When you imagine the upcoming Age of Robots, you probably see the robots replacing humans in jobs that involve manual labor. An assembly line is a good application for robots, for example. And I assume fast food workers will soon be replaced by robots too.
But I predict that one of the first occupations that will be entirely replaced by robots will be middle management, not skilled labor. I think it will be a long time before a robot can replace a sales person or a graphic designer. But it won't be long before a computer can do project management and resource allocation better than humans. Compared to most skilled jobs, management is relatively easy. Management only becomes difficult when there are so many simple projects happening at the same time that a human can't keep them all straight. The individual tasks of management are fairly simple. Management only becomes hard when you add a lot of simple steps together until you have a complicated whole.
Computers are great for handling that sort of complexity. Put a computer in a robot body and it can walk from cubicle to cubicle handing out assignments, checking on progress, and adjusting schedules and budgets on the fly. A robot could easily juggle the complexity of dozens of projects. It could be talking to you in your cubicle while simultaneously having a phone call with another employee and texting a third without you even knowing as it happens.
I don't think it will be hard to teach computers the basics of project management. Most of the common steps for a project have a predictable order. For example, you know you need to get bids before buying hardware. You know you need to prep the space before installing the equipment. And you know you don't put the equipment into production until after it has been tested. A robot can easily learn all of the steps in a common project.
Best of all, a robot would be good at estimating the time and resources needed to complete projects. Robots that are involved in project management would share their experiences through the cloud. Eventually Big Data will help the robots determine how long the various stages of the project should take, and the resources that are needed, based on similar projects elsewhere. The robots will be free of human bias and optimism, so I would expect them to do a better job of estimating budgets and timelines than humans. A human manager will tell his boss what the boss wants to hear. A robot will be entirely objective, creating estimates based on similar projects from history. The robot won't fear being fired if he tells the boss the project won't be done before the CEO visits. For the robot, facts are facts.
One of the biggest advantages of a robot manager is that it can be a hard-ass jerk as often as that is called for. A robot might need to single out weak performers and let the rest of the team know who the problem is so peer pressure does its thing. A human couldn't get away with being so confrontational, but a robot has no feelings. It simply identifies inefficient parts of a system and highlights them. No one would bother wasting an hour of the robot's day crying in its office or complaining about fairness.
My prediction that robots will dominate management before they dominate blue collar jobs is based on The Dilbert Principle which observes that the least skilled employees are promoted to management. You need your most skilled people doing interface design, engineering, and the hard stuff. Management is mostly about optimizing resource allocation, and that is something a robot can learn relatively easily, at least compared to most skilled jobs.
You might wonder if a robot can have enough leadership qualities to be a manager. I would point out that most humans in management have zero leadership skills, so the bar isn't set high. You can see leadership in humans when you start getting to the senior management level. It might take a hundred years for robots to get C-level jobs. But I think robots will dominate middle management in twenty years.