China News Service, Beijing, June 11 (Reporters Chen Hang and Lu Xiaona) AI+ basic research, AI+ scientific instruments, AI+ black light laboratories... China News Service reporters visited a number of new R&D institutions and technology companies in Beijing on the 11th, and experienced the accelerated penetration of artificial intelligence into thousands of industries, and the continuous "new" application scenarios.
As the city with the richest AI innovation resources in the country, Beijing has promoted the integration of technological innovation and industry in recent years and systematically laid out the development of the AI industry. Data shows that there will be more than 2,500 AI companies in Beijing in 2025, with a core industry scale exceeding 450 billion yuan, and 241 large models have been registered, continuing to rank first in the country.
Large models accelerate the move from the digital world to the physical world. Wang Zhongyuan, director of the Beijing Zhiyuan Artificial Intelligence Research Institute, said that the institute has launched the world's first multi-modal neuroscience model that unifies understanding and generation, realizes the mutual conversion of brain signals into text and images, and solves the problem of cross-individual and cross-species brain signal transmission in traditional neuroscience.
With the support of this model, the brain science research results jointly carried out by the institute and Tsinghua University were published in the international academic journal Science on June 4, demonstrating the potential of AI to participate in basic research in complex life sciences. Wang Zhongyuan said that the institute is cooperating with Xuanwu Hospital of Capital Medical University to use AI to support personalized diagnosis and treatment of Alzheimer's disease; it is also working with relevant institutions to develop brain signal encoding and decoding and brain-computer interface virtual signal output to realize dream analysis.
After equipping scientific instruments with "AI brains", the scientific research paradigm has changed from "manual operation" to "intelligent operation". Zhang Zezhong, a researcher at the Beijing Institute of Scientific Intelligence, said that the focused ion beam scanning electron microscope is a core scientific instrument in fields such as materials research and development. It relies mostly on expert experience and manual operations, which is time-consuming and labor-intensive. The institute, Peking University and Shenzhen Technology jointly released an intelligent dual-beam electron microscope system, which allows AI to automatically identify sample area characteristics, calibration position and tracking, and "optimizes" the process through reinforcement learning, enabling the instrument to be unattended for more than 8 hours, and the sample preparation success rate has jumped to more than 90%.
In Beijing Dana Experimental Technology Co., Ltd., the self-learning AI+ black light laboratory has upgraded acrylonitrile detection from more than a dozen tedious manual processes to automated workstation operations. The robot completes sampling, detection, report generation and waste liquid treatment in one stop, freeing humans from high-risk operations.
"The combination of AI and robotics technology can achieve visual recognition, perceptual decision-making, action execution, experience summary and self-evolution." Chi Haipeng, chairman of the company, gave an example. In the testing of shampoo, body wash, etc., a person with rich operating experience makes an error an average of twenty times before getting it right once. However, AI can summarize the rules of success after ten times of operation. As long as it succeeds once, the correct result can be stably reproduced. He hopes that AI will free scientific researchers from repetitive manual labor and accelerate exploration and breakthroughs in energy and other fields.
Wu Jun, deputy director of the Information Center of the Beijing Municipal Science and Technology Commission and Zhongguancun Administrative Committee, said that Beijing will focus on life sciences, materials science and other fields, encourage dominant innovation entities to open up core scenarios, organize enterprises to jointly tackle key problems with universities, scientific research institutes, and new R&D institutions, and create a number of demonstrative, leading, and breakthrough application benchmark demonstrations. (over)
