22 May Bad Bots: Battlefield Robots and Counters to Them as Weapons Against the US
(Last post on battlefield robots.)
An issue of battlefield robots that has not received, so far as I can tell, sufficient attention is what happens when battlefield robots are deployed against the US.
The assumption seems to be that robots are going to be so complicated and expensive that only the US, and maybe a handful of others, such as Japan, will be able to deploy and maintain them. What happens if that is wishful thinking? What happens instead if it turns out that once the investment in R&D is made, they turn out to be just more cheap, reasonably easily replicated consumer electronics – as the Apple slogan goes, designed in California, made in China? What then?
The most scary but also perhaps most realistic scenario for robots against the US is that the hard part, the really hard part, is coming up with autonomous battlefield robots that are able to discriminate and make proportionality calculations. But that’s something the US cares about; not all our adversaries necessarily do, and might be very willing to deploy robots that are independent but lack these elements of judgment that we are still working on programming.
An adversary might be willing to deploy robots that, for example, are capable only of identifying a source of fire, and perhaps to distinguish friend from foe, and fire at it. But is not capable of determining civilians nearby, and so on. The technology to independently shoot at things is deployed by an adversary while the US is still trying to develop the judgment technology and programming.
Somewhat peculiarly, the US might find itself facing robotic technology with independent firing capability in the field, as a weapon against the US, before the US is ready to deploy its own robots. These judgment-incomplete robots, as we might call them, might legally be characterized to be mobile landmines in the sense of weapons that are inherently indiscriminate because, like antipersonnel mines, they lack the ability to distinguish combatant from noncombatant. But they might well be on the battlefield, and the US might need to develop counters to those weapons even before the US has deployed autonomous battlefield robots of its own. What are the counters to these weapons?
In a similar vein, how might nonstate actors and terrorist groups use independent robotic technology? We think of little robots able to crawl into an apartment and blow up just the bad guys with a mini-blast as a really great thing – a big advance in discrimination in targeting – and it is. But what happens when little crawly creatures with mini explosives start independently crawling across a street and into a crowded restaurant in Baghdad or Tel Aviv or Washington DC or London?
What are the counters to uses of independent, but not ethically constrained, robotic technology either against the US in war or as terrorist devices?
Because the crazy part is, the US might very well find itself having to counter robotic technology before it has deployed robotic technology in its fullest “autonomous” form.
(If this is starting to resemble a little bit too much the movie adaptation of the Philip K. Dick novel, Screamers, in which self-replicating battlefield robots take over the world, well …)