For years, we’ve been hearing that automation is coming to white collar jobs, crossing a barrier that we previously thought uncrossable. But is this truly the case? And if so, which white collar jobs are the most susceptible to being replaced?
Blue Collar vs. White Collar: The Automation Dilemma
These jobs are:
- Predictable. First, these jobs are predictable. They tend to unfold in the same ways; the potential problems are easy to anticipate and list, and most of these problems can be resolved in a formulaic way. There are few external variables that can influence the course of action in these positions.
- Repetitive. These jobs are also highly repetitive. The first day on the job looks very similar to a day three years into the job. You can set a schedule and follow it consistently, and there aren’t many curveballs or surprises.
- Low risk. Though some blue collar jobs deal with multimillion dollar equipment and high-stakes orders, many positions are relatively low risk. They deal with production and other low-level tasks, so mistakes and errors aren’t especially costly.
- Easy to learn. For both AI systems and humans, these jobs are relatively easy to learn. There may be lots of facts to memorize and demand for practice, but anyone with a high school degree and a willingness to learn can get the hang of things in a matter of days (or a matter of hours for adaptive AI).
For example, learning to complete a task on an assembly line is relatively easy. Almost anyone, including a machine, can learn the steps necessarily to efficiently and safely produce an item.
By contrast, many white collar jobs are:
- Unpredictable. The problems you face in a white collar position are often more nuanced or qualitative than those in blue collar jobs; this makes them unpredictable, and therefore hard for an AI to learn.
- Multifaceted. High-ranking, white collar professionals typically juggle many different types of responsibilities. Consider the founder of a new business, who must make decisions related to marketing, HR, accounting, and product development all in the same day.
- High risk. The stakes are often higher in high-ranking white collar jobs. People in the highest positions often make decisions and take actions that can influence the movement of millions of dollars—and a mistake can be deadly.
- Hard to learn. Though many white collar jobs are easily learnable, some require many years of effort. A human being may require several years of university-level education, combined with many years of onsite experience, to do the job effectively. An AI, at current levels of sophistication, may require an exceptional length of time to do its job effectively—and even then, it may need additional supervision.
For example, requires you to incorporate many different types of information, and respond to an ever-changing work environment. To be an effective CMO, you need years of education and experience, and you’ll be making decisions that can impact multi-billion dollar corporations in some cases.
AI is constantly getting better, so from a certain perspective, it’s only a matter of time before AI algorithms become sophisticated enough to handle more complicated jobs.
We’ve already seen this play out in a handful of areas. For example, AI is being to crunch numbers and intelligently recommend new strategies. It’s being used to write content on a regular basis—and the content it produces is almost indistinguishable from content generated by human writers. AI is also being used in the medical field, responsible for executing precise surgeries on patients and analyzing and filling prescriptions for patients in a pharmacist role.
Perhaps the most promising area of development in AI and automation is the inclusion of human emotion. Engineers are developing chatbots and other forms of AI that can both “understand” and replicate human emotions; in the near future, you may be able to have an open conversation about your feelings to an . You may hear notes of compassion in the voice of a chatbot when you call a customer service line. You may even rely on an AI algorithm in a therapy session.
Tech optimists see these forms of progress as an indication of where we’re headed. In 1996, AI was sophisticated enough to beat the human chess champion Garry Kasparov in a game of chess. In 2015, AI became sophisticated enough to beat human go players (with go being considered one of the most complex traditional games). It wasn’t that long ago that it was considered impossible for computers to become advanced enough to beat human players in either chess or go.
The mentality here is that we’ve seen AI do “impossible” things on a consistent basis. Every year, we develop machines to accomplish something new that was previously thought to be unthinkable. Following this line of logic, it’s hard to assert that there’s anything truly impossible for machines to do.
AI and Humans: A Perfect Partnership?
Of course, just because AI could have the power to accomplish human responsibilities doesn’t mean AI is going to replace human beings in a takeover of their current jobs. There are a number of possibilities that could allow humans and AI to work together in harmony.
In the first vision, AI is merely used to handle responsibilities that humans can’t handle, for one reason or another. For example, completely automated surgery may be reserved for handling surgeries when human surgeons are busy or unavailable; there’s a doctor shortage in the United States, and AI-based systems could arise to help fill the void, thereby jeopardizing few (if any) human jobs.
In another vision, AI could mostly serve as a complement to human thinking. Rather than depending exclusively on human creativity or AI-powered calculus, the best systems would reflect a partnership between these modes. Human beings in analyst and creative positions would utilize AI as tools to enhance their own skills, knowledge, and abilities. Their .
It’s also worth noting that AI could replace some human responsibilities without actually replacing the humans engaging in those responsibilities. For example, if part of your job is generating marketing reports, the AI could take over that portion of your workday—and you could spend more time handling other, more complex responsibilities. In this vision, the nature of white collar work would gradually change, with white collar workers taking on bigger, higher-level, and more complex responsibilities over time. This could result in an even bigger skill gap between blue collar and white collar workers, and spur a number of problematic economic side effects; however, it wouldn’t mean the true end of any white collar jobs.
Where Humans Still Excel
For all its advancements, it’s unlikely that AI will be able to trounce human beings in certain areas. For example, even the best AI algorithms we can imagine will only be able to replicate human emotions; they won’t be able to offer sincere empathy. Many customers, patients, and other consumers will strongly prefer a customer experience , no matter how “good” the AI alternative is.
Additionally, humans remain far superior to machines when it comes to generating novel concepts, imagining creative solutions, and making sense of complex ideas. at the very least, human minds will need to work together with AI to come up with the best ideas and resolutions.
The Bottom Line
So what’s the bottom line here—is automation truly “coming for white collar jobs?” The answer is yes and no. There are already many white collar applications that are being handled by advanced AI algorithms and machines, and the capabilities of these systems are only going to expand in the immediate future. White collar workers in all industries and at all levels will need to remain adaptable, and prepare to forge a place for themselves in an evolving technological workplace.
However, threats that white collar jobs will be replaced by machines, leaving behind a tormented trail of unemployment and depression, are mostly unfounded scaremongering. It’s unlikely that humans will be totally replaced, so long as they’re willing to grow, learn, and change with these new technologies.
View original article here Source