Dialogue with Ma Xiaoyi from Tencent: AI outlets, globalization, and the next revolution in games

Source: That thing about mobile games

Author: Ben

I don't know when it started, people are always discussing Tencent.

You should have seen similar reports: Tencent’s anxiety, Tencent’s heart disease, Tencent’s wavering, Tencent’s silence, Tencent’s inability to sit still, Tencent’s inability to figure it out...In the increasingly competitive game industry , Tencent’s every move has attracted much attention. In the face of rising stars, everyone seems to be used to asking frequently, how will the big brother of this industry respond?

Taking advantage of the opportunity of the Spark 2023 Tencent game conference, several industry media, including mobile games, chatted with Ma Xiaoyi (Steven) about their understanding of the essence of the game, investment in new technologies, overseas The internal and external layout and thinking, and even talked about Mihayou.

(Ma Xiaoyi at today's press conference)

Let’s talk about the situation of this year’s press conference first. In terms of products, Tencent has produced a total of 35 game products and innovative content, including 20 new game products and innovative content.

(Spark2023 Tencent Game Conference)

Not only are there all "spots" that can be launched or tested this year, but also products that have been verified in the global market such as "Fearless Contract", "Ark of Destiny", "Legend of the Seven: Battle of Light and Darkness".

In terms of game technology, Tencent cooperated with China Southern Xiangyi to build China's first fully self-developed full-motion flight simulator visual system, digital central axis small universe, digital Great Wall, digital Buddhist scripture cave and other projects, using case by case , to verify their idea that games are super digital scenes-to sum up in Ma Xiaoyi's words, the popularization of any new concept is a process of seeing is believing and a matter of course. Only by being seen can it be truly perceived and then recognized.

Many times, people focus too much on some specific products, whether Tencent is anxious or not. But in fact, Tencent has its own strategy, persistence and long-term thinking.

(Digital Central Axis Small Universe)

After the entire press conference, you will find that Tencent can play a lot of bright cards alone, and they still have a lot of high-potential products that are being built with longer-term investment and thinking.

As Ma Xiaoyi mentioned in the interview: "The most discussed issue now is the thinking of our method, not the stress (reaction). What should Tencent do if we think that there are good products on the market? It is not this problem.” He said, “but when the product is good, we have to see if our method is right.”

01. Talk about the essence of the game

"When you see the application of game technology in the digitalization of cultural heritage and the application in the visual system of flight simulators, you will know that the original game can still be like this. Only when you are seen can you be truly perceived and then be recognized. agree."

**Q: Recently, everyone has started discussing "game technology". How would you define this concept? **

Steven: The concept is very good. Because the new experience and new effects that the game is constantly chasing are essentially based on cutting-edge technology. During this process, the game drives the development of new technology, allowing it to be applied quickly and on a large scale, and feed back the investment in research and development through the market. So I think that in two paths, game technology can play a great role in stimulating other technologies in the real world:

On the one hand, the game itself is the most hardware-intensive digital product that ordinary users can accept daily. For example, in the 1970s, Atari sold tens of millions of game devices, which was much higher than that of personal computers at that time, which led to a large-scale eruption of the chip industry; another example is the relationship between AI and graphics cards. Nvidia is a graphics card Born in China, graphics cards are made for games, because games have extremely high requirements for computing power, require 3D environments, and require various details, forcing graphics cards to make breakthroughs one generation after another, promoting the birth of GPUs, until today , GPU provides large-scale concurrent computing power, enough to support large language models, and can generate human-like intelligence. In this era, we start to discuss AI.

It can be seen that the connection between games and technology is very direct. This is a two-wheel drive concept. With the development of hardware, games have more demands and a better experience, so better hardware is needed. These are two very important fields that have been driven since the digital age. From this point of view, game technology itself has always existed.

