Are robots going to take your job?
Maybe, maybe not.
There are estimates out that state 375 million people worldwide will need to transition to new jobs through the use of all types of what is collectively called automation (although that isn’t accurate when you consider AI in the mix) (Mckinsey Global Institute, 2017). That may be true, but it isn’t the full story behind the physical robots because we still need a lot of people in other jobs. While jobs are lost, other jobs are being created. The robotic workforce is not the only thing impacting our economy.
There are two core types of “robots” invading the workplace and some of these robots have been active for decades. They may even overlap and become indistinguishable from each other at times. Then there is a third type commonly lumped in with the word “robot”, although “robot” is not an accurate description.
1. Physical Robots
These may be what most people think of when they think of robots. They have been used in manufacturing for decades. Modern ones often use a small computer or microcontroller (older ones used a lot of relays and electrical wire). A programmable logic controller (PLC) is often the way to store the code that runs the physical robot. It usually has a series of inputs such as sensors to detect poorly made parts (What are Jidoka and Autonomation?) and outputs to send out electricity to fire responses.
People have feared for decades that physical robots would cause everyone to lose their job. That hasn’t materialized yet, although many jobs were lost, others were created, still, others evolved. That is the general trend of ever major shift impacting the workforce.
Books on Robotic Process Automation (RPA)
2. Robotic Process Automation (RPA) or Software Robots
Computers started showing up commonly in the workforce to help make things easier and faster (*side note: this also caused some jobs to be lost, while others were created). Applications were built on these computers to aid in that goal. The computers advanced, the applications advanced with them.
Organizations had some routine tasks that involved the same keystrokes being typed over and over again on these computers. This is where Robotic Process Automation (RPA) began to evolve. In the early days, we just had macros. These were basically “recordings” of keystrokes. You could push a single button and the computer would repeat the keystrokes. If you had a hundred keystrokes, it could save a lot of time.
At some point, businesses didn’t want to pay to keep upgrading their applications. Others wanted ways to get information from one place to another more efficiently. At the core of this was still a repeating of keystrokes, just keystrokes with some variability. The early versions of these robots were coded by hand and often used in spam creation. It is why we have Captchas on things like login pages or account registration screens for many web pages out there.
Businesses used them too, but they took a long time to build and were expensive. It was often cheaper to build newer and better web applications, making the use often limited to easier to automate areas.
Fast forward in time some and a complex RPA can be built by one person in a matter of weeks or even days using drag and drop or low-coding tools. The cost to implement RPA solutions is dropping dramatically while the spending increases. RPAs are starting to impact our working world more and more.
The first time someone sees a more complex RPA in action is usually something they remember and are impressed by. As someone who has built them (manually without drag and drop tools and with drag and drop tools), they actually aren’t that impressive. I have programmed both physical and software robots. The most complicated part is sometimes getting the logic right for what needs to occur next, not the actual underlying code. RPAs are basically macros with variable inputs with, in most cases, more advanced application integration ability. You can build one with an Excel file and lots of patience if you wanted (Hacking Excel: Creating a Search Box Automation Tool).
Most RPAs in use by companies do a lot more than the one shown in the video above, the one above shows a simple tool interacting with a web browser. Some RPAs rely on a backdoor structure avoiding the web application, others work with desktop applications. They take data in, they make defined decisions based on that data and then put it out somewhere else based on the input data. They can issue keystrokes like a macro, move text and make it look like they are smarter than they actually are. Don’t be fooled; they can generally only handle defined processes and if an unexpected issue happens it can stop the bot from running. (*#3 below is changing that)
3. Artificial Intelligence (AI)
Artificial Intelligence (AI) is often lumped in together with Robotic Process Automation (RPA). It is not the same. AI is in fact very different and much more complicated. You still have to tell RPA what to do and how to do it. The goal with AI is to train it and let it make decisions similar to a human. Robotic Process Automation handles routine and regular tasks. AI is not supposed to be routine and regular; it is meant to deal with those situations in which all variables/issues/options are not known in advance.
AI may look like an advancement to RPA, but they came up from two different areas. They evolved independently, but are now merging in multiple ways. More complicated tasks outside of a routine that RPA cannot handle, for example, AI just might be able to do. As such, many RPA tools are integrating Artificial Intelligence within them.
With the other two items on this list, to keep ourselves employed, humans only had to learn how to do more complicated jobs. AI is likely to change that very soon. AI has not reached its full potential, but the list of more complicated jobs it is able to do is growing. At least, for now, it has issues physically carrying out the tasks and humans still have a place. As AI merges with physical robots, it won’t be long before even that area for humans is unavailable.
To be honest, I am not that impressed with AI… yet. I have seen it do interesting things, but not very often better than most people. I have an Amazon Alexa and a Google Home, but I rarely use them as they don’t seem to offer much benefit to my life without spending a lot more money to upgrade all of the lights in my house so they can carry out a basic command that could have been done decades ago with a remote control. I have questioned more than once how intelligent most current AI actually is.
Many of us already work with robots. They make our job easier or they work silently next to us without us even knowing they are there. It will only become more common and the reality is that no one can say for sure what will happen with our jobs because of them. AI has the most potential to be a huge catalyst for change. As it evolves I expect the nature of our work to evolve with it.
Mckinsey Global Institute. (2017) Jobs lost, jobs gained: Workforce transitions in a time of automation. Retrieved from https://www.mckinsey.com/~/media/mckinsey/featured%20insights/Future%20of%20Organizations/What%20the%20future%20of%20work%20will%20mean%20for%20jobs%20skills%20and%20wages/MGI-Jobs-Lost-Jobs-Gained-Report-December-6-2017.ashx