Dialogue in birthplace of Confucius pools wisdom for AI governance

(Xinhua) 15:23, June 28, 2023

JINAN, June 28 (Xinhua) -- A conference pooling wisdom for building a digital civilization in the age of artificial intelligence concluded on Tuesday in Qufu, east China's Shandong Province, the birthplace of the ancient Chinese philosopher Confucius.

The three-day World Internet Conference (WIC) Nishan Dialogue on Digital Civilization, themed "The Age of AI: Building a Digital World of Exchanges, Mutual Learning and Inclusiveness," was the first thematic event the WIC has held as an international organization.

Zhuang Rongwen, director of the Cyberspace Administration of China, urged efforts to seize AI development opportunities, explore good strategies for AI governance, jointly promote exchanges and mutual learning among human civilizations in the AI era, and build a community with a shared future in cyberspace.

Large language models, represented by ChatGPT, are the focal point of global scientific and technological innovation, hence a hot topic in the dialogue.

"Over the past year, the development of AI has been turbocharged, undergoing a generational change almost every week, in terms of technology, products and applications," said Li Yanhong, CEO of Baidu, at the conference.

Li said it is foreseeable that large language models will encroach on more sectors, adding that the deep fusion between the digital economy -- predominantly driven by large language models -- and the real economy will give rise to profound changes in industries and socioeconomic development.

He took car-making as an example, and predicted that large language models would significantly curtail the time and money invested in R&D.

While it took seasoned engineers massive efforts to work out proper configurations among over 20,000 parts and hundreds of thousands of parameters, apart from drawing up blueprints by themselves, large language models can save them all these troubles through efficient information processing and automatic generation of blueprints, according to Li.

Acknowledging the tremendous benefits of the rapidly evolving AI technology, participants also urged caution against AI-related challenges such as security hazards, privacy concerns and ethical problems.

"GPT-4 has achieved the first form of artificial general intelligence, a new entity among us that will change many of our existing societal structures and mechanisms. This requires us to rethink what the impact will be, and how we manage it requires debate and careful thinking," said Bart Selman, a professor at Cornell University.

Zhang Yong, the chairman and CEO of Alibaba Group, warned that the global race on AI has heated up. However, technological ethics in related fields have yet to be clearly defined. "It is equally critical to apply the brakes skillfully while pressing the accelerator."

Zhang applied the ancient Chinese teaching that "A person's virtues take precedence over his or her merits" to the realm of AI, saying technological evolution must conform to social norms and bring more benefits to human society.

"While chasing after cutting-edge technologies, we must hold on to our social responsibilities," said Zhang.

"How are we going to address the AI-related challenges? We'd draw wisdom from Chinese civilization and successful practices of Chinese modernization," said Zhang Weiwei, director of the China Institute of Fudan University in Shanghai. He compared AI to a double-edged sword that, while providing convenience to human beings, may upend business-as-usual in many industries or even be weaponized, eventually threatening the entire human race.

However, Chinese civilization honors a tradition of seeking harmony without uniformity. As no country can take on the challenge of AI single-handedly or stay out of its impact, all countries need to join hands in steering the course of AI development to the benefit of humankind, Zhang Weiwei added.

Ren Xianliang, secretary-general of the WIC, expressed wishes to work with all parties to explore the viable paradigm of global governance of AI and jointly address opportunities and challenges the technology of AI brings to human civilizations.

Approximately 400 representatives of various sectors worldwide joined in the dialogue online and offline.

(Web editor: Zhang Kaiwei, Liang Jun)


Related Stories