Shanghai Finance President Liu Yuanchun: AI governance must be urgently advanced, not a conventional tech upgrade

Shanghai – Artificial intelligence is not a conventional technological upgrade but a "super technology" that is fundamentally reshaping economic growth, higher education, and governance systems, according to Liu Yuanchun, President of Shanghai University of Finance and Economics.

Speaking at the "Digital Humanities and Innovation Ecosystem" high-end dialogue held on March 24, Liu argued that the impact of AI on society and economic growth follows a fundamentally different path from that of traditional technologies.

"Actually, the path through which our traditional technologies affect growth is different from the path through which this super technology affects human society and economic growth," Liu said.

He stressed that understanding AI requires recognizing its disruptive nature. "We need to rethink AI as a super technology. The intensity, time path, and manner of economic growth it may trigger go beyond our general concept of technology."

Three Layers of Governance Challenges

Addressing the frequently discussed topic of AI legislation and governance, Liu outlined several fundamental considerations that must be confronted.

1. Human Agency at Risk

The first and most critical concern, Liu said, is whether AI will ultimately escape human control and replace humans as a new form of consciousness and agency.

This concern is not unfounded, Liu noted, pointing to public calls from leaders of major technology companies for AI regulation. The core reason, he said, is that unchecked development of the technology "could bring devastating impacts to humanity."

Liu cited recent remarks by Elon Musk, who specifically referenced the situation surrounding ChatGPT-6 in a recent video. He urged non-technical professionals to take such signals seriously: "If what these people say is true, then humanity as a whole must systematically guard against the potential destructiveness of this technology in the future. We must comprehensively regulate AI and establish safety nets."

2. Efficiency vs. Equity

The second layer of governance concerns the balance between efficiency and fairness. Liu referenced an open letter from July 13, signed by nearly 200 researchers and entrepreneurs, including 16 Nobel laureates. The letter conveyed three core consensuses:

• This AI technology is fundamentally different from previous ones
• Its impact and implications exceed traditional expectations
• Governments must establish corresponding regulatory and preventive mechanisms

Liu observed that as AI technology has advanced over the past two years, many economists who were previously pessimistic about its impact have changed their views. For example, economist Daron Acemoglu has argued in recent research that AI will affect employment structures and widen wealth inequality.

"Even if it does not bring about devastating consequences at the technical level, it will certainly impact our existing social structure," Liu warned. He called for governance to be moved significantly forward, with a core focus on proactively preparing the entire social system to address the shocks AI may bring to traditional frameworks, even as it benefits humanity.

3. Moving Beyond Spontaneous Order

The third layer of governance, Liu said, involves moving beyond spontaneous order and building a "subject foundation" for technology to serve the public good.

From a historical perspective, every major technological revolution has triggered a revolution in production methods. As a result, the relationship between technology and the humanities has become the core foundation of governance, Liu argued.

The impact on the humanities, the reconstruction of subjective meaning, and the reaffirmation of subjective value are not only key pathways to "technology for good" but also the core defense against technological risks. Liu noted that top technical experts such as Albert Einstein and J. Robert Oppenheimer often turned to philosophical thinking in their later years.

Unlike the three previous industrial revolutions, AI's unique technological characteristics mean that humanity cannot passively wait for the system to self-correct, Liu said. "The idea that spontaneous order can automatically achieve a soft landing must be abandoned in the macro-level organizational planning of society."

Only through proactive institutional design, he concluded, can humanity fully enjoy the growth momentum brought by AI while effectively mitigating its potential impacts on human agency and social structures.

(Source: Yicai)