A footnote to my last post about autonomous killer drones; they're not just worrying because they're (possibly) already here. There's also a compelling military logic to keep developing such things:
...when you talk about drone systems, about remotely-piloted systems, these systems are comparatively easy to detect. I mean, it's not that easy if you actually want to build a counter-drone system, but it's still comparatively easy, and that's because of the uplink and the downlink ... the control links that go to the drone and the information links that come back to the pilot. So if you want to fly an unmanned system in a way that's as stealth[y] as possible, you want to cut these links, which means you need to give the system more autonomy.
Dr Ulrike Franke, of the European Council on Foreign Relations, (about 2'30" into this video).
Communication links are a vulnerablity, and the technology exists to get rid of that vulnerability. So the incentive to make killbots autonomous exists.
The only upside, for those of us who don't like the idea of swarms of killing machines perpetrating algorithmic carnage, is that swarms of drones which communicate and coordinate with one another retain this weakness. Maybe the take out is that it's easier to make killbots as solitary assassins than terrifying swarms, at least if the defenders have the capability to detect the swarm's chatter and use countermeasures.
History suggests that, all other things being equal, these sorts of technological constraints, rather than moral debates, tend to guide how militaries use novel ways of waging war. An interesting book review by David Fedman and Cary Karacas highlights this. They're dissecting Malcolm Gladwell's The Bomber Mafia,* a book which suggests that the savage fire bombing of Japanese cities in World War Two was the result of a doctrinal battle being won by brutal area bombing advocate Curtis LeMay over the more restrained advocate of precision bombing, Haywood Hansell.
Fedman and Karacas aren't having this over-simplification of history, though. In reality, they write, the dividing line between area and precision bombing had already become blurred over Europe and the path chosen owed more to pragmatism and contingency than to some moral and doctrinal face-off between two Great Men™:
It’s important to recognize that, by the time the COA [Committee of Operations Analysts] issued this report, the AAF [Army Air Force] was already engaged in the radar (or “blind”) bombing of German cities. While Gladwell devotes considerable ink to a frustrated Hansell contending with the challenges of the air war in Europe, he says little about the broader evolution of American bombing tactics. Hobbled by poor weather, the Eighth Air Force had come to accept that precision bombing was not achieving results. The use of radar, by contrast, appeared more efficacious, even if it meant accepting that bombs would fall haphazardly across urban areas. Under the pretext of destroying Germany’s railways, moreover, they had begun to bomb large swathes of entire cities.
After a character from my favourite sci fi show in many years, The Expanse, kills the mad scientist who's been unleashing all sorts of horrors on the human race, he says "I didn't kill him because he was crazy. I killed him because he was making sense."
If you're building a drone to evade detection and countermeasures it makes sense to make it autonomous, which is a pretty scary perverse incentive.
*Disclosure - I haven't read it and, after that review, I don't feel inclined to.
0 comments:
Post a Comment