Artificial intelligence (AI) — a top priority for tech companies in the United States, for Industry 4.0, and for digital China — is already reshaping global business. But this major scientific and technological disruption will also deeply impact relations between powerful nations.
While narrow AI has moved from labs to our daily lives, high-profile personalities such as Stephen Hawking, Bill Gates and Elon Musk have rightly raised concerns about the risks inherent with AI capable of equaling, or even surpassing, human intelligence.
Anticipating the emergence of even more powerful and increasingly autonomous AI reinforced by quantum computing, these engaged voices are asking for a collective reflection upon what could constitute a challenge to mankind, a technology that could dominate its creator.
The recent win of the AlphaGo computer program over the Korean Go champion Lee Se-dol was indeed a strong signal of the rapid development of machine learning at the intersection of computer science and neuroscience.
However, a more immediate danger connected with the advancement of intelligent machines is an AI fracture enlarging what is already known as the digital divide.
While AI’s algorithms and big data increase the productivity of a small segment of the global village, half of the world’s population still does not have access to the Internet. “Don’t be evil” can be Google’s slogan, but exponential technologies carry with them the risks of unprecedented inequalities.
While AI’s social and political effects are often discussed, the geopolitical implications of the “fourth industrial revolution” — the introduction of advanced technology to manufacturing, also known as Industry 4.0 — have been surprisingly absent from public debate.
How could AI affect Sino-Western relations or, more specifically, Sino-US relations — the major determinant of today’s international order? For decades, nuclear weapons stood as the frightening symbols of the Cold War; will AI become the mark of a 21st century Sino-Western strategic antagonism?
For humanity, the atomic age has been a time of paradoxes. In the aftermath of the 1945 Hiroshima and Nagasaki nuclear bombings, an arms race involving the most lethal weapons defined US-Soviet relations in what also constituted a permanent existential threat to human civilization.
But analysts will also argue that it is the doctrine of mutually assured destruction acting as a deterrent among rational actors which prevented a direct conflict between the two superpowers.
As the 2015 Plan of Action for the Iran nuclear program demonstrates, 70 years after Hiroshima and Nagasaki, world powers actively collaborate to avoid nuclear proliferation, even if North Korea appears to be an exception.
But the Sino-Western convergence of views on the issue of nuclear proliferation does not apply in cyberspace. Despite a certain level of interconnection between some private Chinese and US Internet companies and financial institutions, overall Sino-US relations in cyberspace are characterized by strategic mistrust.
Besides, in space science and in the exploration of the universe, the US and China are following two separate courses. While China prepares to operate its own modular space station, the International Space Station shows that in this strategic field, the West can work with Russia but not with China.
Any responsible approach to AI has to take into account the combined lessons of the atomic age, digital dynamics and space exploration. Should Western AI and Chinese AI develop on two separate trajectories, one would dangerously increase the risks of creating an irreversible Sino-Western strategic fracture.
In this context and following the appreciation of interaction between AI and global politics, an international AI agency should be established on the lines of the International Atomic Energy Agency (IAEA).
It was in his Atoms for Peace address to the United Nations General Assembly that US president Dwight D Eisenhower proposed in 1953 the creation of the IAEA. Today, our actions must be guided by the spirit of “AI for mankind”.
A UN-led international AI agency involving academics, private businesses, global civil society and, of course, governments should at least give itself the following four objectives.
It has to create the conditions for AI’s awareness across societies and for a debate to take place on AI’s ethical implications. Scientists, engineers, entrepreneurs, legal experts, philosophers and economists have to analyze AI from all possible angles, its future, and its potential effects on humanity.
This international body should take all possible action to prevent an AI fracture that would dangerously enlarge the digital divide. It is not acceptable to have, on the one side, a tiny segment of humanity making use of a series of human enhancement technologies and, on the other side, the vast majority of the world population becoming diminished.
The agency should ask for transparency in AI research at both government and corporate level.
The issue of nuclear proliferation and therefore the creation of the IAEA followed the secretive Manhattan Project and the use of nuclear bombs to end the war in the Pacific. If humanity really wants to protect itself from the military use of strong AI and its tragic consequences, it has to define a set of rules and policies that would maintain research within reasonable and collectively accepted limits.
The IAEA imperfectly manages an existing threat. The AI agency would aim at preventing the realization of what could be an even greater danger.
An international AI body should encourage knowledge sharing and international cooperation. Elon Musk’s OpenAI initiative is certainly a constructive force encouraging openness and collaboration, but the “AI for mankind” ideal cannot depend only on a group of private entrepreneurs.
AI, more than any other technology, will impact the future of mankind, and it has to be wisely approached on a quest toward human dignity and not blindly worshipped as the new master of a diminished humanity. It has to be a catalyst for more global solidarity and not a tyrannical matrix of new political or geopolitical divisions.
David Gosset is director of the Academia Sinica Europaea at CEIBS and founder of the Euro-China Forum. He also established the New Silk Road Initiative.