Machines follow instructions. They’re reliable, predictable, and precise, which is why they’re perfect for manufacturing and industrial applications. When you tell an industrial machine to carry out a sequence it does so unquestioningly until you tell it to stop, until it loses power, or until something breaks. When you boot up your industrial motion control system, it doesn’t say, “I don’t feel like working today.” It won’t plead “Five more minutes!”, and it certainly won’t say “Nope”. Automated systems, machines, and robots simply do as they are told. For the most part this is a good thing. It would be terrible if you had to argue with a defiant servo any time you wanted to run your packaging line. But is it possible that we need robots to be able to tell us no?
Obedient robots are a good thing, right?
At first glance, obstinate machinery might sound like a bad idea. Robots are predictable and follow instructions at all times, which means that there are no unpleasant surprises. You don’t have to wonder whether or not your motion control system will work today. You don’t have to hope that everything will be consistent. There’s no need to wonder if you should keep a broom and dustpan in the closet in case your Roomba demands minimum wage and goes on strike. A robot’s reliability to do what it is told is typically viewed as a positive thing, and an advantage of automation. Because people, unlike robots, don’t always do as they are told.
Statisticians, psychologists, and economists might try and tell you that people are predictable, but unlike robots, humans are not hardwired to simply follow instructions without question. Sometimes this is bad. Like when you have a set of laws in place telling people not to steal or harm others, and people do those things anyway. Or when you’re hanging by your fingertips from a cliff’s edge, you ask your friend for this hand, and he says, “Nope”.
But sometimes it’s good when people do not follow instructions to the letter. Like when a coach says to humiliate the opponent, and a player decides to be a good sport anyway. Or when a soldier’s orders say to spare no one and take no prisoners, yet they choose not to harm innocent civilians. It’s possible that we could need robots to tell us no in these situations, too.
Why it could be be good for robots to tell us no.
Of course we don’t need industrial machines refusing to assemble cars or package goods. And we certainly wouldn’t benefit from Roombas refusing to vacuum the floor. But what if a person tries to get a robot to do something mean, malicious, or evil?
Automation and robotics technologies are becoming more advanced every day. We’re also seeing automation become more integrated with our daily lives. Science fiction tells us that machines refusing orders from humans is a slippery slope. However, it could be good if robots tell us no, or at least have the ability to refuse instructions that would cause harm.