1 Currently, AIGC is redefining the entire process of game development and continues to influence related fields. At the same time, many related jobs are facing the dilemma that they may be replaced, but there will also be more demands for different forms of game jobs.
2AIGC brings a new round of changes to the game industry, which will effectively reduce the cost of game development and bring new opportunities to more small and medium-sized manufacturers. The future game field will no longer be "the world of giant companies" .
3 Through AI, the cost of large games, especially AAA games with beautiful graphics, will drop to a level that small studios can afford. Therefore, combined with the development of cloud game technology, end games with higher image quality requirements may achieve greater development.
When the wave of AIGC comes, all walks of life may face the reshaping of competitiveness, and the game industry is no exception. On the one hand, topics such as the replacement of the original artist have repeatedly rushed to the hot search, on the other hand, game innovation attempts such as enriching NPC dialogues with AIGC content are also endless. Under the influence of AI, what kind of disruption and innovation is the game industry facing? Are there any opportunities for new giants to emerge in the industry? Will the game we see afterwards be completely different? At the BEYOND International Science and Technology Innovation Expo held recently, Tencent Technology had a dialogue with Zhang Junbo, CEO of Unity China, and interpreted these issues.
Zhang Junbo, Executive President of Unity China
The following is the content of the conversation:
AIGC's impact on the game industry: the future will no longer be "the world of giants"
**Tencent Technology: With the popularity of AIGC, many people are paying attention to whether the jobs in the game industry will be impacted and affected. What kind of impact do you think will be brought? What kind of opportunities does the outbreak of AIGC mean for the entire game industry? **
Zhang Junbo: On the whole, this wave of AIGC has a relatively large impact on the positions in the game industry combined with the design direction. For example, original painting, art and other positions, in fact, many companies have already started laying off employees in related positions,** and the layoff rate is 50%-80%**, the proportion of layoffs is quite large. If a game company with 10,000 people can earn 10 billion a year. But we assume that using AI to participate together may only need 1,000 people. AI can further "reduce costs and increase efficiency" for companies.
Next, positions such as modellers, terrain editors, and game testing can also be replaced by AI, and people who write codes may also be replaced. If the code written by Copilot is better, jobs that require manual coding Probably less and less. In addition, game planning positions may also be affected, such as script planning, many game scripts may be written in GPT. There is also a possibility that this will force the job requirements and content of these related directions to be more complicated.
In general, all practitioners in the game industry may be affected to a certain extent. However, based on the outbreak of AIGC, the scale of the game industry will further increase. From this perspective, the number of jobs in the game field may further increase, but the direction of job demand has changed. Here will be several development trends:
**First, game companies in the future will not necessarily be concentrated in a few big companies and other giant companies as they are now. Because the cost of making games is very high, small and medium-sized factories can rely on AI to invest moderately in the future, and they can also make very good games. AIGC may increase the popularity of game development in the future. **
It used to be that the research and development of some games required the investment of hundreds of millions of dollars to make a "masterpiece". Now, the cost can be reduced, and a very beautiful game can be made with tens of millions of dollars. There will be more creativity, players and planning. These can be integrated into your own art style. In terms of gameplay, with the help of AI, it can also bring about certain innovations. More exquisite games may appear, and the lower threshold means more new players will enter the game. **AIGC brings about changes in game creation and subverts the process of game creation. It may bring another wave of start-ups and start-ups to dominate the market and give everyone more choices and more opportunities. **
**Second, the future game ecology will be more diverse and rich. The addition of AIGC will enrich the content of the game and generate more innovations. **
**Third, there will be more demand for game jobs in different forms. **Maybe in the short term some game practitioners will be replaced due to the impact of this wave of AI, but it is impossible to replace all people 100%. There are still many tasks that artificial intelligence cannot accomplish at present. For example, programmers, their job requirements are not only to write code, but also include data transmission and other engineering issues that still need to be dealt with, as well as some work that needs to be done manually, which cannot be done by GPT. For another example, for a large amount of 3D content, rendering processing is also required. This process is very important, but there is no easy-to-use artificial intelligence rendering technology. Artificial intelligence can only give approximate results at present, and we still need a lot of Engineers to do things that artificial intelligence cannot.
In general, with this wave of innovation, more games may appear in the future, and there will be more opportunities to enter the game.