(Full flight simulator visual joint project)

On the other hand, the game technology itself can be directly overflowed. For example, the flight simulator vision system we cooperated with China Southern Airlines Xiangyi needs to generate large-scale real terrain. If it was done by hand in the past, the accuracy was limited. Now we use the Tencent game CROS self-developed engine to create an area of hundreds of square kilometers, which is large enough to train pilots to take off, land, and cruise in various environments, including simulations of various extreme situations. This is a typical direct export of game technology to other industries.

Therefore, the word "game technology" can be said to describe the relationship between the game itself and technology very vividly, and it is inevitable that it will be paid attention to and discussed. It often takes a process for human beings to establish and understand things. With the global attention and discussion on cutting-edge technologies such as chips and AI; it is also accompanied by the spillover of game technology to multiple industries, and more substantive and tangible specific projects appear. Remind and inspire more people to understand game technology.

Overall, our understanding of game technology is that with the development of technology and the continuous improvement of user needs, the game industry has gradually accumulated and precipitated a set of unique technology systems that can simulate the real world, provide immersive experience, and have rich interactive capabilities. We refer to the technological clusters that occur in games to achieve a better game experience and continue to improve with the development of games as game technology.

**Q: In addition to the flight simulator visual system in cooperation with China Southern Airlines Xiangyi, we also saw cooperation on the digital central axis, digital Great Wall, and digital Buddhist scripture cave at the press conference. What is the starting point of Tencent Games for these projects? **

Steven: Technology drives the evolution of cultural expressions, and we think that's what games are really doing. With the advancement of technology, we hope to express this expression better.

In fact, any major change or advancement in any industry requires at least two preparations: first, whether the understanding is in place; second, whether the technological maturity has reached a critical point. Two years ago, we put forward the concept of "super digital scene". The support behind these scenes is game technology, and what is reflected on the user side is the super digital scene.

("A Tour of the Great Wall" (Time and Space Edition))

Why do we keep saying this is a "digital scene" instead of a game? Because many of the scenes shown have jumped out of the game itself. We feel that these projects like the ones mentioned just now are opportunities to embody digital scenes. They are all realized with game technology. Our team collects data in the real world and performs digital restoration and replication.

This is a good demonstration of the possibility of what the digital scene we imagined should look like. At the same time, the Digital Great Wall (space-time version) and the visual system of China Southern Airlines’ flight simulator are made with Tencent’s game CROS self-developed engine, which also allows our self-developed engine to have more evolution in non-game projects. In the future, we will explore more possibilities in more directions; conversely, we may be able to explore some gameplay in these possibilities, which is our various considerations for these projects.

**Q: It has been two years since the "Super Digital Scene" was proposed, when do you think people will really recognize this concept? **

Steven: The super digital scene is our open thinking about the nature, value and possibility of games. The underlying driving force of this kind of thinking and understanding is the technological attributes of games and the development trend of game technology. Moreover, the capabilities and characteristics displayed by game technology are very suitable for more current and future social needs. I think understanding of this concept requires progress in three areas:

First is the scene. The games we see today are relatively single, and I think the current game industry is still in its early stages;

Then there's the real technology. To improve the user experience, technology must be developed. The best example is the VR immersion in "Ready Player One", which can be said to be a truly super digital scene. But there is no such technology today. As we talked about before, VR may not really break out until 2026. Other more immersive experience methods may not appear until after 2030, and we are still waiting or cultivating these technologies;

And finally the content. When the technology develops, there needs to be content to let everyone know what the technology means. I have also given an example before. Nvidia's Rage TNT (3D graphics card) appeared in 1998. At that time, many people did not know what 3D technology meant to ordinary users.

Later, we first saw the rise of shooting games, making full use of the immersion and freedom brought by 3D graphics cards; and then because of 3D technology, we saw the emergence of editors. Because 3D games are more complicated, everyone develops tools The tool set is used for research and development, and the tool set becomes an engine, which eventually evolves into an editor; and with the editor, everyone will see "CS", "DOTA", as well as "Valorant" and "League of Legends" that we know today Wait for the game.

