Can you put an AI robot on trial for killing a human?

I have been attempting (with limited success) to warn all of you about the impending end of civilization at the hands of artificial intelligence and robots for some time now. I don’t hold a grudge if you’re not taking this seriously, and I’ll be sure to say a prayer for all of you from my secret bunker as the Terminator machines from Boston Robotics haul you all away to the work camps. But until that day arrives, there’s still time to take in new information.

Over at the Daily Beast, Libby Torres previews the new documentary from director Maxim Pozdorovkin titled The Truth About Killer Robots. It debuts on HBO tomorrow night and I’ll definitely be checking it out. Keep in mind, this isn’t a movie. It’s a documentary and Pozdorovkin isn’t just speculating about the future. He’s talking about robots who have already killed, such as one robotic worker at a Volkswagen factory in Germany which “malfunctioned” and smashed a 21-year-old auto plant worker against a wall, crushing his chest and killing him. And yes, the robots in that area were able to move and operate independently with some level of AI involved.

The accident at the Volkswagen factory introduces the central question that the documentary, somewhat unsuccessfully, seeks to answer: can robots be guilty of killing humans, however unintentional the death was? And how do we hold them accountable if so? Quite a lot of the doc is spent speaking to experts, both legal and philosophical, about the implications of robot interference in human life—and no clear answer is reached.

The documentary is more engaging when it examines the effects of robots and automated labor on the workforce and world economies. The introduction of automatons into factory jobs has lead to the displacement of hundreds of workers, and has forced surviving workers to work harder, longer, and more intricate jobs.

Before we dig into this conversation, let’s take a look at the trailer for the documentary. It’s short and it may pique your interest.

Since the robots are already running up the body count and they’re only getting smarter every month, this documentary may be dealing with a question that we should be prepared to answer. Can an artificial intelligence system be smart enough to be “guilty” of something?

People have been getting killed in industrial accidents since we’ve had machines. Perhaps a bolt snaps and some scaffolding collapses on someone or an insulation seal breaks down and an arc welder puts 10,000 volts across some poor worker’s chest. It’s an accident. Things happen. You don’t blame the arc welder for it. At most, there might be a lawsuit against the manufacturer for an unsafe design, but that’s about where the liability ends.

But what if the machine is smart? What if it’s looking back at you. What if it can answer questions?

Now we’re faced with a new ethical kettle of fish. Does a robot have rights? Does it get a chance to defend itself at trial? Do we need to provide an AI-enabled robot attorney? We’ll have to figure out if those robots are capable of mistakes or if they have the capacity to demonstrate malice aforethought. (Of course, “aforethought” for a robot might take less than a billionth of a second.)

Because they are machines and we created them, maybe we can skip over all of that messiness and just send it to the scrap heap, eh? But wait… doesn’t that sound an awful lot like slavery at that point? Perhaps this documentary will go some small way toward answering these questions.

But you’re going to want to have those answers ready. Rember the robot who can now do parkour?

He’s faster than you. He’s at least as nimble. (Can YOU do parkour?) And his arms can deliver a crushing bite force greater than a saltwater crocodile. And he’s got a wifi-capable computer onboard which means if the master race AI is already out there hiding in the internet… Maybe we’d best not bother with a robot lawyer for this guy.