‘Smart’ guns and the dangers of trigger-happy technology
Sniping skills are being digitised with 'precision guided' weapons
A few weeks ago, some old Silicon Valley hands suggested I take a look at a weapons company called TrackingPoint. I duly perused its website, not knowing what to expect, and quickly found myself both amazed and chilled.
This tiny tech business based in Texas sells “precision-guided” weapons that “enable anyone to accurately hit targets at ranges from 100 yards up to one mile in some cases”. More specifically, these guns are fitted with devices that use laser beams, sensors and computers to track targets and hit them in almost any conditions, irrespective of shooting skill.
They are so effective that there are videos circulating online that depict a blind military veteran hunting with a gun, and hitting the target; company officials say he was guided by a friend with a computing device.
Until recently, it was assumed that the job of a sniper or a hunter was a human one, requiring fantastic eyesight, great mental and physical discipline and lone-wolf patience. Just think of the images presented in movies such as American Sniper or The Day of the Jackal.
Today, however, we are witnessing the emergence of quasi-robotic snipers equipped with “smart” guns. This does not mean that flesh-and-blood people are redundant: they are still needed to tote the weapons and select the targets. Instead of talking about “artificial intelligence”, it is more accurate to talk about “augmented intelligence”, in the sense that human shooters are enhanced by technology.
Even allowing for that caveat, the TrackingPoint guns take military and recreational shooting to a new level, starting at $6,995 a gun, according to the website, with products aimed at the hunting market as well as military niches.
And TrackingPoint is not alone: although details are elusive (since most material is classified), other US companies are reportedly developing “smart” bullets, which use guidance systems and powerful computers to hit their targets.
Sniping skills, in other words, are being quietly digitised, just like the expertise underlying taxi driving or paralegal work.
Or as TrackingPoint’s co-founder John McHale put it in a 2016 press statement for a new squad-level M1400 338LM bolt-action rifle: “Extreme distance lethality is no longer the exclusive domain of trained snipers. With minimal training, any soldier can reliably deliver lethality well beyond what is possible for today’s expert marksmen.”
The statement adds: “As a soldier … pulls the trigger, the target is automatically acquired and tracked … Total Time-To-Kill (TTK) is approximately 2.5 seconds.”
Is this a good idea? When I first saw the website, I wanted to howl “no”, and not just because I find the TTK idea distasteful. After all, the US is a country that is already blighted by an excess of guns and gun violence. And one obvious problem with combining computers and weaponry is that software can malfunction. Just look at the current furore around the Boeing 737 Max 8.
There is another obvious risk: these guns could fall into malicious hands. TrackingPoint insists in its publicity material that it will only supply the guns to “the US military, other US organisations that can legally fight our adversaries, and qualified US citizens”, and suggests the main purpose is to “deliver mission dominance, force multiplication and remarkable battle overmatch in the war on radical Islamic terrorism”.
Maybe so. But what if someone were tempted to sell this know-how elsewhere? (Which is not impossible to imagine, since TrackingPoint was forced to restructure a few years ago, owing to financial pressure.) What if terrorists stole the technology, or hacked into the software?
“That is the real worry,” observes one veteran Silicon Valley technologist, who says that these days the military is spending almost as much time developing systems to jam robotic weapons as it is on developing the weapons themselves.
Of course, these problems are hardly new: proliferation existed long before robots arrived. And it is not clear that robotic guns are more dangerous than the nonrobotic sort. As with driverless cars, it is probable that automation will produce fewer casualties These weapons are so accurate that there is less danger of hitting the wrong target.
But such arguments will be of little comfort to many. Nor will it be reassuring to hear that the company was recently folded into a bigger weapons group (including Joint Force Enterprises and Talon Precision Optics) based in Jacksonville, Florida. Company officials tell me they hope this deal will turbocharge sales, enabling them to supply more smart guns to the US military.
I suspect the only reason why there has been so little public debate about these guns is that it is not widely known such hardware exists outside Hollywood films. So this is a new twist on the idea of automation. But remember that this is just one tiny — visible — tip of a vast iceberg of innovation, most of which we cannot see because it is classified.
That is truly disquieting; and it puts a whole new spin on the idea of digital disruption.