It can be seen that when a new technology emerges, it takes time to find a game content that is suitable for carrying this technology and to demonstrate the advantages of the technology. When these aspects continue to develop, the entire super digital scene will become more and more clear.

(Digital Sutra Cave)

To put it simply, the super digital scene is all the possibilities based on game technology. It can be the game products we can see now, or more new game forms that have not yet been created, or it may be based on the game A purpose-built digital experience space, or digital solution, of technology.

Because of this kind of thinking, we have expanded the scope of our exploration, otherwise it would be hard to imagine the emergence of these game technology "cross-border" projects we do. Because these cooperations seem to have no intersection, according to the traditional understanding, there is no inevitability at all. And when we think that games are a super digital scene, we can see the versatility and more possibilities of game technology, and we have these practices, and the social response is quite good so far.

In fact, the popularization of any new concept is a process of seeing is believing and a matter of course. The focus has never been on the concept itself, but on the solid practices under this concept. When you see the application of game technology in the digitization of cultural heritage and the application in flight simulators, you will know that the game can still be like this. Only by being seen can it be truly perceived and then recognized.

02. Talk about technology investment

"Games stand at the crossroads of culture and technology. The essence of the game industry is how to present technology to users. The technology may change, but our position remains the same."

**Q: From VR, cloud games, Metaverse, to the current AI. The game seems to catch up with the trend every round. To what extent do you think AI can change the game industry? **

Steven: I remember when I was talking about VR and Metaverse, I said that I was a little more pessimistic than most people in the short term, but I was a little more optimistic in the long run, and it is the same now. In fact, whether it is VR, cloud games, metaverse, or AI, I think as long as time is given, three, five, or ten years will greatly change the game industry. Specifically for AI, there are two paths:

One is the end of the production link, which can greatly improve efficiency. I may be the first person in the industry to talk about industrialization. Industrialization itself requires tools, and AI is a very good creative tool. It is conceivable that with the substantial increase in productivity brought about by AI, the industrialization level of the game can be raised to a higher level. This can be seen in the next few years, but it may not be a leap-forward development, but a step-by-step breakthrough, so the slope of the growth should be a steady increase.

The second is the gameplay level. What new experiences will AI bring to users? Regarding this aspect, I think it is difficult to answer today, because if you gradually add it to the game, everyone will not have such a clear perception. So it is very likely to be a leap. Maybe one day you will suddenly find that a new AI-driven game has appeared. Maybe in three years or five years, we will see more and more examples.

**Q: Some people say that AI is the industrial revolution of the game industry, which will make people feel anxious. What do you think of this statement? **

Steven: I think the emergence of AI has made the game industry have higher requirements for practitioners.

This is not a cliché, I have spoken to many companies in the industry about these views. You will find that it is particularly amazing that when it comes to AI, what companies say they want to reduce is often the manpower that they are relatively good at. Why? Because when a tool evolves, capable people will be able to amplify their abilities, which will make adjustments to ordinary people and ordinary works. I think this is the truth that AI has the most impact on the industry.

**Q: The development of science and technology is changing with each passing day, and the cycle becomes very fast. What do you think is the core technology unique to the game? **

Steven: I think games stand at the crossroads of culture and technology. The essential core of the game industry is how to present technology to users.

When we used to take MBA courses, there was a saying called FAB. Translated into Chinese is "because, so, that is to say". For example, because a Coke cup is big enough, you can pour two cans of Coke into it at a time, that is to say, you don't have to go to the refrigerator to get Coke. The reason why users end up paying is because they don’t want to run to the refrigerator all day long.

For the game industry, I think its biggest advantage is that we stand at the forefront of users, and we can always translate the features provided by technology and culture into the best experience and the best benefit for players. The technology may change, but our position remains the same.

