Are Engineers Designing Their Robotic Replacements?

The profession could be putting itself out of work


“The robots are coming for your jobs!” That was the gist of numerous news reports following the release of the 2016 U.S. Economic Report of the President. My first thought on reading this was that anyone who saw the videos of clumsy robots falling helplessly during the recent DARPA Robotics Challenge must have been incredulous: “That’s what’s coming after my job!?”

My second thought was more sobering. Robots are, after all, only a subset of the computerization leading to the automation of traditional jobs. As engineers we can see steady progress in machine learning, artificial intelligence, and big data. Contemplating this, it suddenly occurred to me to worry about engineering jobs. Are they threatened as well?

History has shown that while automation increases productivity and creates new jobs, it also displaces jobs. However, the president’s economic report notes that in recent decades the rate of job destruction has outpaced the rate of job creation, resulting in less overall participation in the workforce. Regardless, stopping or slowing the pace of technological development has always been a fruitless exercise. While we engineers may be responsible for implementing tech’s advance, we also seem powerless before its inevitable progression. As technology’s handmaidens, we will undoubtedly provide at least complicit support, even as our own engineering jobs are displaced.

An earlier report, by Carl B. Frey and Michael A. Osborne, of the University of Oxford, examined the potential for technological disruption in 702 different occupations. Their study concluded that 47 percent of current U.S. jobs are at risk of displacement. In considering the risk for each occupation, they evaluated the technological bottleneck that would be most likely to occur in automation efforts. The bottleneck could arise in any of three different aspects of work—the need for perception and manipulation, the need for creative intelligence, and the need for social intelligence.

One of the occupations studied was listed as “computers, engineering, and science.” Frey and Osborne’s data show it among the least threatened by automation, which they attribute to the high degree of creative intelligence required in science and engineering. Nonetheless, they qualify this conclusion in a cautionary amplification, saying that although today the roles of the computer and human in these fields are complementary, “it is possible that computers will fully substitute for workers in these occupations over the long run.”

Software development is singled out as particularly amenable to computerization. Inasmuch as software is becoming an ever bigger component of engineering design and development, this might raise doubts over whether this will remain a safe haven for future engineers. However, even though software development involves the management and manipulation of symbols within a set of constraining rules, I suspect that fully automatic programming may be a more distant objective than that predicted by the economists.

As an engineer, I always think that as lower-level tasks are turned over to computers, we move up the stack to working at higher levels of functionality. Accordingly, we keep our jobs. But an economist might say that engineers have thus become more individually productive and so employers need fewer of them.

There’s a lot more I’d like to say about this, but I have to run. I hear clomping footsteps behind me. I see a robot back there and I think it’s gaining on me.