News center > News > Headlines > Context
Musk, the world’s richest man, has entered the field of AI gaming, and the war between China and the United States in AI intelligence has begun
Editor
2024-11-29 15:03:01 9,006

Musk, the world’s richest man, has entered the field of AI gaming, and the war between China and the United States in AI intelligence has begun

Image source: Generated by Unbounded AI

The "World's Richest Man" is forming a new team, this time around AI.

Titanium Media App reported on November 28 that the world’s richest man and Tesla CEO Elon Musk retweeted today that too many game studios have been acquired by major companies. In order to "make games great again", he proposed that his AI company xAI will establish an AI game studio.

This is the first time that Musk has proposed that his company enter the AI ​​gaming industry, and it is also the first time to announce the application field of xAI’s intelligent agents.

Subsequently, Musk responded to another tweet and criticized some game developers, "Can they make games well and not add those awakening sermons?"

In fact, This isn't the first time Musk has shown interest in making games. In October, he said, "If there's one thing that's fun to do, it's to set up a game studio. Because I really do fundamentally like games. I feel that if you can apply AI to games, you really It can make great games.”

Earlier, a video of Musk’s exchange with Trump’s granddaughter circulated on the Internet. Trump's granddaughter asked, "Do you play Fortnite?" Musk replied, "No, no, no." Afterwards, Trump’s granddaughter asked, “Do you play League of Legends?” Musk responded humorously, “League of Legends? If I play this, it will affect my space program.” p>

This may be one of the "triggers" for Musk's plan to make AI games. But as of press time, Musk has not further explained how he hopes to build AI+ gaming products or technology.

As the release of ChatGPT enters its second anniversary, the implementation of large AI models is moving from the "model layer" to the intelligent agent (AI Agent). Microsoft, Meta, Baidu, Tencent, etc. are all developing AI intelligence. body application.

In response to Musk’s establishment of an AI game studio, Chen Di, founder and CEO of the domestic AI game company New Wisdom Heart, told TMTpost App that Musk’s move will undoubtedly have a profound impact on the game industry. The in-depth combination of the two creates a new way of experience, marking that the game industry will enter a new stage of development. AI will develop from the construction of the game world, to companionship during the game, to having equal rights with players. This will greatly enhance the social and interactive nature of the game, allowing players to gain a more realistic and profound emotional experience in the game.

AI and games have sparked sparks, but whether AI can completely replace game creation is still controversial

With the rapid development of generative AI, its Technology has gradually been applied to all aspects of life, such as AlphaGo's victory over Lee Sedol, image beautification, one-click face-changing in live broadcasts, etc. At the same time, AI also has many applications in games and plays a very important role in the game field.role.

In early games such as "Pac-Man", AI technology played a key role, and antagonistic AI characters appeared at that time. Later, this kind of AI character gradually gave rise to more complex NPC AI. Since DeepMind published the DQN article, game AI has begun to attract attention. In just three or four years, DQN has developed many variants and solved many functional problems that could not be solved in the past.

So, specifically, how is AI used in the game development process? TMTpost App summarizes six main directions.

For example, 1. Assisted design uses tools such as StableDiffusion to quickly generate multiple complex game scenes and character images based on text descriptions; 2. Real-scene scanning, conducts real-scene scanning of domestic scenic spots and historical sites, and uses AI technology to generate games Screen; 3. Motion capture, AI combined with motion capture technology, "Black Myth: Enlightenment" The action production efficiency of "Empty" can be increased by about 30%; 4. Procedural generation, using AI technology to automatically create game scenes and elements; 5. NPC characters, introducing decision trees and other related technologies, so that NPCs in the game can show complex behaviors Mode; 6. Anti-cheating and assisting users, using AI technology to monitor player behavior in real time to quickly identify and prevent cheating.

In addition, in terms of generative AI technical capabilities, AI is mainly empowered in two scenarios: game super scores and NPC personalized characters.

For example, a player recently used AI-related technology to clear the 150-level secret realm of "Diablo 4" in 1 minute and 52 seconds, which attracted attention.

Weldon Zhang, technical director of NetEase Game Artificial Intelligence Laboratory, once pointed out that there is no doubt that AI brings more convenience and more diverse gameplay to game development. At the same time, the use of artificial intelligence in a game The quality will also directly affect the playability of the game. How to add more and more novel AI technologies to the existing technical level is also a topic worth discussing in game development.

Currently, AI is gradually penetrating into the gaming field. According to Gamma data, more than 99% of game industry respondents’ companies or departments have introduced AI technology and tools. At the R&D level, about 80% of respondents perceive that AI improves project efficiency by more than 20%. Nearly 10% of respondents believe that AI can improve efficiency by more than 50%. Currently, AI is most widely used in game auxiliary content production, such as "improving the freedom of content such as NPCs and plots" and "assisting the creation of art, music, code, etc." accounting for nearly 70%.

However, there is still controversy in the industry as to whether AI can replace game creation.