For the game engine, we will integrate many AI capabilities to make the development process more efficient. We will not make a separate AI product, but will integrate AI technology into our various capabilities to help developers use AI functions anytime and anywhere, and they can be replaced by AI to improve work efficiency. Integrate for the purpose.
In addition, in terms of the use and guidance of developers' functions, there may be many documents and functions, etc., which developers may not necessarily use. Technical support or manual learning is required, and the cost of occupation is very high. If it is possible that AI can help us, let developers understand functions faster. For example: we have a technology stack called DOTS. DOTS is different from traditional programming. It used to be an object-oriented programming language. It is not the optimal data structure in graphics. Data-oriented programming is actually a better structure. more efficient. The problem with data-oriented DOTS is that DOTS code is difficult to write, because it needs to change the programmer's thinking. If with the help of AI, programmers can use natural language or object-oriented language to describe, AI can help programmers generate inverted codes. If you want me to do the first (AI fusion) thing, this is the first thing I want to do.
Game form brings innovation: end games and TV cloud games will rise
**Tencent Technology: Games are divided into mobile games and console games. Do you think AIGC has the same level of impact on these two application scenarios? For example, among all the impacts, what is the impact on mobile games and console games? **
Zhang Junbo: From the perspective of AI, there is not much direct impact, but under this trend, end games (PC games/console games) may rise further.
This is related to its production cost. From the point of view of mobile games, the screen of the mobile phone is small, even if the pixels are a little lower, the effect is not good enough to play. However, if you want to experience console or PC games, etc., to experience games with good effects, the requirements for pixels and image quality are very high. Correspondingly, the production cost is very high. For example, the experience of users playing 720P games on TV is very bad.
However, if more players join in through the combination of the latest AI technology, the overall R&D cost will be reduced, and a lot of technology in the game industry will be mentioned. For example, the future game standard has become 4k or 8k games, and the game experience in the TV scene will definitely explode further. In general, if there are more and more high-definition and high-quality content in the future, end games will rise, and cloud games on TV will become the mainstream. **
**Tencent Technology: In terms of product form, in March this year, Unity CEO John Ricitiello also mentioned the possibility of using generative AI for NPC dialogue. In the future, in what form will generative AI and game development be better integrated? **
**Zhang Junbo:**There is a very natural connection and application scenarios between artificial intelligence and game development. For example, the NPC in the game, how to describe and simulate real players is one of the most suitable fields for AI.
In addition, AI can also be used to drive characters in games, and drive virtual characters, such as your Avatar or digital avatar, through the facial expressions and language of real people. These are all relatively feasible, and we could see them implemented soon. **The application scenarios of future NPCs can be wider and more subdivided. **
At present, it is easier to deal with NPCs in fixed fields, and its actions and routes are fixed, but it is still difficult to deal with in complex environments. Especially NPCs with special status. For example, in the open world of the World of Warcraft game, AI may support NPCs to form the same growth path as players, so that the equipment of NPCs and the equipment played by players can simulate the real situation, rather than simply setting it. This requires more development experience and technical support. In the future, we may place more AI in the game, allowing players and NPCs to grow together, creating a more realistic gaming experience.
Combination of scenes: Metaverse and car machines bring new opportunities
**Tencent Technology: You said in an interview before that China may have the largest Metaverse market. What is the current development of the Chinese market? What are the advantages or advantages? **
Zhang Junbo: China is definitely the largest metaverse market, and related companies in China have already taken the lead in digital twins.
For example, many ports and mines in China are unmanned in the scenes of vehicle traffic coordination and various port terminals. These scenes naturally require digital twins. In addition, many e-commerce companies and Chinese customers are already leading in the use of virtual anchors and 3D product displays in the e-commerce field, including software clothing, home decoration and hard furniture, and are continuing to invest and develop.
In the marketing process, many brand companies are building their own metaverse communities. Although some companies still take some detours in this regard, they are constantly trying and exploring. Ultimately, if the virtual-real Metaverse becomes a reality, China will certainly be at the forefront. Because except for some specific industries, other industries in the Chinese market are really full of competition.
Many Chinese private enterprises and entrepreneurs are very innovative, and they are constantly trying and applying new technologies. Many enterprises are very innovative, and they are very willing to try and apply the use of new technologies.
