When it comes to unmanned airplanes, how much autonomy is too much?
The Washington Post offers one glimpse of the future of UAVs. It talks about an ongoing project to increase the capability of identifying, targeting and firing upon targets — with no humans anywhere in the kill chain. (Scared yet?) The article didn’t identify which organization specifically was behind the push; an exercise took place at Fort Benning, Ga., but no military officials were quoted. Still, it’s not much of a secret that the defense industry has been pushing the Defense Department hard to increase “autonomous strike,” a nice way of saying a computer algorithm will decide whether one person lives or dies.
So how does this play at that five-sided black hole on the Potomac? Not so well, unsurprisingly. Chief of Staff Gen. Norton Schwartz was in our building last week and talked about a need for more autonomy in UAV operations — but I highly doubt he’d be a proponent of autonomous strike. (In never arose as a topic of discussion.) The chief told Air Force times last October that the lethal nature of today’s UAVs requires an officer making a strike decision.
The Air Force, in its official technology look-ahead study, doesn’t foresee autonomous strike either. Then-Chief Scientist Werner Dahm unveiled the Technology Horizons study in September, but he repeatedly said autonomous strike is not part of the future for the Air Force.
That’s something that we have policies [against], for well-founded reasons, that we will not go down that path in the timeline that we’re talking about here,” he also told a rountable interview in September. “And frankly, we lose almost nothing by having the human on the loop to make the strike decision. The real benefits of autonomy are not in that very last moment where the strike decision is made, but in the whole chain of events and the coordination that leads up to that.”