**Q: As for Tencent games themselves, what attempts have you made in AI? **

Steven: First of all, we disassembled a lot of links in the tool chain of game development, and evaluated whether each link can be enhanced by AI. At present, it has been identified that many links can be improved with AI, such as the UI design of the game, the adjustment of the original painting, etc.; for example, we can already use AI to fill the gap between the actions of the characters. In the future, everyone will see the actions of our characters. more and more natural. All toolchains, we're doing it seriously.

Secondly, we are also exploring on the user side. Since the entire tool chain is so automated, can it be turned into a UGC tool? Because there are some examples in the industry that rely entirely on creativity. For example, the success of "Rocket League" stems from a pure creativity. In fact, players often open their minds, but they may not have the conditions for game development. With AI blessing and a simple tool chain, players may one day be able to use these tools to create games.

I think this is to bring the game industry back to its essence, and the threshold for creative realization will become lower, which is a great benefit to the industry.

**Q: Will these layouts and applications in AI eventually reduce the difficulty of making 3A-quality games? **

Steven: Definitely. For example, for a high-quality original painting, it may take us 15 people/day to make it. Now with AI, it basically only takes about 3 hours.

This is just one of the links. If more links can be accelerated in the future, I think 3A quality games, or games with large investment will be greatly improved in efficiency. As for whether the cost will drop? Not necessarily, because when everyone's efficiency improves, an arms race will begin. This is a good thing for players, players can see richer details, larger worlds, and higher-quality art. And for this industry, it can take another step forward.

**Q: In the game industry, what do you think AI cannot solve? **

Steven: AI can help us create more content more efficiently. But whether it is gameplay or high-quality content, the baseline still depends on the level of team members. If the people in your team are 9 points, they can use AI to create a 9-point product; but if your team is 6 points, the product is only 6 points. Therefore, the upper and lower limits of the product are determined by the personnel of your team. In this regard, AI cannot replace the role of humans.

**Q: In fact, the outside world is very curious. For example, you will say "Moon Landing Project" and make high-quality games; but some companies will say that the post-Genshin era should not be an open world. As a head manufacturer, do you think it is necessary to pursue these things? **

Steven: First of all, it is not easy to make high-quality games. Secondly, if you look at the review, I think that in today's more mature market, the water mark for success is very high.

Looking back at this industry, for example, from 2013 to 2016, although the competition is fierce, we may be able to achieve good results by using the popular gameplay + IP in the market. But in a mature market, because the water mark of success is high enough, you are only a 60-point product, which is not enough to win a large number of users. Perhaps today, only products with a water level above 80 points can emerge from the water and be successful.

However, products with a water mark of 80 points often require a team to accumulate in the industry for a long time.

**Q: So you must make high-quality games, and you must "go to the moon"? **

Steven: Yes.

03. Talk about the competitive landscape

“Today we see that the so-called border of this industry is far from the border. In the foreseeable future, such as ten years, the game industry may still triple to quadruple.”

**Q: Let’s talk about overseas first. Looking back, among the overseas studios invested in recent years, what impressed you most? **

Steven: In fact, there are many studios that have shown very good standards in the past one or two years.

In terms of holdings, such as last year's "V Rising", we did it with a small team in Sweden. They are in a very small city, how small? There are a total of 35 of them, and their CEO and I stand on the balcony of the company, and we can find out which house each employee is in, but they have done it very successfully.

Another example is "Darktide", they do it from the field they are good at, and it went online at the end of last year, and the results are also good; there is also Riot, which is good at making competitive games. "Valorant" reached a very high level last year, and it is already top-notch E-sports games.

(At today's press conference, "Fearless Covenant" announced an internal finalization file)

In addition to holding shares, we have also invested in some studios, such as FromSoftware, and the results are very good. If we look at the list, we will find that among the new games released every year, there are more and more games related to Tencent.

