A SUPERPOWER arms race to build killer robots could wipe out humanity if left unchecked, experts fear.
The doomsday warning comes after a UN conference failed to agree a ban on Terminator-style “slaughterbots” – which are being developed by China, Russia and the US.
Major powers are investing billions to create advanced AI weapons that can hunt and strike targets with no input from controllers.
Last year a Turkish-made kamikaze drone made the world’s first autonomous kill on human targets in Libya, a UN report revealed.
But experts warn the technology is advancing so fast, governments and societies have not properly considered the dangers.
They say machines making their own decisions are prone to unpredictable and rapidly spreading errors.
These arise from codes called algorithms which even the programmers don’t always understand and cannot stop going awry.
If AI weapons in the future are armed with biological, chemical or even nuclear warheads, the results could be unintentional Armageddon.”It is a world where the sort of unavoidable algorithmic errors that plague even tech giants like Amazon and Google can now lead to the elimination of whole cities,” warns Prof James Dawes of Macalester College.
“The world should not repeat the catastrophic mistakes of the nuclear arms race. It should not sleepwalk into dystopia.”
MIT professor Max Tegmark, co-founder of the Future of Life Institute, issued a similarly dire warning this week.
He told Wired: “The technology is developing much faster than the military-political discussion.
“And we’re heading, by default, to the worst possible outcome.”
A potential ban on so-called Lethal Automomous Weapons Systems (LAWS) was discussed last week at the UN’s five-yearly Convention On Certain Conventional Weapons.
Some of the 120 nations taking part – including Brazil, South Africa and New Zealand – argued LAWS should be restricted by treaty like landmines and some indenciaries.
A growing list of countries including France and Germany support limits on some automous weapons including those that target humans. China said it supports a narrow set of restrictions.
Other nations, including the US, Russia, India, the UK and Australia resist a ban, saying that continuing to develop killer robots is essential to avoid being at a strategic disadvantage.
Life and death decisions
Already a terrifying array of AI weapons have been deployed around the world, including self-firing machine guns in Korea’s Demilitarised Zone no man’s land.
At least 14 countries have suicide AI drones, including Israel’s devastating Harop unmanned attack plane which was used to hunt down Hamas terrorists.
Harops also devastated Armenia’s army during the clash with Azerbaijan last year – although it was not revealed with they acted with or without human input.
Russia’s deadly robot weapons include the new Checkmate stealth fighter, which combines AI systems with a human pilot.
Designers say there could be an future version with no need for a pilot.
China began tests over a decade ago on a robot submarine designed to track and destroy enemy ships autonomously.
China is also building a robot warship armed with torpedoes, satellite pictures revealed in October.
All these killing machines can be programmed to seek human approval before attacking the targets they find.
But they could also be used as the ultimate “fire and forget” technology, loitering for hours or days ready to fire on targets chosen entirely by a computer.
Experts say leading militaries are kidding themselves if they believe they can control the spread of these advanced new weapons.
They say rogue states and terrorists will inevitably try to get hold of them to unleash massacres.
“If you can buy slaughterbots for the same price as an AK-47, that’s much preferable for drug cartels, because you’re not going to get caught when you kill someone,” Prof Tegmark told The Sun.
“Even if a judge has lots of bodyguards, you can fly in through a bedroom window while they’re sleeping and kill them.”
And governments will be more likely to go to war in future, experts say, as robot weapons mean fewer soldiers in the firing line.
They also offer the false promise of clean strikes that don’t harm civilians – with zero accountability for the inevitable war crimes they will commit.
The threat from China and Russia has convinced US military planners of the importance of not being left behind in the 21st century arms race.
The Defense Advanced Research Projects Agency has begun trials involving large numbers of drones and ground vehicles that work together in ways that operators struggle to control.
The US Air Force is also investigating how AI could help or even replace human fighter pilots in a new military landscape ruled by machines.
In April, a Pentagon official confirmed they are considering whether it will one day be necessary to remove humans from the chain of command in situations where they cannot respond fast enough against robot enemies.
The superpower arms race shows it may already be too late to stop the rise of the machines.
Putin’s fearsome robot army goes into action amid fears of war in Ukraine