**Tencent Technology: We can see that some overseas companies, such as Meta, a company that focuses on XR and Metaverse, integrate AI into the concept of the company. What do you think of the trend of combining XR with AI? Including that Apple will release the first XR product this year, if the combination of related AI technologies is added, what kind of impact may it have on the industry? **
**Zhang Junbo: The entrance to the Metaverse does not have to be realized with a helmet, but should be multi-entrance, and there will be countless stations in the future. **Maybe the starting point for Meta’s consideration is that there are enough data and resources to realize this vision in various ways.
The idea of Meta is similar to the concept of Web2.0, hoping to expand its influence and attract more users to participate in the creation and sharing of content.
Perhaps to a certain extent, in some applications, users can get a better experience improvement, such as the positioning technology used in AR to describe the positioning of the space. Another example is technology related to extended reality - NeRF-based technology, etc., which can help to quickly generate high-quality video and other content. The core of Meta is to create a platform that provides opportunities for creation and sharing. This model may help Meta replicate Apple's successful experience, allowing users to create content and provide data on the platform, thereby achieving a closed-loop ecosystem.
In conclusion, regarding the combination of artificial intelligence and extended reality technology, it is actually not directly related. While some AI technology may be used in the headset to process the data, it's not necessarily linked.
Player entry competition: Computing power and data are the core competitiveness
**Tencent Technology: Just now you mentioned in the dialogue that various major manufacturers are releasing AI products. What is Unity's current plan in this regard? Will you release your own AI product? **
Zhang Junbo: It is still in the stage of exploration. This also involves the issue of large models and small models. We believe that in the future, we will need to have our own small models in the vertical field. The problem that needs to be paid attention to and solved is: even if it is a small model, where does our data come from. Here are a few aspects:
First of all, from a theoretical point of view, because our software carries a large number of 3D models and game logic, compilation, but compilation belongs to the developer in terms of copyright, whether it is possible to use the data involved in our software to compile, Retrieve, train? We must consider issues related to personal privacy, data security, copyright and so on.
Second, billions of games and applications running every day generate a large amount of 3D data, can it be used for training? This also requires the consent and authorization of all parties. If it cannot be used, what we can do now is to use the existing knowledge base to do some training, which can provide some help. According to some open source codes available to help developers do something.
In fact, in the final analysis, it is still a matter of data. I don't think game developers will really use OpenAI to develop games. Because there's a lot of privacy involved here. The problem we often encounter is that I have some 2D CAD models, can I convert them into 3D running models? AI can do it. But do I upload this model? After uploading, other people can have it, so this is a very painful and difficult choice.
If you see online, some people claim to use ChatGPT to make games for a few minutes, those are actually relatively basic, or in layman’s terms, they are “stupid games” and have not reached the point where they can be developed directly with AI .
For Unity China, we often use some interfaces of OpenAI to develop and try. But for game development and future operations, we still prefer domestic platforms for access and cooperation. What we are considering is: which platform to connect to in China, which is the best or the most credible and reliable, but judging from the current situation, basically all the entrants are going hand in hand and looking for it.
All in all, there are two elements here: first, there must be computing power, and second, there must be data. After the training is over, it can be explained how to use it.
**Tencent Technology: From games to solutions for smart car cockpits, what do the two engines have in common? **
Zhang Junbo: Most of the vehicle-mounted devices and platforms are developed based on Android, Linux and other platforms and Qualcomm, Xinchi and other vehicle-mounted chips, so its computing power is basically similar to that of mobile phones, which means that through Unity’s powerful With cross-platform capabilities and lightweight features, many applications can be easily ported to vehicle-mounted devices. Based on this background, more and more developers will use the cloud platform to develop 3D interactive in-vehicle applications, instead of using traditional software development, which is a costly and inefficient way. This is what current car equipment and mobile phones have in common.
But unlike the game engine that squeezes all the computing power in exchange for a more realistic performance of the game, you must reserve enough performance and resources in the car to deal with driving-related functions. These functions have priority and cannot allow other applications to take up too much memory. and resources lead to the collapse of the basic functions of the car itself. Therefore, on in-vehicle devices, developers need to consider resource management and optimization more carefully. There are also some other points to note:
First, the startup time of the in-vehicle application must be fast. Because if it takes dozens of minutes or even longer to start the game, it will greatly affect the user experience for in-vehicle applications.
In addition, the management methods of in-vehicle applications and data also need to be different from traditional computers and other equipment, because once the hard disk has no available space, the in-vehicle needs to carry out effective resource management. For example, more and more OTA systems are now available. For normal use, you must ensure that it has enough hard disk space to ensure the stable operation of the system itself.
