Will AI Advances Cause Nuclear War? artificial intelligence

Will AI Advances Cause Nuclear War?

The world is focusing on research to secure competitiveness in robot development, such as the introduction of artificial intelligence, ahead of the 4th industrial revolution. However, some academics have pointed out that if a nuclear war breaks out, it will be related to the development of artificial intelligence.

Tesla founder Elon Musk said the race for dominance in artificial intelligence technology would be the cause of World War III.

A report published by the Rand Institute, which specializes in US military affairs, supports this view. Doug Irving, author of the Rand Lab blog, said, “Recent advances in artificial intelligence technology have given rise to machines capable of learning and thinking.”

The online tech site Motherboard cited the study, noting that it is inevitable for machines to reach superintelligence that surpasses human thinking.

In addition, the report presented two scenarios that could occur with superintelligence. The first is the prospect that superintelligence will solve human problems. The second is the prediction that humans will become extinct by superintelligence. It is pointed out that positive superintelligence will save mankind from nuclear war, but in the opposite case, superintelligence causes human extinction in various ways other than nuclear war.

This research is part of the 'Security 2040' special project jointly conducted by Rand Research Institute. In this project, artificial intelligence experts and nuclear strategy experts were invited and held meetings and workshops.

Artificial Intelligence Development and Expected Scenarios

The Rand Institute had experts envision a nuclear weapons system in 2040 and predict the positive and negative impacts of artificial intelligence at that point in time. Experts gathered at the workshop presented three possible scenarios as follows.

1. Doubts about artificial intelligence

Many AI experts predict that by 2040, artificial intelligence will not advance enough to play an important role in nuclear war decisions.

2. Aggravation of tensions due to nuclear weapons

Nuclear strategists argue that perfect artificial intelligence is not needed to exacerbate tensions from nuclear weapons. Even the mere thought of a country under a nuclear attack triggers a response. This will lead to competition and additional distrust between countries with nuclear attack capabilities.

3. AI, learn that non-participation is a winning strategy

Some experts predict that advances in artificial intelligence will rather prevent the possibility of nuclear war. Artificial intelligence, for example, could track nuclear disarmament activities to monitor countries' compliance with non-proliferation treaties.

In addition, artificial intelligence can prevent mistakes or bad decisions from being made under the psychological pressure of a nuclear weapons confrontation. As one expert said, there is a possibility that artificial intelligence will realize that the only way to win a nuclear war is by not attending, as in the 1983 movie 'War Game'.

false alarm

The incident that occurred in the United States and Russia in 1983 proved the possibility of a nuclear war caused by computers and artificial intelligence.

Russian Colonel Stanislav Petrov was instrumental in preventing an all-out nuclear war between the United States and Russia at the time. According to his judgment, Colonel Petrov treated the nuclear missile attack warning flashing on his computer as a false alarm.

That night, Colonel Petrov had to treat the attack alert on his computer as false three times. He showed the courage not to issue orders to launch a Russian nuclear missile in the face of a false nuclear attack from the United States on the screen.

It was later discovered that a Russian computer system mistook the twinkling of sunlight above the clouds for a US nuclear missile. As a result, Colonel Petrov's composure and intuition saved the whole world, not just Russia and America.

Of course, the computer at the time was a technology in 1983, and it should be taken into account that many technological advances have been made since then. However, it is an event that awakens the fact that human intuitive judgment can be clearer than mechanical information gathering.

what are the risks

Computers today can scan thousands of photos to find patterns that the human eye cannot see. The future is not far off, when a nation's weapons and defense systems can be accurately predicted from drone footage, satellite reconnaissance, or even social media posts.

Such systems leave room for competing nations to think that their enemies are not capable of counterattacking if they launch a nuclear strike. Therefore, it is trying to make its military power equal to that of its enemies.

Such scenarios increase the likelihood of nuclear war. The US, for example, is developing automated ships that can detect enemy submarines within thousands of miles. China and Russia are also rapidly pursuing them, and China is developing a swarm intelligence algorithm that allows multiple drones to attack enemies at the same time. Russia, on the other hand, is developing underwater drones that can cross oceans and launch nuclear warheads large enough to destroy major cities.

Russian President Vladimir Putin even said that the country with the upper hand in AI technology will rule the world.

Edward Geist, assistant policy researcher at the Rand Institute, said, "As artificial intelligence develops, people will think that they lose the moment they hesitate. ” he asserted.