Unity China President and CEO Zhang Junbo once admitted to TMTpost App that AI can improve production efficiency, which is a good thing for the game industry. Programmers’ “jobs” may be robbed, but game creators Their “jobs” will not be robbed, and it will be difficult to speed up game development.development process. At the same time, AI can improve production efficiency to a certain extent, reduce the proportion and cost of creatives, and reduce costs and increase efficiency in the game development process.

"Can we produce more and better games with the power of AI, and then everyone will have more choices, allowing everyone to release their creativity. Therefore, I think AI has the potential to make everyone's Creativity is realized at a lower cost, and players have more choices with AI. It’s more beneficial for creatives, because in the past, you needed a lot of people to implement ideas. Now with the blessing of AI, creative people have more room to use their skills. Overall, this is a process that continues to improve the cost-effectiveness. ), if AI improves production efficiency, it will increase the output per unit area, and everyone will be able to play more and better games." Zhang Junbo views AI technology's changes in games very rationally. In his view, the use of AI tools can greatly reduce game development costs. In the past, the research and development of some games required 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.

However, many people who are optimistic about the future development of AI believe that the integration of AI and games will become increasingly close, and may completely change the game creation model.

Chen Di told TMTpost App that although AI still faces challenges in terms of illusion and creativity, the future of AI in the gaming industry is by no means limited to its role as an auxiliary tool.

“The hallucination problem is the biggest problem that AI scientists are currently seeking to solve. Its root lies in our incomplete understanding of the working principle of neural networks. In terms of creativity, although the current performance of AI is not as good as humans, But this is mainly because the training data comes from the real world, and many artificial restrictions are imposed on the model output, just like the Diablo system. From the 2nd generation to the 4th generation of series of games, the innovation in gameplay is not significant, only the graphics have been improved. This reflects that the pace of product innovation of game companies is relatively slow, compared with the technology industry or the film industry. Lagging behind. But we have reason to believe that with the continuous advancement of technology and the in-depth training of vertical field models, AI will bring huge changes to the game industry," Chen Di said.

Chen Di co-founded Beijing New Wisdom Heart Technology in 2024, which mainly develops game auxiliary plug-in products based on AI technology. The game skills it developed, GameSkills-Fearless Contract AI sparring, will help Fearless Contract players in the future. In the customized single-player mode, in-game prop throwing and use training is conducted on all maps, all heroes, and all skills, helping players quickly master advanced skills to improve their technical and tactical levels, and greatly improve training efficiency.

Chen Di revealed that the game skills use the XPU architecture computing power of Intel Core Ultra series chips to achieve real-time capture and feature recognition of dynamic images in the game, and to analyze the hero characters, map positions and perspective images in the game screen. Recognition processing is performed one by one, with auxiliary aiming prompts and tutorial videos, in the case of low power consumption.Under the circumstances, the in-game Lineup training function is implemented, so that it does not occupy resources with the game itself and does not reduce the game performance. In the near future, the product it developed will be open to professional teams in the Fearless Contract China Division for testing and training, and will be available to ordinary players in the future. In 2025, game skills will launch more advanced functions for Fearless Contract, and more AI assistant functions for different games will also be launched.

As a terminal manufacturer of AI games, Zong Ye, general manager of Intel China’s channel and partner business department, also believes that the outbreak of AI will definitely lead to a new round of PC replacement. The arrival of the AI ​​PC era, It will also accelerate the integration of AI into all walks of life, and the gaming industry has always been at the forefront of new technologies due to its industry characteristics. So far, Intel has shipped more than 20 million AI PC devices and built an ecosystem of more than 300 AI applications and more than 500 AI models.

In February this year, Musk said that Tesla should have produced a game long ago. "Our real-world simulation and video generation are the best in the world, but Tesla can't start making games until we release unregulated FSD, which is safer than regulated FSD."

Today, Musk is accelerating this desire and rushing to the AI ​​game track.

Microsoft, Meta, and Baidu have all come to an end, and China and the United States are fiercely fighting for the AI ​​​​intelligent industry

AI games are only for AI Agent (intelligent) applications One of the scenes.

In fact, the current new round of AI competition is moving from the "battle of hundreds of models" to the application of AI agents, and a new round of AI agent competition in China and the United States seems to have already begun.

On November 20, Microsoft announced at Microsoft Ignite, its annual conference in Chicago, the United States, the launch of 10 newly upgraded AI intelligent products including Microsoft Copilot Studio and Copilot Actions. Among them, Microsoft Copilot Studio can call any of more than 1,800 customized Azure AI models, which has attracted attention.

Microsoft CEO Satya Nadella said that agents will be new applications in the AI ​​world, and they are rapidly adding new features to solve the biggest pain points in personal work and drive real business results. , "A major change is coming."

Microsoft Vice President Bryan Goode bluntly said that these agents can be regarded as real applications in the AI ​​era. "Every business system that exists today will be reimagined as an Agent on top of an AI intelligent assistant."

