White collar jobs aren’t safe from automation, and eventually, robots and AI software will be able to do your job better than you can.
Automation has slowly been creeping into industry for many years now. The shift to machine work first began with the industrial revolution, where 70 percent of all American workers made their livings on farms. When the industrial revolution swept through the food and manufacturing industries, robots replaced all but 1 percent of those workers. The trend continues, but essentially all of the history of automation has dealt with manual to mechanical labor, never have we seen a technological revolution that faces the world right now.
You have likely already been involved in a conversation discussing how AI and robotics will replace factory workers, taxi drivers, and essentially all in the service industry. Unfortunately, this staggering assumption will soon become true, based on all estimates and technological projection. The one thing that human workers have on their side is time. This takeover of AI and robotic workers will occur slowly, but steadily. Bloomberg projects that by 2020, 5 million jobs will be lost to robots and AI. This is a distressing prediction, but one that only represents a fraction of the workforce. Replacements to manufacturing line workers have been happening already, but the progression of manufacturing robots has resulted in highly specialized and expensive machines. What the world has yet to see, is the highly-versatile and cheap machine. A machine that can learn, one that is. . . intelligent.
So, the total replacement of blue collar workers appears obvious within say, the next thirty to fifty years. Common stigmas place white collar workers like engineers and computer programmers outside of the potential to be overtaken by robotic workers. There will always be a need for someone to design and program robots, right?
Future AI computers, if we can call an intelligent machine that, will be able to learn, and even design, just like we can. An expansive article from Jason Tanz at Wired takes a look to the future of computer programmers. Programming code will soon become a task of the past, rather ‘programmers’ will train AI programs much like training dogs. Early AI programmers will essentially become teachers to newly born artificially intelligent computers. However, this ‘student,’ if you will, can retain knowledge for an infinite amount of time, and instantly translate knowledge to its robotic counterparts. While it may sound like something straight out of the Terminator series, it is the very real future that we face in the technological future.
To break away from speculation and dive into existing proof-of-concept technology, let’s look at a robot named Baxter. Baxter is an industrial robot designed to learn any and every task, designed to work right alongside humans. The robot was designed by a former MIT professor, who also invented the Roomba. Baxter can learn: he can watch anyone and everyone and understand what he needs to do to complete a task. He is possibly the most practically advanced AI robotic system in existence, and his price tag sits at cheaper than your family SUV, $22,000. This is only one example of AI technology already in existence, and robots have their eyes set on white collar engineering and technical jobs.
A look at Baxter, the learning universal robot.
As engineers, we have to juggle hundreds of constraints and designs per project. This can be a stressful task, and ultimately one that takes many years to master. AI computing machines will be able to develop the same skills that took you years to develop, in only months, days – minutes. Designing is ultimately a qualitative process, and AI will eventually be able to do it better than you can. Imagine a machine that can consider virtually every design constraint and possible flaw simultaneously and instantly. The main argument for the safety of engineers from automation lies within the skills that engineers take years to develop. That argument rests on a common fallacy when considering AI, that AI systems are bound by their ability to think.
There is no foreseeable hurdle that will stop the progression of artificially intelligent machines from advancing to a point comparable with authentic human thought.
This is the scary reality of technologies current state. Moore’s law may soon be coming to an end by current projections, but with the advances in quantum computing and AI software, automation won’t be limited.
Here’s the good news, based on the current progression of AI, you likely won’t see engineering white collar work being automated until the very end or after your career. Even then, there will still be a need for engineers, but their duties will drastically shift from the present. However, designers, engineers, and those in the technical fields aren’t safe from being replaced by the computers they will likely build.