Smart guns? No thanks.

I recently read an opinion article on CNN.com by Jeremy Shane, and I had a few thoughts to share.  Shane basically says that it would be great if guns were smart enough to refuse to fire when they shouldn’t be fired (like at a group of children).

One of the things I’ve been training my brain to do, as part of my job as a software developer, is to figure out all the ways the code I write is broken or could be misused.  Is it possible for certain input to crash the system or cause other problems?  Can a malicious user use this to get information or access they should not have?

One of the other things I’ve been training my brain to do is to figure out if the code I’m writing is going to prevent users from doing something they should be able to do.  If I add this constraint, will I prevent users from using the service the way they’re used to using it?  Will I break something that used to work?  (This concept is called “backward compatibility”.)

This is what Jeremy Shane gets wrong.  He forgot to ask himself two questions: “What could possibly go wrong?” and, “Is this backward-compatible?”

For example, he wrote:

The root of the problem is that guns are “dumb.” Pull the trigger and they discharge bullets mindlessly, regardless of who is doing the aiming or where they are aimed. Guns should “know” not to fire in schools, churches, hospitals or malls. They should sense when they are being aimed at a child, or at a person when no other guns are nearby.

At face value, I agree that guns being “dumb” could be viewed as a problem.  There are plenty of situations in which it would be good to have a “smarter” gun.

But the purpose of a gun is to shoot things.  The last thing you want is a gun that can arbitrarily decide not to fire when you need it most.

The problem is that what Shane proposes is not backward-compatible.   If I carry a “smart” gun for self-defense, and I am in a mall when a guy comes in to shoot up the place carrying a “dumb” gun, my “smart” gun will refuse to fire, resulting in everyone present being killed to death with bullets.

Put another way, Shane’s idea only works if everyone turns in all their guns (including the criminals!) in exchange for “smart” guns.  I’ll let you guess what would happen if you tried to force people to do that.

Now, Shane tries to partially address the “what could possibly go wrong” question:

Building software into guns need not affect gun owners’ desire to protect their homes. Trigger control software could be relaxed when the gun is at home or in a car, while other safety features stay on to prevent accidental discharges. Guns used by the police would be exempt from such controls.

The problem is that this isn’t really a solution.  What happens if my “smart” gun can’t figure out whether it’s in my house or out on the street in front of my house?

What happens if I pull my gun on a mugger, but it has run out of batteries and so refuses to fire?

What happens when criminals simply start stealing the guns meant for law enforcement, just like they used to steal higher-capacity magazines meant for law enforcement when those were banned?

What happens when (not “if”) someone writes a hack that either disables the safety control software, or disables other people’s guns entirely?

If you read Shane’s article, you’re about to point out that he proposes a solution:

Couldn’t gun software be hacked? Perhaps, but the risk can be reduced by open-sourcing code, requiring software patch downloads, and notifying gun makers or law enforcement if software is disabled.

People don’t even keep their desktop computers up to date.  What makes Shane think that gun owners will keep their guns’ built-in computers up to date?

Shane does propose phasing in “smart” guns slowly… starting with what he calls “the most lethal assault rifles”.  The trouble is, that wouldn’t be the right place to start.

You see, yes, carbines (which I’m sure Shane would describe as “assault rifles”) are one of the most commonly sold guns in the country, but the vast majority of gun homicides are committed with handguns — and when I say “vast majority” I mean somewhere around 90%.

What I’m getting at is that I can’t take someone seriously when they claim they’re trying to stop gun homicide but they want to start with rifles.  If guns are the problem — and I don’t think they are — then the problem is handguns, not rifles.

I do like that Shane is trying to think of solutions to gun violence that do not involve unenforceable and counterproductive legislation.  But as with all solutions, we need to make sure we don’t rush into something without figuring out whether it will work.

Sorry, Shane, but your idea would only work in a world where there are no criminals in the first place.