China’s version of RoboCop is a little scary, but not for the reason you’re thinking

Promotional still from the 2014 film Robocop.

You may already have seen by now: China has unveiled its own version of RoboCop. The squat, egg-shaped “AnBot” was developed at China’s National University of Defense Technology. Sadly, it doesn’t really look like RoboCop. It looks like this:

It does boast some impressive features, though, like:

• A top speed of 18 kmph
• Remote-control Taser-style electric weapon for non-lethal “riot control”
• Self-guided patrolling
• Panic button that will alert police
• Smart monitoring that will alert police if someone shouts “help”
• Automatic recharging
• “Optional modules for environmental monitoring, biochemical detection, and clearing explosives”

Needless to say, this announcement was met with quite a bit of mockery both at home and abroad. The folks at Shanghaiist, for example, pointed out that it wouldn’t be much of a threat to criminals or rioters who have mastered the use of stairs. And here at Tech in Asia, we managed to obtain some exclusive footage of the AnBot helping a civilian in danger:

Jokes aside, though, this thing does kind of creep me out, and not for the reason you might expect.

When people talk about police robots, the concern is usually that they’ll be used to create some kind of dystopian nightmare – think the robots from Elysium, if you’re one of the other three people in the world who saw that movie.

But my concern for the moment is actually the opposite. This kind of technology will tempt lazy administrators into replacing actual human cops with shitty robots like this in low-crime areas to save money. And then, if and when something actually does happen, instead of having a trained, thinking human officer there to handle the emergency, you’ve got an egg-shaped version of Siri that can’t climb stairs.

We need to be very careful about the responsibilities we give to robots until we’re sure they’re ready.

Obviously, having this sort of thing is better than nothing, and if it’s only used to augment existing police forces, then I’ve got no real problem with it. I think that’s unlikely, though. There are always middle-management types looking to cut corners with the budget and willing to make a few bets that nothing will go wrong in the process. But if and when something does go wrong, it’ll be some innocent citizen who’s yelling “Help!” at a stupid robot that can’t actually help them as they lie bleeding in the street.

Automation of virtually every job is coming, and security and police jobs are no exception. But especially in those jobs, I think we need to be very careful about the responsibilities we give to robots until we’re sure they’re really ready to take over. Because while a shitty robot waiter malfunctioning or just being poorly-designed can’t screw up much more than your order, a robotic security officer that screws up or simply can’t handle the kind of emergency it’s being presented with could be the difference between life and death.