ON WEDNESDAY, over 50 artificial intelligence researchers reportedly co-signed a letter announcing that they will be boycotting a university in the Republic of Korea until it pledges to refrain from developing AI weapons without "meaningful human control". Beijing News commented on Monday:

According to reports, a laboratory, co-sponsored by Korea Advanced Institute of Science and Technology and an unidentified enterprise, plans to develop AI-based weapons.

It should be noted that the AI researchers opposed "autonomous weapons", but many media outlets have mistaken that for "smart weapons" in their reports. Actually, there is a deep redline between the two, and that redline should never be crossed.

Smart weapons have long been in existence. For example, a smart missile launcher might be able to automatically locate its target, a military Unmanned Aerial Vehicle might identify its route with the help of sensors and positioning systems.

However, no matter how smart they are, they are still under human control. It is a person that decides when to pull the trigger.

Autonomous weapons are different. Just like the researchers said in their open letter, such weapons "lack meaningful human control" and it is the AI that decides whether to launch a strike upon a military target.

That is, of course, extremely dangerous. Most military targets are humans or have humans inside, and it is deep violation of human dignity if a machine is empowered to decide a human fate.

Worse, AIs lack moral or ethical obligations. In case a total war breaks out, they might totally destroy humankind. That's why the majority of AI researchers oppose the development of autonomous weapons. Armed conflicts might not be totally eliminated, but at least that bottom line should not be crossed.