Michael Blumberg
4 answers

Contrary to popular belief the machines are predicted to take over certain tasks from many different types of jobs, not entire jobs themselves. Also when machines do take over some human activities in an occupation, it does not necessarily spell the end of the jobs in that line of work...it may actually increase human activity. A good example is the deployment of bar-code scanners in retail lowered the cost of point-of-sale, but increased the employment of human cashiers in retail.

A large portion (25%) of human workplace activity is predictable physical work. Predictable physical activities figure prominently in sectors such as manufacturing, food service and retailing; and are the most likely to be automated rapidly. These activities include packaging, equipment maintenance, loading, cutting, welding, etc.

Further out is the automation of activities related to collecting and processing data, which by some estimates, amount for up to 30% of all human activity in a workplace. In industries like insurance and investment banking, almost half of all activity is either collecting or processing data. This includes several types of tasks performed by individuals whose annual pay exceeds $200,000!

The hardest activities to automate are the work of knowledge workers. Activities like complex planning, decision-making, creative collaboration and managing people. Healthcare and education are two large industries, which stand out in terms of being least susceptible to widespread automation.

Understanding the activities that are easiest to automate could provide a unique opportunity for managers to think about how many of their organization's activities could be more efficiently executed by machines, freeing up staff time to focus on the core competencies that no robot can replace.