Promoting Artificial Intelligence (AI) to Empower Society (Technology Perspective)

Reporter Liu Shiyao and Yu Sinan

In recent years, the emergence of artificial intelligence large model technology has set off a research boom in the development of a new generation of artificial intelligence, and also brought new development opportunities for the application of artificial intelligence in all walks of life.

What are the possibilities of the future of artificial intelligence, and how should we deal with it? The reporter interviewed relevant experts.

Artificial intelligence research is shifting from academic traction to demand traction

As an important driving force for a new round of scientific and technological revolution and industrial transformation, artificial intelligence is widely believed to have a profound impact on economic and social development.

In recent years, artificial intelligence research is shifting from academic traction to demand traction, and the development goals of artificial intelligence are undergoing a major transformation.

In the view of Wu Fei, director of the Institute of Artificial Intelligence of Zhejiang University, artificial intelligence is expanding from the original "simulating human intelligence with computers" to three directions: "machine + human" (combining machines and people into an enhanced hybrid intelligent system), "machine + man + network" (combining machines, people and networks to organize into a new group intelligence system), "machine + man + network + thing" (using machines, people, networks and things to form a more complex intelligent system such as a smart city). Driven by the application of "empowering society", a series of smart technologies are booming.

Liu Wei, director of the Human-Computer Interaction and Cognitive Engineering Laboratory of the School of Artificial Intelligence of Beijing University of Posts and Telecommunications, introduced that from the use of mathematical models to establish machine intelligence, the realization of machine learning with the help of statistical probability, and the attempt to realize context-aware behavioral intelligence systems with the help of transfer methods, these three types of artificial intelligence development trends reflect the development path towards human-machine environmental systems. "The ability to perceive, understand and interact with the environment is a general artificial intelligence in the true sense. Liu Wei said.

In 2017, the State Council issued the "Development Plan for the New Generation of Artificial Intelligence", which clarified a variety of technical forms of artificial intelligence, including cross-media intelligence, swarm intelligence, hybrid augmented intelligence, etc.

For example, Wu Fei said that the surgical operations involved by surgical robots cannot be completed by doctors or robots alone, showing the characteristics of human-machine collaborative hybrid enhanced intelligence. "More verticals are also accelerating the exploration of human-computer interactions, either by introducing the role of humans into intelligent systems or by introducing biologically inspired intelligent computing models into AI systems. Through continuous interaction and mutual learning, man and machine can achieve the purpose of continuously improving their own intelligence level. Wu Fei said.

Avoid AI risks through technical and legal means

Artificial intelligence benefits humanity, but it also has potential risks.

"The risks of artificial intelligence mainly exist in the following possible situations: first, malicious use of artificial intelligence and misuse of artificial intelligence, second, software vulnerabilities and hardware failures cause artificial intelligence failures, and third, environmental changes cause artificial intelligence to get out of control. Liu Wei said.

The risk of inappropriate use of AI has been a cause for concern. For example, AI technology can collect and analyze massive amounts of data, which may lead to personal privacy breaches, data breaches, and cyber attacks, and AI technology can generate fake photos and videos to facilitate online fraud and other illegal acts. The advanced scenarios depicted in sci-fi films on the theme of artificial intelligence also make people think about the ethical issues that AI technology may bring, for example, how will humans get along with "intelligent" machines in the future?

In Wu Fei's view, human beings have always been the master switch and determiner of the height, breadth and depth of artificial intelligence, as well as the coordinator of people and artificial objects. Therefore, on the one hand, we should not simply equate artificial intelligence with the human brain and blindly believe in the idea that "artificial intelligence replaces humans", but on the other hand, we should also be good at using artificial intelligence as a human helper to create a better future in human-machine collaboration.

Experts believe that AI-related risks can be avoided through technical and legal means, and humans have the ability to find effective means to mitigate them.

China has always been an active advocate and practitioner of AI governance. In September 2021, the National Committee for the Governance of New Generation Artificial Intelligence issued the Code of Ethics for the New Generation of Artificial Intelligence, emphasizing the integration of ethics into the entire life cycle of artificial intelligence. In October 9, China released the Global Artificial Intelligence Governance Initiative during the 2023rd Belt and Road Forum for International Cooperation, which systematically elaborated China's plan for AI governance in three aspects: AI development, security, and governance, fully reflecting China's responsibility as a major country in dealing with AI governance issues.

Establish a good ecosystem to support AI-powered society

In recent years, more and more countries have placed the development of artificial intelligence at the national strategic level, and added concepts such as preventing the monopoly of artificial intelligence technology and the disorderly expansion of technology to the top-level design.

How to prevent the misuse of AI technology?

According to the interviewed experts, multidisciplinary and multi-field cooperation should be established, and scientists, engineers and the public should be encouraged to participate in discussing the development direction, application fields and ethical issues of artificial intelligence. The whole world should work together to adhere to the principle of equal emphasis on development and security, and ensure that the development of AI technology is in the overall interest of mankind.

Liu Wei said that from the perspective of technical means, in order to increase the fairness, trustworthiness, explainability, transparency, security, etc. of artificial intelligence, the R&D subject should maintain the design ideas of the human-machine environment system in the entire process such as top-level design, process control, and project completion.

Tian Tian, CEO of Rely Intelligence, believes that it is necessary to do a solid job in tackling the basic theories of artificial intelligence, improve the intrinsic security performance of artificial intelligence, forge and reserve the security evaluation ability of the new generation of artificial intelligence, and master the methods and means to evaluate its accuracy, efficiency, content security, anti-attack ability and other dimensions. At the same time, efforts should be made to develop and provide necessary security tools, such as providing deepfake detection tools to help the public identify the authenticity of information and prevent online fraud.

Data and algorithm security is at the core of AI security. Experts said that it is necessary to pay attention to the construction of high-quality corpus, optimize and train through fine standard data and feedback data, and finally realize the selection and elimination of the inferior model output.

Wu Fei believes that promoting AI technology to empower society requires the overall support of a good ecosystem, not just technology research and development itself. It is necessary to grasp the characteristics of the high degree of integration of artificial intelligence technology attributes and social attributes, and strengthen forward-looking prevention and restraint guidance. "A good ecosystem that connects ethics and morality, technology and industry, capital and markets, and every AI actor from different fields can benefit from it. Wu Fei said.