Killer robots. Invisible drones: not science fiction any more. Welcome to war.

Giving your soldiers superior technology has always been a significant factor in winning wars, but now it’s the technology itself threatening to do the killing.

drone soldiers in front of statue image

Time is running out for the UN to enforce a ban on invisible weapons and killer robots.

Photo: michael mucci

Autonomous killer robots that can hunt down humans and make their own decisions about who to exterminate are not science fiction or a problem for the future. They are a threat now.

Any doubt about that can be removed by reading US Department of Defence budgets covering the past 12 years and by talking to scientists at the forefront of military-funded research. Having done so, the following scenario becomes frighteningly real.

I can assure you a lot more innocent civilians will be killed by drones if we take humans out of the loop.

Professor Toby Walsh, UNSW

A swarm of autonomous drones cloaked in the latest invisibility technology arrives at a house. It hunts like a pack of wolves, informing itself of movements around the target site with virtual pheromones that mimic the communication system of insect colonies.

New drone designs make the lethal Predator, shown here, seem primitive. Photo: Supplied

The swarm identifies an access point behind the house and sends a rotor drone through an open window. It flies swiftly around furniture and through doorways using a clutter avoidance system, its facial recognition software locks on to a human sitting in an armchair and, in a microsecond, the hovering robot makes its own decision to kill.

As the 2011 US defence research budget makes clear, the military’s “robust robotics” program has been developing techniques that enable robotic agents to achieve effective levels of autonomous reasoning “whether humans are present or not”.

In fact, earlier budgets show the US military has been developing autonomous hunter-killer weapons, robot learning and reasoning technologies, unmanned combat armed rotorcraft and drone swarms for more than a decade.

The 2016 budget outlines a newer strand of research. It shows the military developing “hybrid biological-computational platforms” and new approaches for using physiological and neural signals to make human-machine systems more efficient.

It also warns that “current modelling approaches are heavily reliant on human insight and expertise, but the complexity of these models is growing exponentially and has now, or will soon, exceed the capacity for human comprehension”.

What exactly that means is unclear. The budgets read by Fairfax Media are unclassified, and only top-secret versions would reveal whether autonomous weapons have actually being produced, what they are, and if they’re being stockpiled.

The technology is so menacing that more than 1000 top scientists and technology gurus, including physicist Stephen Hawking and Apple co-founder Steve Wozniak, have signed an open letter calling for a UN ban on autonomous weapons.

One of the signatories is Toby Walsh, professor in artificial intelligence at the University of NSW and CSIRO’s Data61, who spoke at a UN event last year on the need for countries to act on concerns arising from rapid advances in AI weapons technology.

“It is certainly the case that there will be an arms race, if we’re not already in the beginnings of one,” he told Fairfax Media. “It’s very hard to know what the Chinese and the Russians are doing. If I were them I would of course be working furiously on developing autonomous weapons because they’re likely to be facing them in battle.

“That’s what is really worrying about this scenario. There’s only a narrow window of opportunity to get the UN to put a ban of some sort in place.

“I know the current technology, and it would struggle to reach the capabilities of the human mind. I can assure you a lot more innocent civilians will be killed by drones if we take humans out of the loop. Another aspect is that if you have autonomous weapons, they will be falling into the hands of terrorists and rogue nations.”

One military-funded scientist, Boubacar Kanté, a professor in electrical and computer engineering at the University of California, says his team has developed a new cloaking device that can be used to create invisible drones in the near future.

The system uses an ultra-thin Teflon substrate studded with cylinders of ceramic that can bend light waves around objects coated with it. However, Kanté takes no responsibility for the deadly potential of his work.

“I am a scientist, so I research and become excited about scientific problems that can be overcome,” he says. “I don’t necessarily think about the final applications. It is for the whole of society to decide what use to make of the technology.

“It’s very difficult to put boundaries on what should be done. There is a global competition to see who can make the first discovery, and discoveries are very difficult to control. Even if the research is not militarily funded, there are enough organisations with sufficient resources to do it by themselves.”

Despite claiming he doesn’t consider the final applications, Kanté says he is already in contact with US companies interested in making an invisible drone.

Walsh describes Kante’s attitude as unfortunate.

“Most scientists would say we do have a very strong responsibility,” he says. “After the Manhattan Project [in which the nuclear bomb was created] we woke up to the fact that what we do is very important.

“We don’t have to develop something just because we have the ability to do so. We have made conscious decisions about chemical and biological warfare, and we are not creating such weapons even though we easily could.

“There was a successful 1998 UN ban on blinding lasers. The technology existed and in fact two arms companies, one US, one Chinese, were ready to sell them. Now no arms company anywhere in the world will sell you a blinding laser.”

The professor says senior diplomats from many countries agreed with his concerns when he spoke at the UN, but he believes that achieving a ban on autonomous killer robots would probably be harder than for nuclear weapons.

The “if we don’t do it, the enemy will” attitude of military powers will be one obstacle. Former US Army officer and autonomous weapons expert, Sam Wallace put the hawkish side of the argument in a written response to calls for a UN ban.

“Machines excel at making split-second tactical decisions that humans have trouble with,” he says. “This could lead to a significant disaster if we choose not to develop any systems capable of combating these types of threats.

“How would you fight against 50,000 small agile robots invading a village if you only have four or five drones controlled by humans trying to find and shoot them down?”


Henry Sapiecha

Leave a reply

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>