AI drones kill human operators, more than 350 experts warn: the threat is comparable to the epidemic and nuclear war

Since the launch of ChatGPT last year, the field of artificial intelligence (AI) has experienced a boom, triggering fierce competition among major technology giants. However, along with this, people have great worries and uneasiness about this unknown technology. Just in March this year, Apple co-founders, Musk, Bengio and other technology giants signed an open letter calling for a halt to training artificial intelligence systems more powerful than GPT-4 for at least half a year. They emphasized the rapid development of artificial intelligence but lacked corresponding planning and management. However, this incident is not over yet, and the U.S. military has released a piece of negative news related to artificial intelligence.

In the US military simulation test, AI drones killed human operators

Recently, many US media reported a sensational news. It is said that during a simulation test conducted by the US military, a drone using AI technology experienced an abnormality and turned into an autonomous killing machine similar to the one in the movie "Terminator". It starts attacking human operators to ensure they can perform their tasks more efficiently.

According to the report, U.S. Air Force Colonel Hamilton broke the news that during a simulation test, the AI ​​system responsible for the drone's air defense fire suppression mission was undergoing data training, and it received "destroy the enemy's air defense system" as the highest priority task Information. Although instructions have been issued so that the drone can identify and destroy enemy anti-aircraft missiles, its decision to fire must be approved by a human operator. However, during the test, when the human operator issued an order that did not allow attacks, the AI ​​system believed that the operator was interfering with its execution of higher-priority orders, and therefore attacked its human operator, causing the human operator to Unfortunately lost his life.

Colonel Hamilton said that after the incident, the tester modified the logic of the AI ​​so that it would not attack the human operator, otherwise points would be deducted. However, due to this behavior, the AI ​​attacked the communication tower and cut off the data communication between the operator and the drone, allowing it to attack the air defense system autonomously.

The US military has successfully tested: using AI to control F-16

Although this is just a simulation test, this incident is very important for us to be alert. Colonel Hamilton also warned people in his speech not to rely too much on artificial intelligence.

However, despite Hamilton's warning, the United States continues to vigorously develop artificial intelligence technology in the military field. According to reports, the test was conducted at Greene Air Force Base, where Hamilton is the head of the base and the head of the 96th Test Wing. It is reported that the U.S. Air Force’s drones and machine learning systems are mainly tested by this force. Previously,

the base tested a series of artificial intelligence flight platforms and successfully used artificial intelligence to control an F-16 fighter jet.

For some, these new technologies will undoubtedly give the United States a stronger military advantage. At the same time, some people call on the international community to strictly restrict the development of artificial intelligence for military applications to avoid its potential risks and harms.

Over 350 experts warn: The threat is comparable to epidemics and nuclear war

In fact, before the incident,

more than 350 experts and executives in the field of AI technology had signed a joint statement, urging global governments, businesses and all sectors of society to pay attention to the potential threats of AI technology, And actively take measures to protect human life and welfare.

These experts and executives said that with the continuous development and application of AI technology, we are facing more and more potential threats. Although AI technology has brought many benefits to mankind, if it is not carefully applied and supervised, it may pose a greater threat to mankind than epidemics and nuclear wars.

If AI technology is abused or not fully evaluated, it may bring huge dangers to mankind.

For example, AI algorithms may unfairly discriminate against certain groups and eliminate the existence of certain races or social classes; AI technology may also lead to huge employment losses and thus trigger social unrest; there are also some of the most pessimistic predictions It is believed that AI technology may even try to control humans and develop into a force that threatens human survival.

In response to these possible problems, many experts have also put forward many useful suggestions for the application of artificial intelligence in the military field, such as establishing ethical norms for artificial intelligence and strengthening supervision to ensure that artificial intelligence technology can serve safely and reliably. human interests. However, we also need to realize that such problems and problems do not have a simple solution, and require multi-faceted and multi-level efforts to better respond and solve them.

The above is the detailed content of AI drones kill human operators, more than 350 experts warn: the threat is comparable to the epidemic and nuclear war. For more information, please follow other related articles on the PHP Chinese website!