The roles scientific and technological advances have played in our lives are varied. For every gun, bomb and missile manufactured, humans have also developed genetically modified crops and medicine. But with countries increasingly focussed on the military capabilities offered by Artificial Intelligence (AI), spending billions in the process, a tipping point is approaching.
Armies around the world are already using a raft of deadly AI military machines, explains Professor Noel Sharkey of Sheffield University in an exclusive interview.
Swarms is the big buzz word at the minute, where you have overwhelming amounts of attacking robots controlled by very few people
He said: “The thing that scares me the most is Kalashnikov already built the Armata T-14 super tank which is 10 – 20 years ahead of anything else on the planet.
“The Armata is an artillery gun capable of finding targets by itself, trained by near networks or deep learning.
“The Armata can already be remote-controlled and Russia is trying to make it as autonomous as possible, and they are talking of using them in swarms.
AI apocalypse: Could an AI arms race building military machines trigger another world war?
AI apocalypse: Billions is being spent on autonomous killing machines
“Swarms is the big buzz word at the minute, where you have overwhelming amounts of attacking robots controlled by very few people.
“And Russia are talking about sending swarms of these super tanks to the European border.”
And this is not the only devastating AI weapon already in production.
Professor Sharkey said: “The X47-B is an unmanned fighter jet prototype, capable of refuelling by itself midair, with 10 times the normal reach of the F-35.
“This was designed to counter the threat of Chinese hypersonic aircraft-busting missiles.
“The US’s idea was to have swarms of autonomous vehicles to fight the Chinese but the trouble with America’s simulations was they were based on the presumption of the Chinese military stagnating, which of course has not happened.
“The Chinese have the world’s strongest military and have a great advantage over other, as they don’t have to worry about dissenters.”
And the most autonomous weapon currently in production is the Israel-made IAI Harpy radiation detection loitering munition.
AI apocalypse: The Israel-made IAI Harpy is arguably the most autonomous weapon in production
AI apocalypse: The X47-B unmanned fighter jet was designed to counter the Chinese
Professor Sharkey explains: “It looks like an aircraft with wings, but it is launched like a rocket.
These are fired before an airstrike and when they detect radar signals, they look them up on a database and if it is not one of the Israelis’, it transforms into a dive-bomb and blows up the radar system, presumably connected to anti-aircraft missiles.”
However, the issue with this weapon is it is completely indiscriminate: if a radar was placed on a hospital or school roof, it would not know the difference.
Professor Sharkey, who has 40 years experience in the AI sector believes there are several areas for concern with AI military machines.
He said: “Once you get systems in place that turn out to be really reliable, you begin to rely on such machines alone.
“This is called automation bias, where you come to rely on a machine to do a job.
“The AI robots are nearly always right except for when they go badly wrong.”
Another fear is how AI robots could potentially destabilise global security.
AI apocalypse: Automation bias is where you come to rely on a machine to do a job
AI apocalypse: There are several areas for concern with AI military machines
Professor Sharkey said: “The issue here is wars accidentally being started.
“We are already seeing the beginnings of an arms race among weapons manufactures at the very least.
“There is really big money to be made in this sector, and that will mean this arms race will spread widely.
“People talk about how AI weapons will save soldiers lives but I believe that would be true only for a short time.
“Once everybody possesses AI weapons we will be back to square one, only things will much much worse than they were.”
And there is additionally the moral issue about whether the decision to kill a human should ever be delegated to a machine.
As for a solution to concern over an AI apocalypse, Professor Sharkey believes a new legally-binding international treaty to be ratified by the UN is vital.