In 1875, Karl Marx described what he saw as a workers’ utopia as requiring contributions "From each according to his ability, to each according to his needs."
A practical failure in achieving this utopia, though, is that there are numerous tasks that no one wants to do, or that are considered beneath some people’s ability, yet remain necessary — e.g., toilet cleaning, or manual data entry.
Regardless of one’s political philosophy, we are seeing more businesses developing effective means of enabling workers to deliver according to their abilities, while better addressing worker, company, and market needs.
Robotic Process Automation (RPA) is well on the way to automating away repetitive and deterministic tasks via software, while technical improvements in computer vision and robotics are making it more effective to automate physical work as well. Processes with outcomes that have distinct, measurable value, but that involve repetitive tasks that require no introspection, no judgement, and no human touch, are perfect candidates. Those are the tasks that RPA efforts are focused on — and there are a lot of them.
But, how many tasks really fit that bill? The real answer is that it doesn't matter, because we'll eventually automate them all. The reality of tasks where humans add no value is that nobody wants to pay human wages to do them. The only reason that people do them in the first place is that there have been no affordable alternatives. Now, with automation looming, we need to be sure that the business roles we are creating are ones where humans have something to add.
Cost avoidance is only part of the picture, though. The other part is increased employee productivity.
If we go back to our utopian framework above, we need to get from everybody according to their ability. This is no longer an automation problem, this is a two-fold knowledge problem.
The first problem, and the less important one, is that employees don't all have the skills they need. I say this is less important because it is solvable. Businesses need to invest in training for the skills they need, to do the human tasks that have value. By internalizing the processes that build necessary skills, businesses can give to each according to their need.
The second is much less tractable. Companies too often do not know what specific knowledge or talents that their employees have. As the saying goes, “if my company knew what my company knows...”. The value in employees may be great, but when talented employees are preoccupied by tasks with limited value, it has been impossible to show this.
Marx's failing was to assume that the second category would emerge as self-evident. It has not. Mired in the flow of tasks needed for the daily operation of business, the potential of a company's workers is difficult to tap. In a world with rampant technological change, the distribution of available skills often does not match what the business needs.
Automation is beginning to lift rote tasks away and, to the amazement of many, this is not leading to widespread layoffs or the Robot Apocalypse. And employees increasingly embrace rather than reject automation, because they see benefits for themselves. For example: few, if any, workers today aspire to manual data entry, and now workers can avoid that task and instead focus their efforts where humans do add value.
Even in these early days, though, we can see that companies are going to need to solve both of the above problems before long. To start, companies have a responsibility to themselves, their shareholders, and their employees to ensure that they can build the talent they need. The continual shortage of high-skill jobs is not going to be filled by robots any time soon. Training programs need to be redesigned for a post-automation age, and they need to focus on building the skills that matter, not the tasks that don't. Knowledge workers, like laborers before them will need to learn to operate the machinery (software), rather than doing the work themselves.
Secondly, companies need to invest in building a knowledge model to supplement their commercial and business models. This needs to be more than a hasty volunteer organization, and needs to go beyond Q&A, Enterprise Social Networks and wikis that they have built before. Self reporting, skills evaluations, and recommendations need to be used to build a model of the true talents, extracurriculars, interests, and expertise of the employees within a business. By building this knowledge model, the business can overlay it on top of operational and commercial models to bring the skills from the right people to the right places.
I am highly optimistic that we already have the technology and the abilities to complete the required steps. What remains to be seen is whether companies have the will to pick up their end of the bargain, and help their workers make transitions to a more automated workplace. We may not be able to build a utopia where we only do the work we all want to do, but we can get closer.