Due to the irregular shape of car screens, unlike mobile phone screens, the display and layout of applications require special consideration. The input and output methods in the car also need to be different, and may need to use voice input and so on. And the audio system in the car also needs special treatment in order to allow users to get a better music and sound experience.
In short, in the development process of in-vehicle applications and data, many factors that are different from ordinary computers or mobile phones need to be considered in order to provide users with a better experience.
Future trends: more interactive forms besides vision and hearing appear
**Tencent Technology: What new interactive forms will be born in the future game industry? **
Zhang Junbo: There will definitely be new interactive forms. In addition to interacting mainly through vision and hearing, there will be more possibilities in other aspects. For example, some gloves can be used as input or output devices. In addition, there are many functions in the headset, such as interaction through recognition of eyes and facial expressions. These technologies are also related to AI. In addition, based on control methods such as gestures, eyeballs, and voice, there are actually many types of these input and output methods.
The more radical ones are those who are working on brain-computer interfaces, considering that they may do some related combinations, but this involves too many ethical and security issues, which will cause more problems that we cannot imagine, and it will take a long time. time to resolve and implement.
**Tencent Technology: In August last year, Unity established a Chinese independent joint venture company (Unity China). What is the logic of the separate split? What are the important changes in the company's overall strategy in recent years? **
Zhang Junbo: There are many reasons for the logic of China's split, mainly in two aspects.
First of all, overseas markets attach great importance to China, and the core market direction will determine the company's development logic**. For example, where is the company's headquarters, and product strategies and routes need to be formulated according to corresponding needs. Over the years, we have often seen that the demands of developers have not been met in China, and splitting and combining with the domestic market is a necessary condition.
**The other is platform cooperation. ** There are many platforms abroad, such as Google, Amazon and Apple, and there are basically corresponding platforms in China. If there is no independent team in China, the priority of all integrations with these platforms will be very low. If it is split, we can put some research and development priorities in China, realize some innovations in China, and feed them back to foreign countries.
** After the split, the company's strategy includes two changes, one is cloudification. **Before 2015, we have been the traditional engine software, and after 2015, we have gradually converted the product, including bringing some equipment advertisements. In addition, in terms of operations and products, the transition to maximize SAAS, and this transition is still accelerating, our goal is to gradually turn the Unity engine into a pure cloud product in the future.
View Original
The content is for reference only, not a solicitation or offer. No investment, tax, or legal advice provided. See Disclaimer for more risks disclosure.
Interview with Zhang Junbo, Executive President of Unity China: In the AIGC era, end games and TV cloud games will rise
Source: Tencent Technology
Editor: Li Haidan Hao Boyang Zhao Yangbo
Focus on:
When the wave of AIGC comes, all walks of life may face the reshaping of competitiveness, and the game industry is no exception. On the one hand, topics such as the replacement of the original artist have repeatedly rushed to the hot search, on the other hand, game innovation attempts such as enriching NPC dialogues with AIGC content are also endless. Under the influence of AI, what kind of disruption and innovation is the game industry facing? Are there any opportunities for new giants to emerge in the industry? Will the game we see afterwards be completely different? At the BEYOND International Science and Technology Innovation Expo held recently, Tencent Technology had a dialogue with Zhang Junbo, CEO of Unity China, and interpreted these issues.
The following is the content of the conversation:
AIGC's impact on the game industry: the future will no longer be "the world of giants"
**Tencent Technology: With the popularity of AIGC, many people are paying attention to whether the jobs in the game industry will be impacted and affected. What kind of impact do you think will be brought? What kind of opportunities does the outbreak of AIGC mean for the entire game industry? **
Zhang Junbo: On the whole, this wave of AIGC has a relatively large impact on the positions in the game industry combined with the design direction. For example, original painting, art and other positions, in fact, many companies have already started laying off employees in related positions,** and the layoff rate is 50%-80%**, the proportion of layoffs is quite large. If a game company with 10,000 people can earn 10 billion a year. But we assume that using AI to participate together may only need 1,000 people. AI can further "reduce costs and increase efficiency" for companies.
Next, positions such as modellers, terrain editors, and game testing can also be replaced by AI, and people who write codes may also be replaced. If the code written by Copilot is better, jobs that require manual coding Probably less and less. In addition, game planning positions may also be affected, such as script planning, many game scripts may be written in GPT. There is also a possibility that this will force the job requirements and content of these related directions to be more complicated.
