is a direction worth exploring.
The handbook puts it neatly: SpaceX goes outward; miHoYo goes inward. Both are futures.
The reality
The handbook is still a company handbook. It may be a grand vision painted by the boss. I felt my blood run faster reading it; new employees must feel even more excited, as if they are about to do something historic. In practice, most people will still do their jobs. miHoYo itself admits it is still far from industrialized content production.
I do not think the founders are the kind of capitalists who paint a vision only to squeeze employees. The vision is more likely a promise they made to themselves and to the company.
But can it be realized? How many ambitious technical people start with a grand goal and end up as capitalists? miHoYo is now visibly caught in that tension.
An article titled When miHoYo Starts to Contract described Liu Wei’s speech at an all-hands meeting, where he reflected that the company had become inflated. He pointed to poor communication between senior and junior leaders; senior leaders sometimes did not even know what kinds of people junior leaders had hired. Many new employees did not read the handbook and did not practice its culture of saying what you mean, speaking directly, recognizing only results, and pursuing the extreme. Liu Wei said the company would pay more attention to passing culture from the top down and preventing values from becoming formalism.
Much of what I have described comes from the handbook. When Liu Wei says new employees do not read it, the implication is that many new employees do not share miHoYo’s spirit. More bluntly, miHoYo’s top-tier salaries and Genshin’s success attracted people who want to make money, and their values do not always match the company’s. That can pull miHoYo off course.
Genshin is popular around the world. miHoYo is at a peak of rapid expansion. And the founders are already reflecting on whether the company is drifting. That fact itself suggests the three founders have not yet lost their original intention.
People change. No one can guarantee miHoYo will never rot. miHoYo has anticipated that too. The handbook says that if one day the company really goes astray, employees should take the handbook, put it in front of the boss, and tell him to get lost. It calls the handbook miHoYo’s constitution.
A personal note
After all this talk about ideals, reality is still bony. I am an ordinary student at an ordinary 211 university. At my current level, I certainly cannot get into miHoYo. That does not stop me from learning from the company’s spirit: aim high, pursue excellence, keep improving.
A goal may not be reached, but it can still point you somewhere. So I will set a small goal: keep working, and try to join miHoYo after graduation.
Tech Otakus Save the World
