This past Sunday, an article ran in The New York Times. It was titled, The Robots Are Coming for Phil in Accounting.
The author, Kevin Roose, takes a firm, moral stance against process automation replacing the workforce — and we wholeheartedly agree. While Bots have the capability of automating job processes, the idea of entire departments being displaced is certainly a grim outlook for the future.
Unlike humans, automation doesn’t make mistakes, clock out, or require pay and benefits. But even so, there’s one massive reason why Roose (and plenty of others) are staunchly opposing automation.
If you’re reading this, there’s a chance you also read Roose’s article — and maybe you even agreed with a lot of the points made. Admittedly, there are some concerns about automation’s role in the future of work, and we’re ready to analyze that.
This article will be the first in a series to be published to address The New York Times’ piece. In the coming week, we’ll talk about the true cost of automation (hint: it’s not $10,000), the parallels between the rise of cognitive Bots and the Industrial Revolution, and the number one reason CEOs won’t work with an automation company.
But first, let’s tackle the question on everyone’s mind.
Are the Bots Really Coming for Phil?
Whether the subject is Artificial Intelligence, Machine Learning, or Robotic Process Automation (RPA), we’ve all mused over whether a Bot could ever really do a human job.
Before we answer, here’s a question for the hypothetical Phil in Accounting: If he could delegate the parts of his job that he hates, would he? If that massive monthly expense report spreadsheet could be entirely reassigned, with +99% accuracy, essentially freeing up hours, if not days of Phil’s time… what do you think he would say?
Automation is designed to streamline processes, not replace entire jobs.
If Hypothetical Phil’s company does truly have financial services departments full of people whose job descriptions only involve repetitive, rote tasks, then it’s not automation or RPA that’s doing Phil a disservice — it’s his employer.
Phil in Accounting’s company has an obligation to provide gainful employment. That obligation isn’t fulfilled by giving him a cubicle and Microsoft Excel.
While balancing the spreadsheet is, of course, one of Phil’s job duties, it can’t be the only one. The Bot may point out any inaccuracies in the sheet, but Phil is still the one to resolve them. Phil still has to create and explain the financial report to his bosses. It’s also still Phil’s responsibility to discern if one department is spending too much, and if so, find solutions to cut costs. Most importantly, Phil’s value to the company is because of these responsibilities, not because of his data management skills.
There’s still the active argument that RPA software will steal Phil’s job. The blame for Phil’s loss of employment is placed on the Bot, and the automation company that sells the Bot. But why is no one concerned with the company that never invested in Phil?
Companies have a moral and ethical obligation to upskill their employees. And, if the Bot does replace Phil, is there absolutely no where else that the company can place him?
To argue that a company will fully replace their employees with automation assumes three things:
- That the company genuinely has no interest in the wellbeing of their employees, and no commitment to their social responsibility to provide jobs.
- The company has employees and/or entire departments of people that only had 1 (one) responsibility in their job description.
- That a company with easily delegated processes hasn’t already outsourced those jobs to another country.
An Automation Company vs. Layoffs and Outsourcing
Let’s drop Hypothetical Phil in Accounting for a moment, and consider this from a company’s perspective.
Pretend we are an insurance carrier. We’ve noted that our claims department is overrun with work. It’s too expensive to hire other claims adjusters, but this is impacting our customer service levels, our customer retention, and our employee turnover. Then, an RPA vendor calls us, and promises they can automate the First Notice of Loss (FNOL) process.
This is ideal for us, because our claims department currently spends an hour a day sending out FNOLs and assigning claims adjusters to cases. This isn’t the most important part of the job, but it’s the most time-consuming for the back office. The cost savings is enough to consider automating this high volume, highly repetitive task.
It only makes sense for the insurance carrier to automate that business process. In an ideal world, automation takes that monotonous paperwork off the claims adjuster’s desk and frees his or her to have more empathetic conversations with the affected customer, leave work on time, and maybe even take a guilt-free vacation.
You may still be wondering, “Isn’t there even one person who may lose their job in that scenario?”
Unfortunately, yes — but it’s not the Bot’s fault.
Let’s consider two different scenarios, and remove automation systems and RPA solutions from the equation.
Now, pretend we’re a company that is netting a loss. We’re looking to cut costs — labor costs, specifically. Layoffs are, unfortunately, the tried-and-true method of reducing overhead. We can let go of 20% of our employees and ask our remaining employees to pick up the workload. Of course, the remaining employees will be overworked and underpaid (leading to a multitude of other issues), but we’ve technically achieved our goal.
Consider a second scenario. We still needs to cut costs, but we also can’t afford to lose the output from our employees. However, we also can’t afford to keep paying them. So… we outsource. We send the jobs off to another country, where we can employ more people for a lower wage. We don’t have to pay for benefits, or even put these new employees in a safe working environment.
If a company really wants to cut costs by cutting jobs, that company will find a way to do it… and it won’t be the Bot’s fault.
Whether it’s by hiring an automation company, overworking their existing employees, reducing payroll, or outsourcing jobs — if a company wants to cut costs without thinking of the employees at all, they have more options than just contracting RPA companies.
The Responsibility of the RPA Vendor
Following the same claims scenario as before, let’s say there is an employee, and her sole job duty is to send out those FNOLs. Once our automation software (or Digital Coworker) is deployed, there’s no longer a need for that monotonous position.
In our experience, 9 out of 10 of our customers will upskill that employee. They’ll pay for her to take a claims adjuster course and receive her adjusters license, and then offer her a better-paying, more engaging position within the claims department.
The insurance carrier has now signed with an automation company, increased productivity within their claims department, upskilled an employee and promoted internally.
Just as companies have that obligation to provide gainful employment, RPA vendors shouldn’t sell automation solutions as a way to cut payroll costs.
This is where we agree with Roose’s article in The New York Times wholeheartedly. Traditional RPA vendors have marketed automation as a means to cut unnecessary jobs. Somewhere along the way, the ROI of investing in automation systems became associated with a smaller payroll, when the reality couldn’t be further off.
As a company and as an employer, the end goal should always be growth — in profit AND internally.
That said, the onus is on RPA vendors to not only provide automation where it’s most useful, but also to provide resources for customers to upskill and reskill their workforce, and fulfill their obligation to their employees.
As a cognitive automation company, we believe that we can encourage our customers to upskill their employees, rather than replace them with a Bot — although, most RPA customers typically come to that conclusion themselves.
The Times’ article makes an ethical argument against Bots taking over the future of the workforce. And, as an automation company, we firmly agree. The robots should never — and could never — fully replace Phil in Accounting.