핵 통제 실패 80년, AI는 다를까 [유레카]

Seventy years ago, in August 1945, atomic bombs dropped on Hiroshima and Nagasaki reduced the cities to ruins and claimed hundreds of thousands of lives. The international community responded with fierce competition to control nuclear weapons. The scientists behind the Manhattan Project, which developed the atomic bomb, argued that nuclear weapons should be under international control while atomic energy should be used only for peaceful purposes.

In December 1945, the United States, United Kingdom, and Soviet Union agreed to establish the United Nations Atomic Energy Commission (UNAEC). In January 1946, the United States released the Achinstein-Reilerson report, proposing that the commission manage all uraniu mine sites and weaponization facilities, allowing peaceful use only for individual nations, with the ultimate goal of eliminating nuclear weapons.

However, President Harry Truman demanded assurances that the Soviet Union would not develop nuclear weapons, and the UNAEC's American representative, Bernard Baruch, released the Baruch Plan in June 1946. This plan included safe harbor for nuclear facilities and veto rights for the five permanent members of the Security Council, which the United States refused to accept.

The Soviet Union's first nuclear test in 1949 ended American nuclear monopoly and unchecked nuclear arms race continued. Now, history is repeating itself with artificial intelligence (AI). In a recent statement at the UN General Assembly, former President Donald Trump vowed that the United States would reject any "globalist scheme" to control AI, warning that it would not stifle the "next industrial revolution."

China's response is the same, suspecting that any regulation on American AI companies like Anthropic is aimed at monopolizing AI and following in its footsteps. The same response as the Soviet Union's. The United Kingdom's Oxford Future of Humanity Institute research team sought lessons from the initial attempts to control atomic bombs in their 2021 report, "Strong Technology Control: Lessons from the Baruch Plan."

They warned that the expectation that political leaders would voluntarily manage such risks is misplaced and that public opinion and technology experts have a crucial role to play in international cooperation. Ultimately, it falls to the most knowledgeable experts on technology and awake citizens to lead the way.