Beijing Pours Money Into Building Global Innovation Hub
AsianFin--At the Future AI Vanguard Forum of the 2024 Zhongguancun Forum Annual Meeting on Saturday, Lin Jianhua, Deputy Director of the Beijing Development and Reform Commission, announced a plan to build Beijing into an artificial intelligence (AI) hub.
The plan focuses on enhancing intelligent computing power supply, strengthening industrial basic research, supporting high-quality factor aggregation, accelerating innovation and application of large models, and creating a first-class development environment.
Lin mentioned that Beijing will invest over 100 billion yuan (US$ 14 billion) in the next five years to optimize the investment and financing environment. Additionally, for AI technology research projects included in national major strategic tasks, the highest reward can reach as high as 100 million yuan.
In the field of models, Beijing has been working to strengthen the key core technology research of large models and support innovative entities in conducting key core technology research through methods such as "open ranking" and "competition," with the highest support of 30 million yuan, and up to 100 million yuan for projects included in national major strategic tasks.
For innovative breakthroughs in disruptive AI technologies, Beijing will dedicate special innovation funds to exploring in fields such as new architecture chips, brain intelligence, and brain-like intelligence. These measures will be preferentially included in the municipal-level scientific research plan with the highest support of 30 million yuan.
In terms of application, Beijing will make efforts to promote the verification and iteration of AI large model applications in application scenarios, focusing on key areas such as embodied robots, intelligent healthcare, and intelligent education.
This will involve the construction of industry large model application scenario joint research and development platforms to create a batch of exemplary AI benchmark applications, with maximum support of 50 million yuan.
In the environmental aspect, efforts will be made to optimize the investment and financing environment and increase direct investment in high-growth companies with key core technologies.
Yue Li, member of the Standing Committee of the Haidian District Party Committee and Executive Deputy District Chief, said at the event that the southern part of Haidian has a 53-square-kilometer area with top AI resources and close international interaction, and it is fully committed to building an AI innovation district.
Haidian District has effectively connected 52 national key laboratories, 10 new research institutions, 106 national research institutions, 37 top universities (including Tsinghua University, Peking University, Renmin University), 89 of the top 2000 AI global scholars (accounting for over 70% of the city), 12,300 AI scholars (accounting for over 80% of the city), 1,300 AI companies (accounting for about 70% of the city), over 90 large model-related units, and 37 registered large model units (accounting for over 70% of the city).
Haidian District is sparing no efforts to build a Beijing AI public computing platform, with 3,500P computing power already deployed, and plans to deploy 10,000P computing power by the end of the year; in terms of data, Haidian District is promoting the construction of the Beijing AI Model Corpus Center, with 1,100TB of high-quality data currently provided to large model enterprises; in terms of talents, Haidian District has established an "AI Haidian" organization and is building an "AI Haidian" talent powerhouse.
Zhang Hong, Vice President of the Zhongguancun Industry and Commerce Alliance, mentioned that the Zhongguancun has witnessed rapid development of AI industry in recent years, with an average annual growth rate of 31.4%, accounting for 26% of the global total.
He introduced that in addition to the traditional areas of industrial and agricultural production, commerce and trade, and scientific research and education, AI has rapidly expanded its application scope in the fields of medical health, environmental protection, transportation, and public security.
In 2023, the scale of AI industry reached 2.6 trillion yuan in Zhongguancun, ranking first in the world and accounting for 26% of the global total.
Beijing, a global hub for technological innovation, continues to spearhead advancements in artificial intelligence with groundbreaking developments and widespread applications. Recent endeavors by prominent institutions and industry leaders highlight the city's commitment to shaping the future of AI.
A team led by Dai Qionghai, a professor at Tsinghua University, has unveiled "Taiji," a large-scale intelligent optoelectronic computing chip. This cutting-edge innovation introduces an interference-diffraction fusion computing architecture, enabling the creation of intelligent optoelectronic computing models. Taiji is capable of depicting the scalability limits of deep optical networks, supporting complex intelligent tasks with an impressive computational efficiency of 160TOPS/W.
Its capabilities span from lightning-fast intelligent analysis of billion-pixel scenes to training and inference on models with billions of parameters. The team's research paper, published in the prestigious academic journal Science, underscores the significance of their achievement.
During a speech, Baidu's Chief Technology Officer Wang Haifeng emphasized the pivotal role of intelligent agents in driving the development of large models. Wang highlighted the transformative impact of intelligent agents, comparing their cognitive models to human-like learning processes. He showcased Baidu's Intelligent Coding Assistant, which democratizes programming, enabling individuals to code more efficiently. With an adoption rate of 46% and a significant increase in new code contributions, the platform empowers both novice and professional programmers alike.
Wang revealed impressive statistics regarding Baidu's AI initiatives, with millions of users and native AI applications. Leveraging the collaborative optimization of the Paddle deep learning platform, Baidu has achieved remarkable efficiency gains in model training and inference. This ecosystem has fostered the growth of millions of developers, tens of thousands of enterprises, and hundreds of thousands of models.
Experts and industry leaders convened to discuss the ethical and governance challenges posed by AI development. Recognizing AI's transformative potential, speakers emphasized the importance of responsible innovation and global cooperation in addressing emerging challenges. Efforts to establish ethical guidelines, enhance transparency, and promote inclusive governance reflect a collective commitment to shaping AI's impact.
As Beijing continues to flourish in the AI field, breakthroughs in research and applications pave the way for transformative change. From pioneering technologies like Taiji to initiatives promoting accessible AI, the city's ecosystem fosters innovation while addressing ethical and governance considerations.
With collaborative efforts from academia, industry, and policymakers, Beijing is poised to help shape the future of AI, driving progress towards a more inclusive and sustainable society.