Today, you may have seen countless articles about the inevitable apocalypse of automation. Some of these headlines read: “How x job will be disrupted by AI,” or a more popular one: “x billion jobs to disappear by the year y.” All of these headlines read as doom and gloom and most will at least reference to some kind of research or quote by a professional in the field. However, to truly understand where any of these topics are coming from, and whether they are just being over the top, one needs to understand the fundamental idea of automation.
Automation is not an AI, it’s not a computer, it’s simply a term used to describe taking a task and making it process on it’s own.1 The idea of automation goes back for centuries, but a more recent example can be seen in the 19th century with the Luddites:2 A group of workers who feared, destroyed, and protested the use of machinery. They feared that the new machinery would inevitably remove them from the workforce. In a way, they were correct, but it took many years for manufacturing of any kind to become so automated that it was no longer considered a great career path for the every day worker.
The truth is that automation can be arguably one of the greatest things that has occurred throughout history. If you take the case of the Luddites, it not only created faster manufacturing which in return lead to cheaper pricing, but it also created more jobs. When you automate a process to such a high level you usually create positions to manage those high levels. You also create new positions for new problems that needed to be solved, but weren’t possible due to the constraints of previous tasks. More people needed to maintain that machinery and with faster production came to more expansion, allowing industries to hire more workers to use these machines.
There’s a catch to the positives of automation. Sometimes you can look at automation as a way of trimming a tree. In some cases you trim one end of the tree and more branches grow on the other side, but you may also have permanently trimmed those branches. In this case, if you automate a process to the point where you don’t need as much growth to handle it, you may have taken the first steps to what we see as today’s automation problem. Take a basic example of an accountant. Today, when you file your taxes, you most likely use a software tool that you pay for to get your tax return. Some people may still go to an accountant and/or financial adviser, but for the most part financial software such as “Turbo Tax” have dominated that industry. Let’s say that eventually some company creates a software tool that makes doing any kind of accounting work for a large and small scale company so simple, that you can take what dozens of accountants do in a year and reduce it to two. This creates a very large problem for anyone who’s still in school going for accounting and anyone who is currently an accountant. What happens when you automate a field so much that you just don’t need a large quantity of people to fulfill it?
Then comes another scenario where new jobs open up due to an automated process. If we continue our example of this financial software what kind of new processes would stem off of that? One possibility could be more developers to handle software bugs and updates. Another possibility could be hiring more IT staff to handle the large infrastructure demands that a software application of that scale could require. It’s important to note that these positions aren’t entirely related to accounting nor do they require the same type of skills. Some may even argue that these types of new positions require even more skill or fine tuned types of skills. It can create a difficult situation in the market in this case, because a company that used to hire a lot of accountants is now hiring a lot of IT to handle their switch into more automated processes. Even though accountants are still needed, they aren’t necessarily in high demand in our scenario.
To automate isn’t about having an AI becoming smart enough to do your job. Automation is about designing tools that make your job so easy to do, that you just don’t need a lot of people to do them. Look at the manufacturing industry today. While a lot of people will tell you manufacturing is dead, it’s really more alive than ever3. However, it’s nowhere near the same as when the Luddites had their protests for this introductory technology. Manufacturing has become automated to the point that it’s no longer considered unskilled labor. Manufacturers are looking for highly skilled, degree holding workers that understand the ins and outs of the tools that automated the unskilled workforce.
Automation isn’t about getting rid of your job, but it’s about crunching the numbers and realizing that a company can turn more of a profit with less people doing more work, than with more people doing the same thing over and over again. The problem is that as each job is introduced with automation, the requirements for skills gets more demanding. There is no way to tell how much damage automation will do. Even though history is repeating itself with the fear of automation, it is different this time. There’s no longer a need for more workers. Instead there’s a need for well educated and highly trained employees who can take the technology that is removing the unskilled work force and further it to automate even themselves.