In general, all practitioners in the game industry may be affected to a certain extent. However, based on the outbreak of AIGC, the scale of the game industry will further increase. From this perspective, the number of jobs in the game field may further increase, but the direction of job demand has changed. Here will be several development trends:
**First, game companies in the future will not necessarily be concentrated in a few big companies and other giant companies as they are now. Because the cost of making games is very high, small and medium-sized factories can rely on AI to invest moderately in the future, and they can also make very good games. AIGC may increase the popularity of game development in the future. **
It used to be that the research and development of some games required the investment of hundreds of millions of dollars to make a "masterpiece". Now, the cost can be reduced, and a very beautiful game can be made with tens of millions of dollars. There will be more creativity, players and planning. These can be integrated into your own art style. In terms of gameplay, with the help of AI, it can also bring about certain innovations. More exquisite games may appear, and the lower threshold means more new players will enter the game. **AIGC brings about changes in game creation and subverts the process of game creation. It may bring another wave of start-ups and start-ups to dominate the market and give everyone more choices and more opportunities. **
**Second, the future game ecology will be more diverse and rich. The addition of AIGC will enrich the content of the game and generate more innovations. **
**Third, there will be more demand for game jobs in different forms. **Maybe in the short term some game practitioners will be replaced due to the impact of this wave of AI, but it is impossible to replace all people 100%. There are still many tasks that artificial intelligence cannot accomplish at present. For example, programmers, their job requirements are not only to write code, but also include data transmission and other engineering issues that still need to be dealt with, as well as some work that needs to be done manually, which cannot be done by GPT. For another example, for a large amount of 3D content, rendering processing is also required. This process is very important, but there is no easy-to-use artificial intelligence rendering technology. Artificial intelligence can only give approximate results at present, and we still need a lot of Engineers to do things that artificial intelligence cannot.
In general, with this wave of innovation, more games may appear in the future, and there will be more opportunities to enter the game.
For the game engine, we will integrate many AI capabilities to make the development process more efficient. We will not make a separate AI product, but will integrate AI technology into our various capabilities to help developers use AI functions anytime and anywhere, and they can be replaced by AI to improve work efficiency. Integrate for the purpose.
In addition, in terms of the use and guidance of developers' functions, there may be many documents and functions, etc., which developers may not necessarily use. Technical support or manual learning is required, and the cost of occupation is very high. If it is possible that AI can help us, let developers understand functions faster. For example: we have a technology stack called DOTS. DOTS is different from traditional programming. It used to be an object-oriented programming language. It is not the optimal data structure in graphics. Data-oriented programming is actually a better structure. more efficient. The problem with data-oriented DOTS is that DOTS code is difficult to write, because it needs to change the programmer's thinking. If with the help of AI, programmers can use natural language or object-oriented language to describe, AI can help programmers generate inverted codes. If you want me to do the first (AI fusion) thing, this is the first thing I want to do.
Game form brings innovation: end games and TV cloud games will rise
**Tencent Technology: Games are divided into mobile games and console games. Do you think AIGC has the same level of impact on these two application scenarios? For example, among all the impacts, what is the impact on mobile games and console games? **
Zhang Junbo: From the perspective of AI, there is not much direct impact, but under this trend, end games (PC games/console games) may rise further.
This is related to its production cost. From the point of view of mobile games, the screen of the mobile phone is small, even if the pixels are a little lower, the effect is not good enough to play. However, if you want to experience console or PC games, etc., to experience games with good effects, the requirements for pixels and image quality are very high. Correspondingly, the production cost is very high. For example, the experience of users playing 720P games on TV is very bad.
However, if more players join in through the combination of the latest AI technology, the overall R&D cost will be reduced, and a lot of technology in the game industry will be mentioned. For example, the future game standard has become 4k or 8k games, and the game experience in the TV scene will definitely explode further. In general, if there are more and more high-definition and high-quality content in the future, end games will rise, and cloud games on TV will become the mainstream. **
**Tencent Technology: In terms of product form, in March this year, Unity CEO John Ricitiello also mentioned the possibility of using generative AI for NPC dialogue. In the future, in what form will generative AI and game development be better integrated? **
**Zhang Junbo:**There is a very natural connection and application scenarios between artificial intelligence and game development. For example, the NPC in the game, how to describe and simulate real players is one of the most suitable fields for AI.
