One of the closing statements last class was the question I posed to Prof. Kane: “Are there any jobs that you think are immune to automation?”. Without a doubt, this is nearly an impossible question to answer. The visiting professor stated her belief that robots could do anything besides love; if you’ve ever seen Spielberg’s Artificial Intelligence, you may not agree with her statement. After spending a class talking about the threat of being replaced by robots, I thought I’d do some research to explain why this Intelligence Revolution is different than the automation that came with the Machine Revolution of the 1900s. I also would like to explore which, if any, professions will be able to survive this revolution in our lifetime.
Although jobs have been destroyed by automation in the past, there have always been other jobs that opened up as a result. For example, horse and buggy drivers in the early 20th century became cab drivers when cars got on the roads. There are some today that use history as an argument that this will continue into the future.
Besides that classic example, you can even go more recently to the example below. We continuously need fewer and fewer farmers to feed an ever growing population due to advances in equipment and biotech. Secretary was the most popular job 30 years ago, but those jobs have been replaced by software developers. Today, the most common job is truck driving. Some believe that these truck drivers will be able to find jobs in fields not yet in existence once autonomous vehicles become the norm.
In the video below in my sources, the narrator gives a hypothetical story of two horses speaking to each other. The one horse says to the other that they’ll still be able to find better, less demanding jobs now that cars have replaced them. To me, the greatest quote from the video in the link below is the following: “There isn’t a rule of economics that says better technology makes more, better jobs for horses. It sounds shockingly dumb to even say that out loud, but swap horses for humans and suddenly people think it sounds about right.”
So, why is now different than before? Old AIs were built for specific purposes (i.e. playing chess or screwing a nail into a car) and the new are general-purpose. They are able to learn how to complete tasks through either programming or in some cases even watching.
It’s pretty obvious that truck drivers will be replaced once we are confident in our autonomous vehicles, but what about other professions? A 2016 study conducted by McKinsey of over 800 occupations concluded that technologies could automate 45% of activities people are paid to perform and that 60% of occupations could see about 30% of their activities automated with only the technologies available today; these percentages will continue to increase down the road (the video likens our current point in AI to computers in the ’70s.) Of course, companies will not automate for the sole purpose of automating: there must be cost savings, an increase in productivity, etc.
As seen above, given current technology, robots are able to replace predictable work (repetitive tasks such as packaging) but currently cannot replace more unpredictable work reliably (collecting trash or making beds in hotel rooms, where every situation is different from the last).
The lowest potential for replacement? Developing/managing people and applying expertise in decision making. Although we can program a truck to find the optimal route, we still need to tell the program what the goal of our route is and double check that the suggestion makes sense. Through these two actions, McKinsey zoomed out and identified two low threat industries: healthcare and education.
We identified healthcare as a potential sector in our class through the readings/videos, but McKinsey goes further. They’ve identified fields that require physical touch (such as dental hygienists) to have less than half the activity replacement of a nurse (13% to 30%). But this is not to say that complex activities will not be replaced; McKinsey believes that simple anesthesiologists can be replaced by robots at the current state of technology.
Surprisingly, McKinsey identified education as the safest field. Given the fact that many of my classes have used prerecorded lectures, I came into this thinking that education could be replaced somewhat soon. McKinsey went beyond my original hypothesis: “To be sure, digital technology is transforming the field, as can be seen from the myriad classes and learning vehicles available online. Yet the essence of teaching is deep expertise and complex interactions with other people. Together, those two categories—the least automatable of the seven identified in the first exhibit—account for about one-half of the activities in the education sector.” Creating connections with people is a defining aspect of being human. Many of my favorite teachers were awesome not because they were the most knowledgeable, but because they knew how to connect with their students and foster a positive learning environment. Looks like you made a good choice Prof. Kane.
Looking down the road, McKinsey touches upon the rapid advancements in machine learning that will eventually lead to the ability to perform unpredictable activities. Something that we did not touch on in class is what would happen if computers were able to acquire language skills on par with the average person and use this to easily have conversations. McKinsey concludes that this would lead to great automation leaps in multiple occupations. Lastly, they emphasize that while the technology may not be available to replace certain tasks right now, there’s no telling how this will change in the near future.
What is very concerning to me is the lack of government concern for potentially 10s of millions of jobs lost. Even the greatest tech giants like Elon Musk are openly proclaiming that a universal income will be necessary. When it comes to our largest incoming problems, automation and climate change, our governments’ mentalities seems to be to deal with them as they happen. This might work with budget constraints, but I don’t think they’ll easily deal with millions of bored, unemployed, angry, and hungry citizens.