It’s not just Microsoft. American technology giants such as Google, OpenAI, and Meta are also stepping up efforts to build their own AI intelligent products.. According to Axios reports, Meta is forming a new product team to create AI agent applications and tools for enterprise customers to achieve more revenue growth.

Meta CEO Zuckerberg said that there may be more AI agents than humans in the future, and humans will live in a world with hundreds of millions or even billions of AI agents, becoming the core of every company. "Standard", just like today's email, website and social media.

Compared to the American giants on the other side of the ocean, domestic technology leaders such as Tencent, Baidu, ByteDance, and Zhipu are all deploying AI agent applications.

Baidu founder and chairman Robin Li this month released two AI technologies that enable applications: retrieval-enhanced venography technology (iRAG) and the code-free tool "Miaida". He said that Wenxin iRAG is used to solve the illusion problem of large models in image generation, greatly improving practicality; the code-free tool "Miaida" allows everyone to have the ability of a programmer and will create millions of "super useful "application. As of now, the average daily calls of Baidu Wenxin's large model has exceeded 1.5 billion, an increase of about 30 times compared with a year ago; and the number of Wenxin Yiyan users has reached 430 million. At the same time, Baidu has more than 100,000 Wenxin agents equipped with commercial components, and the number of agents with monetization capabilities has increased by as much as 395%, with the highest revenue from a single monetization reaching 102,000 yuan.

Robin Li emphasized that agents are the most mainstream form of AI applications and are about to reach a breaking point and will become the new carrier of content, information and services in the AI ​​native era.

Alibaba and Tencent are also developing AI agent applications. Among them, Alibaba has made a large number of updates around AI search during the year. According to PConline statistics, its PC version has ranked first in the number of downloads of domestic computer-side AI applications this year; Tencent finally launched the AI ​​intelligent workbench ima.copilot, which includes AI search capabilities, at the end of the year. Its own Hunyuan large model provides technical support. In September this year, Tencent revealed that there are more than 100,000 agents on the Tencent Yuanbao app providing services to users every day.

Earlier, ByteDance’s self-developed large model Doubao model was released. According to the latest data, Doubao App has been downloaded more than 100 million times, has more than 30 million monthly active users, and more than 8 million agents have been created. Enterprises such as China Merchants Bank and Haidilao Hot Pot have already launched on Byte's AI application development platform Butou. Intelligent bodies have been built, and famous universities such as Fudan University and Zhejiang University have also built AI "teaching assistants" for courses and experiments. In addition, ByteDance Doubao also released its first AI smart headset Ola Friend this year.

Obviously, AI agents have become a must-have for exploring the implementation of AGI (artificial general intelligence).

The so-called AI agent, Wu Hequan, an academician of the Chinese Academy of Engineering, said that AI agents have memory, planning, tools, and action capabilities. With AI agents, they can complete some tasks on behalf of people, and at the same time, they can allow the terminal to Equipment, industry, spatial computing, intelligent drivingIt has general AI capabilities in areas such as driving, credit management, and “embodied intelligence.”

Fu Sheng, chairman and CEO of Cheetah Mobile and chairman of Orion Star, told TMTpost Media App that 2025 is expected to be a prosperous year for AI applications.

“OpenAI has strengthened its product capabilities and application capabilities this year, and the same is true for 4o. The new conversation mode is particularly smooth, and it can also be regarded as an application, which means that the model itself is constantly changing. Scaling Law has actually slowed down, so I think next year should be a year of great prosperity for AI applications,” Fu Sheng said.

Chen Di talked about the future applications of AI games to TMTpost App. He bluntly said that in the field of e-sports, AI that plays games with users in the future will become one of the closest friends in users' lives; professional players will have AI Agents with the deepest understanding of e-sports techniques and tactics to help them improve their skills.

“The future of AI in the game industry is by no means just an auxiliary tool, but a brand-new productivity tool and interactive tool that can bring more possibilities.” Chen Di said that it is foreseeable that , AI will profoundly change the face of the game industry in many aspects such as game design, content generation, and player interaction, leading an unprecedented game revolution.

Chen Di added, “This change will also profoundly affect the way players interact. In the future, every player may have an AI friend who will accompany him for a long time. Players understand him and he is also familiar with players. , you can fight side by side in the game, accompany players to improve scores, and even ride in different game worlds together, just like we may not see old classmates for many years in real life, but we often see them in the game. "We often team up together."

According to the latest data released by the US consulting firm MarketsandMarkets, the global AI intelligent agent market will achieve substantial growth in the future, increasing from US$5.1 billion in 2024 to US$4.71 billion in 2030. billion, with a compound annual growth rate (CAGR) of 44.8%, far exceeding the growth of global AI output value.

Nadella believes, “The problem with AI is that computers don’t care at all, but we do. Technology is not technology for technology’s sake, but to transform technology into real results, and the transformative power of technology can drive Business growth, improved efficiency and operational balance (leverage).”

Nadella emphasized that AI technology will change people’s lives, and in fact, it has already changed it.

Keywords: Bitcoin
Share to: