AI+ Included in Government Work Report for the First Time: What Can AI Companies Do?

[baoshi.rao]

From ChatGPT to Sora, the emergent intelligence capabilities demonstrated by large models have become a societal focus. Over the past year, large models have been increasingly applied in vertical scenarios, gradually showcasing their potential to transform industries and competitive landscapes.

Against this backdrop, the 2024 Chinese Government Work Report emphasized the need to vigorously advance the construction of a modern industrial system, accelerate the development of new quality productive forces, and deepen the research and application of big data, artificial intelligence, and other technologies. It also highlighted the launch of an "AI+" initiative to build internationally competitive digital industrial clusters. Zhang Yunquan, a member of the National Committee of the Chinese People's Political Consultative Conference and researcher at the Institute of Computing Technology, Chinese Academy of Sciences, pointed out in an interview with China News Service: Compared to the previous wave of AI enthusiasm driven by deep learning, generative AI, as a new production tool, possesses more comprehensive skills and broader application scope. It can replace certain mental labor tasks, be implemented across various industries—especially those facing labor shortages—and ultimately translate into tangible productivity gains.

Creating new economic growth points requires developing new quality productive forces, and the continuous advancements in AI technology will serve as a crucial engine for forming these new productive forces. In Zhang Yunquan's view, "AI+" primarily relies on large models, big data, and massive computing power to achieve the practical application of AI across industries. This will enhance industrial automation levels, reduce costs, improve efficiency, and propel the digital economy to new heights.

Accelerating the industrial implementation of large models and more efficiently unleashing their technological value remains one of the core topics in the industry. According to Deloitte AI Institute's quarterly research report "The State of Generative AI in the Enterprise," three-quarters (79%) of respondents expect generative AI to drive substantial corporate transformation within the next three years. Zhou Hongyi, a member of the National Committee of the Chinese People's Political Consultative Conference (CPPCC) and founder and chairman of 360 Group, highlighted the significance of vertical and industrial applications for large AI models in his proposal. He stressed that only by developing vertical and enterprise-specific large models can we better meet the personalized needs of enterprises, improve production efficiency, and enhance service quality.

He Han, a CPPCC member and deputy general manager of Tianyu Digital Technology, proposed encouraging industries such as manufacturing, finance, commerce, transportation, pharmaceuticals, government affairs, education, culture and tourism, and media to deeply explore and actively create various scenarios for applying large AI models. This would achieve deeper integration of AI technology with various industries.

Liu Qingfeng, a deputy to the National People's Congress and chairman of iFlytek, suggested accelerating the empowerment of large models in the industrial sector to improve quality and efficiency. He also emphasized ensuring the use of domestically developed hardware and large models in key sectors such as finance, energy, and telecommunications. While ensuring the autonomy and controllability of core infrastructure, efforts should be made to accelerate the development of an application ecosystem for domestically developed large models. In 2024, countries are engaging in comprehensive competition across foundational large models, industry applications, hardware, and supply chains, with China and the U.S. particularly contending in the depth of model applications and strategic demands. At this juncture, the government work report's emphasis on the 'AI+' action plan has provided a significant boost to the artificial intelligence industry.

Zhou Hongyi believes that elevating 'AI+' to a national action signifies that the country will strengthen top-level design and accelerate the formation of new quality productive forces driven by artificial intelligence. As an innovator and entrepreneur in the AI industry, he feels an even greater responsibility to continue deepening efforts in areas such as 'AI + security.'

Currently, in the competition around general artificial intelligence represented by large models, there remains room for improvement in the products offered by various manufacturers. Liu Qingfeng mentioned in a media interview that iFlytek's Spark model is expected to reach the current best level of GPT-4/4V within six months. However, with the release of GPT-5, this gap could extend to over a year, with a dynamic difference of approximately six months to a year. If resources in computing power, data, and model training are well-organized for a full-scale catch-up, Liu estimates this gap could be closed within 1 to 2 years.

To accelerate technological advancement, Liu Qingfeng proposed promoting the opening and sharing of high-quality national-level training data, adopting new mechanisms to cultivate top-tier AI talent, and formulating a national "General Artificial Intelligence Plan" to address global AI "systematic competition." In response to lagging regulatory frameworks and legal mechanisms, he emphasized accelerating the formulation and review of laws and regulations related to general AI technology. Additionally, he suggested establishing soft research projects for ethical and humanities studies in general AI.