In addition, AI can also be used to drive characters in games, and drive virtual characters, such as your Avatar or digital avatar, through the facial expressions and language of real people. These are all relatively feasible, and we could see them implemented soon. **The application scenarios of future NPCs can be wider and more subdivided. **
At present, it is easier to deal with NPCs in fixed fields, and its actions and routes are fixed, but it is still difficult to deal with in complex environments. Especially NPCs with special status. For example, in the open world of the World of Warcraft game, AI may support NPCs to form the same growth path as players, so that the equipment of NPCs and the equipment played by players can simulate the real situation, rather than simply setting it. This requires more development experience and technical support. In the future, we may place more AI in the game, allowing players and NPCs to grow together, creating a more realistic gaming experience.
Combination of scenes: Metaverse and car machines bring new opportunities
**Tencent Technology: You said in an interview before that China may have the largest Metaverse market. What is the current development of the Chinese market? What are the advantages or advantages? **
Zhang Junbo: China is definitely the largest metaverse market, and related companies in China have already taken the lead in digital twins.
For example, many ports and mines in China are unmanned in the scenes of vehicle traffic coordination and various port terminals. These scenes naturally require digital twins. In addition, many e-commerce companies and Chinese customers are already leading in the use of virtual anchors and 3D product displays in the e-commerce field, including software clothing, home decoration and hard furniture, and are continuing to invest and develop.
In the marketing process, many brand companies are building their own metaverse communities. Although some companies still take some detours in this regard, they are constantly trying and exploring. Ultimately, if the virtual-real Metaverse becomes a reality, China will certainly be at the forefront. Because except for some specific industries, other industries in the Chinese market are really full of competition.
Many Chinese private enterprises and entrepreneurs are very innovative, and they are constantly trying and applying new technologies. Many enterprises are very innovative, and they are very willing to try and apply the use of new technologies.
**Tencent Technology: We can see that some overseas companies, such as Meta, a company that focuses on XR and Metaverse, integrate AI into the concept of the company. What do you think of the trend of combining XR with AI? Including that Apple will release the first XR product this year, if the combination of related AI technologies is added, what kind of impact may it have on the industry? **
**Zhang Junbo: The entrance to the Metaverse does not have to be realized with a helmet, but should be multi-entrance, and there will be countless stations in the future. **Maybe the starting point for Meta’s consideration is that there are enough data and resources to realize this vision in various ways.
The idea of Meta is similar to the concept of Web2.0, hoping to expand its influence and attract more users to participate in the creation and sharing of content.
Perhaps to a certain extent, in some applications, users can get a better experience improvement, such as the positioning technology used in AR to describe the positioning of the space. Another example is technology related to extended reality - NeRF-based technology, etc., which can help to quickly generate high-quality video and other content. The core of Meta is to create a platform that provides opportunities for creation and sharing. This model may help Meta replicate Apple's successful experience, allowing users to create content and provide data on the platform, thereby achieving a closed-loop ecosystem.
In conclusion, regarding the combination of artificial intelligence and extended reality technology, it is actually not directly related. While some AI technology may be used in the headset to process the data, it's not necessarily linked.
Player entry competition: Computing power and data are the core competitiveness
**Tencent Technology: Just now you mentioned in the dialogue that various major manufacturers are releasing AI products. What is Unity's current plan in this regard? Will you release your own AI product? **
Zhang Junbo: It is still in the stage of exploration. This also involves the issue of large models and small models. We believe that in the future, we will need to have our own small models in the vertical field. The problem that needs to be paid attention to and solved is: even if it is a small model, where does our data come from. Here are a few aspects:
First of all, from a theoretical point of view, because our software carries a large number of 3D models and game logic, compilation, but compilation belongs to the developer in terms of copyright, whether it is possible to use the data involved in our software to compile, Retrieve, train? We must consider issues related to personal privacy, data security, copyright and so on.
Second, billions of games and applications running every day generate a large amount of 3D data, can it be used for training? This also requires the consent and authorization of all parties. If it cannot be used, what we can do now is to use the existing knowledge base to do some training, which can provide some help. According to some open source codes available to help developers do something.
In fact, in the final analysis, it is still a matter of data. I don't think game developers will really use OpenAI to develop games. Because there's a lot of privacy involved here. The problem we often encounter is that I have some 2D CAD models, can I convert them into 3D running models? AI can do it. But do I upload this model? After uploading, other people can have it, so this is a very painful and difficult choice.
