io9 has an interesting futurist piece about the upcoming workpocalypse that many fringe thinkers are kind of hoping will happen because it would be nice to be right for once.
The key argument is that we will hit a point where robots and computer algorithms are so advanced that they’ll start replacing the legions of middle-aged workers without a useful skill set. Unemployment will skyrocket, riots will break out, and then the robots will take over and we’ll all be plugged into the Matrix. OK, we made up that last part.
Still, sorry, guys. The future you’re worried about has already happened, and been found lacking. Oh, robots and computerization will become more and more integral to our daily lives, and there will undeniably be some job loss. For example, the military is going to be very different in fifty years.
But robots are a poor replacement for humans, and there are other forces at work a lot more powerful than technological innovation.
#6) Computers Aren’t As Flexible as Humans
This piece of doggerel pretty much defines the limits of complex machinery:
I really hate this machine
I wish that they would sell it
It never does quite what I want
but only what I tell it
The inherent problem with computers is and always has been their literal nature. And it gets worse with a robot: a robot has to assess its environment, make a plan, and implement that plan; it has to think very hard about what comes naturally to us. And it does it literally.
If you think this isn’t a problem, try writing out all the things you do, in order, for a task you handle every day. Even personal grooming involves a lot more steps than you think, and a robot has to do those steps in order. This is why robots are great at complex manufacturing tasks, and lousy whenever you ask them to step outside that narrow range of tasks; they’re just too literal to handle anything else.
#5) Robots Have Been Found Lacking In The Workforce
We’re not talking manufacturing here, but other jobs robots were supposed to throw everyone out of work a long time ago, like cashiering.
Go into any chain grocery store, and you’ll notice something: they have “self-checkout” lanes, but they also still have a long row of cashiers. If you watch closely, you’ll see why: when a human cashier has a problem, they punch in a code into the register, or otherwise can generally solve the problem on their own quickly and efficiently with minimal training. Very rarely will they call the manager. If you confront the checkout robot with something it doesn’t understand, and these things have a lot of trouble with anything that isn’t a bar code, it has to call the manager. In fact, you’ll often see an employee loitering near these machines specifically to fix them when they freak out. Some supermarkets are even just yanking them completely, finding them not worth the cost and effort.
An even better example are Interactive Voice Response systems: yes, the dreaded customer service phone robot. Companies quickly found that they had to program voice-controlled systems to send you to a customer representative whenever they heard profanity. Go ahead, try it: bring up an IVR system, and just calmly say a swear word. Odds are better than even it’ll go straight to a human.
#4) Robots Cost More than People, and Always Will
Economics has a much stronger role to play here than most futurists like to think.
Take your typical highly skilled construction worker, a welder or a master carpenter, supposedly a job under threat by robots. You can pay up front for a hugely expensive robot, who will need maintenance, spare parts, fluids, storage, and, oh yeah, will be obsolete in about two years, or you can pay some guy in a union fifteen bucks or so an hour plus benefits, and his health, feeding himself, and housing is someone else’s problem.
This isn’t to say technology won’t give us new tools: the welder of the future will have a lot less grunt work to do, as he can focus on the hard stuff while the robot handles the simple welds. But he’s still going to be welding.