Far from being autonomous merchants of death, killer robots are our friends, writes Matthew Sinclair.
It is understandable that many people are uncomfortable with the idea of machines doing our fighting for us. The Terminator and countless other science fiction masterpieces have ingrained in our psyche the fear that one day violent automatons we build will turn on us.
There’s an unease about the drones used today too. It seems almost unfair that an operator of a drone can kill a terrorist in Syria from the comfort and safety of the UK or the US, in the same way early machine guns felt unfair in old colonial wars. Add to that the concern that any new weapons will end up in the wrong hands, and it is easy to see why our lethal machines get a bad press.
But are the killer robots increasingly a reality in modern warfare really bad news?
Dumb weapons will tend to kill more innocent people than smart ones. They are unable to discriminate between the innocent child and the guilty adult. We saw the peak of dumb destructive power in strategic bombing during the Second World War. The smarter our weapons get, the fewer innocent people will die.
Killer robots are also exactly what we need to fight nasty actors like the Islamic State. It is no accident it was a drone that killed Jihadi John.
It is hard for an ageing and comfortable society to beat desperate terrorists. Even if we are able to kill ten or twenty fanatics for every one of our soldiers who dies, and however brave individual British and American soldiers are, the price can become too high for politicians and voters here to bear before the bad guys quit. The other side knows that, increasing the incentive for them to hold out longer in the hope we give in.
At the same time, our tolerance for casualties among civilians caught in the cross-fire is also lower. Our enemies know that too and can exploit it by employing human shields, hiding in places like schools, further increasing civilian casualties in a conflict. If robots are able to do our fighting for us, at a greater and greater level of precision, then fewer people will die and wars will be shorter.
New weapons have always had the potential to change the kind of people who win wars. In his 1903 short story The Land Ironclads, in which he predicted what would become the tank, HG Wells depicted how that new technology could allow “devitalized townsmen” to beat tougher tribes. The clerks, factory hands and students might be “poor amateurs at war” but they were better at making tanks.
The new technologies being developed today favour sophisticated societies with market economies over random fanatics with pick-up trucks and scratched together small arms. They favour the good guys.
The contention that the transfer of these weapons to non-state actors is inevitable and terrorists will turn these weapons against us is unconvincing. They have neither the motive (avoiding unnecessary casualties) nor the means (technological sophistication). They might buy, steal or even manufacture some and that is a risk but, unlike with – say – nuclear weapons, it is not a potential catastrophe. We are not talking about some unstoppable science fiction Terminator arriving on the scene, but a weapon we would have more of and be better able to use. Our robots would win.
Many of the weapons being used by Islamic State in Syria today are a century old or more, complemented with simple consumer technology (like pick-up trucks and mobile phones). The Assad Regime has some modern weapons, but it is improvised explosives and poison gas, old technologies, that are creating the real humanitarian disasters. Dumb weapons work fine if your objective is an atrocity. Experts in Silicon Valley wildly overstate the easiness of high-tech manufacturing in the remoter corners of the world where terrorists are forced to hide.
There is much more to fear from a world in which liberal states retreat from conflict, resorting when pushed to the enormous destructive power of old weapons, than from military investment in AI and robotics. Killer robots are our friends.