China and the U. S. have achieved parity in synthetic intelligence progress, but China’s implementation of this generation in products is expected to advance in 2023.
As Kaifu Lee, a keen observer of China’s AI progress, put it, “We are now in the era of AI implementation. “Even though the West, the United States and Canada in particular, will continue to be at the forefront of AI research. , those Western advances will be temporarily followed in China, where the large market, abundance of young engineers, government support, and a relentless entrepreneurial culture are driving business innovation in AI.
“The virtual and genuine economies are accelerating their integration,” said Haifeng Wang, head of leadership generation at Baidu and director of Baidu Research.
Both China’s Fourteenth Five-Year Plan and Vision 2030 place emphasis on the progression of the virtual economy, seeing the sector as a source of super-untapped innovation strength and a growth area.
Today, smart computing centers, forms of deep platforms, and AI models are already being evolved to shape a new AI infrastructure in China, offering a foundation for a virtual transformation in the genuine economy, especially in production and autonomous vehicles.
In the coming years, the structure of this new AI infrastructure is expected to become one of the vital pillars of the progress of the virtual economy and an investment hub for local governments in China, thus particularly contributing to regional economic progress. expands, we can expect to see China expand new products and business models with commercial applicability.
However, China lacks the school infrastructure that has fueled technological progress in the United States. Even as China has created tough AI institutes, such as the Beijing Academy of Artificial Intelligence (BAAI), the country’s higher education formula, decimated in the 1960s, is decades short of matching that of the West. Teaching and studying in China compete with private sector wages, and universities are also unable to attract the most productive foreign educational talent.
However, thanks to the scaling up of giant deep learning models, multilanguage, multitasking, and multimodality technologies are adapting to a major trend in the existing progression of AI in China. According to Chinese AI giant Baidu, “large commercial companies” models are being deployed in the aerospace, financial, and energy sectors. With an AI infrastructure built around the desires of the express industry, China is creating an “AI industry” structure. Baidu expects a broader ecosystem of industrial models to emerge, leading to the intelligent upgrade of thousands of AI-based industries.
WEZHEN, CHINA – OCTOBER 21: Baidu Chief Technology Officer Wang Haifeng speaks at theArray. [ ] World Internet Conference (WIC) at Wuzhen International Internet Expo
“As large pre-trained AI models are the new technical cornerstone, the intelligence point has increased significantly, leaving more room for large-scale programs, from industry to clinical research,” Baidu’s Wang said.
Baidu, one of China’s AI giants along with Alibaba and Tencent, runs ERNIE-ViLG 2. 0, a 24-billion-parameter style trained on millions of pairs of Chinese images and text.
Another large-scale Chinese style is Taiyi, from the IDEA think tank, led by Harry Shum, co-founder of Microsoft Research Asia, Microsoft’s largest research arm outside the United States. Taiyi is made up of 20 million leaked Chinese images and texts. pairs and has a billion parameters.
Meanwhile, BAAI unveiled the most recent edition of Wudao in 2022, a pre-trained deep style with 1. 75 trillion parameters.
The maturity of those key technologies, coupled with deep knowledge and cloud computing, has given impetus to the advancement of automation, with AI-based robots of all kinds looking forward-looking for major innovations in real-time insight, intelligent decision-making, and process optimization. control. . The increasing maturity of the generation of AI-based robots means they can be used in construction, mining, crisis relief, and other labor-intensive painting scenarios.
As autonomous driving becomes less and less unusual in urban environments, the difficulty of perceiving complex environments and the difficulty of processing gigantic amounts of knowledge are seeing fundamental improvements. Traditional small models no longer meet the demands of high-level autonomous driving, making the giants in AI models the newest concentrate of self-driving technology.
This AI-style behemoth enables autonomous vehicles to successfully gain greater semantic popularity and exponential power to solve long-tail problems. These immediate technological advancements mean that by 2023, the commercialization of autonomous driving in China’s major cities will witness simultaneous expansion. in the scope of operations and fleet size, raising the market penetration rate of smart cars with autonomous driving features to new levels. The intelligent vehicle industry will move from the experimental level to full maturity.
According to Baidu, the increasing commodification of knowledge highlights the pressing need for a reliable “horizontal and vertical” knowledge flow network. The progression of such networks will continue in 2023.
Tech corporations and scientists are also actively exploring “explainable” AI technologies, with the goal of achieving more effective human-machine communication in the context of price alignment, enabling AI to actually perceive human intentions and achieving more predictable AI governance. A more virtual society, with manageable and controllable AI generation functions, will become a new competitive merit for companies.
The new infrastructures and emerging virtual teams have enabled the creation of a content-rich virtual global, which has gone from being a parallel global to one incorporated into the physical global.
Virtual simulation and artificial intelligence technologies will enable the creation of more accurate and intelligent virtual twins in shopping, manufacturing, entertainment, and more.
AI-generated content will bring a new style of content creation for drawing, painting, and other media. Combined with the immersive pleasure offered by VR/AR and the high-speed knowledge transmission capability provided through 5G, the availability of a wide and rich metaverse industry is expected to accelerate, hosting styles of products and services that offer a new wave of disruptive innovation.
In 2022, China issued “Opinions on Strengthening the Ethical Governance of Science and Technology” and submitted a “Position Paper on Strengthening the Ethical Governance of Artificial Intelligence” to the United Nations, calling for the precept of “human-centered intelligence for good. “. so that synthetic intelligence remains secure, reliable, and controllable.