Of course, the global market cannot only look at PCs and consoles. Internally, we will divide the global market into three battlefields:

The first is China. It has almost 600 million users, but it is just in the transition period from high-speed growth to a mature market. You will see that users have higher and higher requirements for quality, and the pursuit of categories is becoming more and more scattered, and each category is basically There are leading companies occupying them.

The second is Europe, America and Japan. In the mature market in the true sense, their user scale is smaller than that of China, maybe 500 to 600 million, but their revenue is more than 2 times. The main competition in the market is PC and console, and mobile games only account for 1/3. And a lot of gameplay is transferred to more markets after the success of these markets, so we have invested more teams there. We hope they succeed, and at the same time have more explorations in gameplay, helping us create new categories globally.

The third is emerging markets. Countries like Latin America, Southeast Asia, and the Middle East are all rising rapidly. Those markets combined have almost 1.2 billion users. With so many people, their income may only be half that of China. But compared to the markets mentioned above, emerging markets have more opportunities for new dominant gameplay, which is called "national-level games" as we used to say. For example, our "PUBG MOBILE" is a national-level game in many emerging markets.

In these three markets, we will have different layouts and try to seize different opportunities. Of course, if these opportunities are successful, they can also be transformed into each other.

**Q: In the past two years, it can be seen that your overseas efforts have continued to expand. What is the current plan? **

Steven: First of all, China is moving from a fast-growing market to a mature market. We increasingly feel that the quality and taste requirements of Chinese players are consistent with those of the global mainstream market. So now we also require internal products to target the global market. We will not make products for a single market. Even if the product is a Chinese IP, we hope that the art style and gameplay can be more adaptable to players all over the world.

Secondly, every time in our financial report, we will talk about the proportion of overseas game revenue, and the proportion of our overall revenue is constantly rising. We don't pursue how much we can achieve all at once, but we hope to increase steadily and continuously, and hope that this ratio will eventually exceed a little more than half.

In the end, we don't want to increase the income ratio through crowd tactics. What we want to explore is greater success. It can be seen that in recent years, whether overseas or domestic, the speed of Tencent’s game release has slowed down, because we want to invest more in some products with a higher upper limit of success. Maybe its success rate is not so high, but we know Games are always risky, and such products are more valued internally.

**Q: In terms of domestic distribution, what is your current core competitiveness? **

Steven: We are evolving from a platform company to a product company.

Back before 2015, Tencent Games existed more as a platform in China. It can be said that we defined a platform with social networks. At that time, there were more light users.

As the game market becomes more and more mature, when categories are shaping the industry, you will see that the role of social networking is actually declining, while the quality of the game itself and the role of categories are increasing. During this process, Tencent Games has followed this trend and transformed itself from a platform company to a product company, so we invest more in game categories and focus more, rather than doing all types of games.

**Q: The global game market has been declining in the past two years, and everyone has been more or less affected. As the head, what is your method for seeking breakthroughs? **

Steven: There are actually many ways. Everyone said, is Tencent very anxious about Mihayou? In fact, these new external companies provide good ideas. Internally, Tencent is good at gameplay. We always believe that this is the main driving force of the industry. If you want to expand the user base, a fun core gameplay is still very important.

In terms of gameplay, the SOC we are talking about and the more open world categories we have seen have provided good ideas. There are many games in the industry that have already shown their potential. These evolutions may explode one day, and the industry will expand a lot in a short while. We will continue to make breakthroughs in gameplay.

But on the other hand, breakthroughs in gameplay are not common. If there is no breakthrough in gameplay, if there is more content to be filled, then the content-driven method is needed. In this regard, MiHoYo has taught the industry a good lesson, and we also need to learn this about what is content-driven. I personally think that gameplay-driven and content-driven are the left foot and the right foot. An excellent game company must be able to walk with both feet.

As for Tencent, our layout is similar to this route, and we are optimistic about it in the long run. I personally think that today we see the so-called border of this industry, which is far from the border. In the foreseeable future, such as ten years, the game industry may still triple to quadruple.