Jack McDonald

On Killer Robots and IEDs

Lethal autonomous weapon systems are back in the news because a lot of tech company founders have penned a letter calling upon states to “find a way to protect us all from these dangers” of lethal autonomous weapons systems. I find it amusing that this letter is signed by Elon Musk and SpaceX, since that company is now integrating military satellite deployment into its business model. That aside, the letter also contains this chunk of text:

Lethal autonomous weapons threaten to become the third revolution in warfare. Once developed, they will permit armed conflict to be fought at a scale greater than ever, and at timescales faster than humans can comprehend. These can be weapons of terror, weapons that despots and terrorists use against innocent populations, and weapons hacked to behave in undesirable ways.

This roughly breaks down into four objections:

Reading the reporting of this letter, it seems many journalists take these premises as a given, or at least don’t subject them to critical thought or consider the implicit comparisons being made. Here, I’ll deal with that first objection, because it’s the one that irked me the most.

A “third revolution” in warfare?

The idea of military revolutions has its basis in “the military revolution” of 1560–1660, referring to the way early modern armies arranged and organised themselves. There is an extensive literature of historians arguing about the nature of “military revolutions”, whether it happened earlier, or later, and the underlying causes for it. Whatever position you take on this, it seems clear that Europeans organised to fight each other in a different way over this period, and subsequent to it.

Which leads nicely to the twin problems of subsequent revolutions, and Euro-centrism. LAWS as a “third revolution” presumes that there is agreement over the number and types of military revolutions. To be blunt: there isn’t. For example, in War and Technology Martin van Creveld divvies up military history into four phases of tools, machines, systems, and automation. William S. Lind and the Fourth-generation of Warfare (or 4GW on a thousand Pentagon slide-decks) phases warfare by tactics (line/column, fire and movement, infiltration, dispersion). Andrew Krepinevitch said there might be as many as ten. You get the drift.

What most theorists of military technology and military revolutions share is the belief that military change is social, as well as technical. Yes, the invention of explosive weapons changed things, but military organisations shape the adoption of weapons systems. Taking as given that LAWS will create a revolution ignores these underlying dynamics, and without a bit more explanation is pretty much futurism gone wild.

The problem with LAWS as a military revolution is that it ignores the likely countermeasures that people will use to defeat that capability. Nowhere is this more apparent than in the continued persistence and utility of improvised explosive devices. IEDs are one decades-long “fuck you” to uni-directional theories of technological change. Expecting states to develop LAWS without expecting their opponents to defeat them in the way that IEDs have defeated tanks and armoured vehicles for decades defies logic.

This leads to my second point — Euro-centrism — since (at least in my reading) Euro-centric military revolutions theory tends to ignore how other people fought each other, or how Europeans (or their Limited Companies) waged war in the wider world. The line/column thing was great for giving Europeans in Europe a kicking, but worse than useless for European settlers/colonists (and their descendents) fighting native Americans in the new world.

How big states imagine deploying LAWS (or like systems) to fight their opponents is unlikely to reflect the global use of autonomous systems in conflict. When thought turns upon the idea of revolution, or similar evolutionary models like a punctuated equilibrium, it drowns out the incremental day-by-day change of the wider world. Big militaries are focused upon human augmentations and integration rather than independent killer robots (i.e. Terminator-like platforms), and non-state armed groups will happily bootstrap almost any civilian item into a weapons-platform if it works. The future is contingent upon circumstances and local utility, not a predictable outcome of an undefinable class of weapons systems.

As it stands, LAWS will always be embedded in military organisations, and those organisations, and the societies that deploy them, will always be susceptible somewhere, somehow, to violence, a point RUSI’s Elizabeth Quintana makes buried at the bottom of this BBC report. In my mind, the interesting issue with LAWS is that they’ll increase the vulnerability of big-ticket items (like aircraft carriers, F-35s, and tanks), and target-identification problems will limit their use against individuals. As LAWS will no doubt be vulnerable to LAWS, this would, in effect, increase the vulnerability of everything except maybe infantry. I haven’t read anything raising this possibility, but the irony of LAWS might be that the things anti-war campaigners are so worried about end up blunting the ability of states to wage war, full stop. At least, that’s as likely as a robot-hellscape of popular imagination. Either way, journalists should maybe push-back on the idea of a revolution in robot warfare.

Mirrored at Medium.com