Hawking & Wozniak Say No To AI Weapons

"Comic-Con 2004 - Terminator statue" by popculturegeek.com - originally posted to Flickr as Comic-Con 2004 - Terminator statue. Licensed under CC BY 2.0 via Wikimedia Commons - https://commons.wikimedia.org/wiki/File:Comic-Con_2004_-_Terminator_statue.jpg#/media/File:Comic-Con_2004_-_Terminator_statue.jpg

“Comic-Con 2004 – Terminator statue” by popculturegeek.com – originally posted to Flickr as Comic-Con 2004 – Terminator statue. Licensed under CC BY 2.0 via Wikimedia Commons – https://commons.wikimedia.org/wiki/File:Comic-Con_2004_-_Terminator_statue.jpg#/media/File:Comic-Con_2004_-_Terminator_statue.jpg

Leading figures in the science and tech world have called for a ban on weapons that can “select and engage targets without human intervention.” Stephen Hawking and Steve Wozniak are among those who say such weapons would undermine the wider use of artificial intelligence.

The letter was released to mark the International Joint Conferences on Artificial Intelligence, an event held this week in Buenos Aires.

It deals with what’s currently a hypothetical but, according to the signatories, increasingly plausible possibility: autonomous weapons. The letter gives the example of quadcopters with the ability to seek out and attack people who meet particular criteria. It’s possible this could involve assassinating specific individuals or even simply targeting people based on appearance to carry out so-called ethnic cleansing.

The signatories say their objection is not to all technology used in warfare. Their definition of autonomous weapons doesn’t cover missiles or drones where both the flight path and the act of firing are controlled entirely by humans.

According to the letter, were any military power to start using autonomous weapons, it would create a particularly dangerous arms race: unlike with nuclear weapons, virtually any “significant military powers” could afford autonomous weapons.

The letter concludes that weapons which take advantage of artificial intelligence could provoke “a major public backlash against AI that curtails its future societal benefits.”

As an example of how a global ban could work, the signatories point to a convention that took force in 1998 and now has the support of more than 100 countries banning the use of any laser weapon specifically designed to permanently blind people. It was the first time in over a century that such a treaty banned a type of weapon before it had actually been used in combat.