If you see online, some people claim to use ChatGPT to make games for a few minutes, those are actually relatively basic, or in layman’s terms, they are “stupid games” and have not reached the point where they can be developed directly with AI .
For Unity China, we often use some interfaces of OpenAI to develop and try. But for game development and future operations, we still prefer domestic platforms for access and cooperation. What we are considering is: which platform to connect to in China, which is the best or the most credible and reliable, but judging from the current situation, basically all the entrants are going hand in hand and looking for it.
All in all, there are two elements here: first, there must be computing power, and second, there must be data. After the training is over, it can be explained how to use it.
**Tencent Technology: From games to solutions for smart car cockpits, what do the two engines have in common? **
Zhang Junbo: Most of the vehicle-mounted devices and platforms are developed based on Android, Linux and other platforms and Qualcomm, Xinchi and other vehicle-mounted chips, so its computing power is basically similar to that of mobile phones, which means that through Unity’s powerful With cross-platform capabilities and lightweight features, many applications can be easily ported to vehicle-mounted devices. Based on this background, more and more developers will use the cloud platform to develop 3D interactive in-vehicle applications, instead of using traditional software development, which is a costly and inefficient way. This is what current car equipment and mobile phones have in common.
But unlike the game engine that squeezes all the computing power in exchange for a more realistic performance of the game, you must reserve enough performance and resources in the car to deal with driving-related functions. These functions have priority and cannot allow other applications to take up too much memory. and resources lead to the collapse of the basic functions of the car itself. Therefore, on in-vehicle devices, developers need to consider resource management and optimization more carefully. There are also some other points to note:
First, the startup time of the in-vehicle application must be fast. Because if it takes dozens of minutes or even longer to start the game, it will greatly affect the user experience for in-vehicle applications.
In addition, the management methods of in-vehicle applications and data also need to be different from traditional computers and other equipment, because once the hard disk has no available space, the in-vehicle needs to carry out effective resource management. For example, more and more OTA systems are now available. For normal use, you must ensure that it has enough hard disk space to ensure the stable operation of the system itself.
Due to the irregular shape of car screens, unlike mobile phone screens, the display and layout of applications require special consideration. The input and output methods in the car also need to be different, and may need to use voice input and so on. And the audio system in the car also needs special treatment in order to allow users to get a better music and sound experience.
In short, in the development process of in-vehicle applications and data, many factors that are different from ordinary computers or mobile phones need to be considered in order to provide users with a better experience.
Future trends: more interactive forms besides vision and hearing appear
**Tencent Technology: What new interactive forms will be born in the future game industry? **
Zhang Junbo: There will definitely be new interactive forms. In addition to interacting mainly through vision and hearing, there will be more possibilities in other aspects. For example, some gloves can be used as input or output devices. In addition, there are many functions in the headset, such as interaction through recognition of eyes and facial expressions. These technologies are also related to AI. In addition, based on control methods such as gestures, eyeballs, and voice, there are actually many types of these input and output methods.
The more radical ones are those who are working on brain-computer interfaces, considering that they may do some related combinations, but this involves too many ethical and security issues, which will cause more problems that we cannot imagine, and it will take a long time. time to resolve and implement.
**Tencent Technology: In August last year, Unity established a Chinese independent joint venture company (Unity China). What is the logic of the separate split? What are the important changes in the company's overall strategy in recent years? **
Zhang Junbo: There are many reasons for the logic of China's split, mainly in two aspects.
First of all, overseas markets attach great importance to China, and the core market direction will determine the company's development logic**. For example, where is the company's headquarters, and product strategies and routes need to be formulated according to corresponding needs. Over the years, we have often seen that the demands of developers have not been met in China, and splitting and combining with the domestic market is a necessary condition.
**The other is platform cooperation. ** There are many platforms abroad, such as Google, Amazon and Apple, and there are basically corresponding platforms in China. If there is no independent team in China, the priority of all integrations with these platforms will be very low. If it is split, we can put some research and development priorities in China, realize some innovations in China, and feed them back to foreign countries.
** After the split, the company's strategy includes two changes, one is cloudification. **Before 2015, we have been the traditional engine software, and after 2015, we have gradually converted the product, including bringing some equipment advertisements. In addition, in terms of operations and products, the transition to maximize SAAS, and this transition is still accelerating, our goal is to gradually turn the Unity engine into a pure cloud product in the future.