The Future of Employment: AI Edition

Will AI take my job? Will AI replace it with a crappy one? What will my industry look like five years, or five months, down the road?

Ask all you want, but not even ChatGPT can give you the answers.

With the future so uncertain, all we can do is look to the findings of the experts. Fortunately for us, they’ve been thinking about this problem long before AI became so suddenly mainstream.

In 2013, an academic paper entitled “The Future of Employment: How Susceptible are Jobs to Computerisation" was published.

Computerisation back then didn’t have as much to do with AI as it does now. Regardless, all those things that computers could do better, faster, and cheaper still apply in today’s world. Maybe even more so.

The authors studied 702 occupations and ranked them on a scale of least likely to most likely to be replaced by computerization. At the very top of the least likely list was Recreational Therapists. At the bottom, with a 99% likelihood of being replaced by computers, were Telemarketers. We live in a just universe. Early casualties at the time this paper was written included bookkeepers, cashiers, and telephone operators. While not all these jobs have vanished, their ranks have been drastically reduced. Talked to an operator lately?

One of the more useful aspects of the paper is the vocabulary it provides. As much as we see AI bringing about great change, we can’t always label what is happening. Below are terms for some of the forces we see at work on a regular basis.

First off, is labor market polarization. In other words, there are increasingly fewer higher paying jobs and many more lower paying jobs. The jobs that pay the best call for the ability to perform cognitive tasks which have a strong correlation to educational attainment.

Next is the theme of “deskilling.” In the 19th century, manufacturers largely substituted skills for the simplification of tasks. Over time, artisan shops and the craftspeople therein began to vanish. The skills of the artist became devalued and fewer people pursued craftsmanship as a path to success.

A third theme is educational wage differentials. While these differentials narrowed in the middle of the last century, they have been expanding since the 1980s. AI will most likely accelerate this trend. This has caused what economists call a “hollowing-out” of middle-income jobs.?

You may wonder that since technological advancement is nothing new, why aren’t all of us already out on the street? For this we can thank David Ricardo who in 1817 laid out his argument for why machines won’t necessarily replace humans. In his famous “The Principles of Political Economy and Taxation” his logic goes something like this:

An increase in the efficiency of production will which reduces the price of one good, will increase real income and thus increase demand for other goods. Technology has two opposing effects on employment. First as technology replaces labor, there is a job destruction effect, requiring workers to reallocate their labor supply. Second comes a capitalization effect as more companies enter higher productivity industries where profits are robust and, in turn, hire more workers.

Whether we keep our jobs or not, it seems clear there will be no escape from the nauseating churn of professional flux.

To address “workers needing to reallocate their labor supply,” we will all need to learn new skills and apply them to industries where jobs are just emerging. If there was ever a time to continually "reskill," this would be it.

As much as we can argue that creative jobs qualify as non-routine, i.e. not easily done by a machine, we should not relax. If big data has done anything, it is to encroach upon those tasks that until recently we thought computers could not perform. These are things like reading radiology charts, in-person monitoring of various infrastructure, and even doing something such as writing funny ad copy. The arts would seem to be decidedly human stomping ground, but as Sam Altman, co-founder of OpenAI, recently pointed out, our belief that the arts would be the last refuge from AI incursion is turning out to be just the opposite.

Strangely, the savior of creative jobs from the onslaught of AI is, of all things, a subtle thing called subjectivity – the curious sliding scale of what constitutes “good.” As we all know, art is subjective. Creativity is a moving target. What is great today, may look outdated tomorrow. What was deemed excellent last week, may look merely okay this afternoon. Few would want to admit it, but the merry-go-round of arbitrary human taste is what keeps creatives gainfully employed. We all love options:

“In these and many other applications, generating novelty is not particularly difficult. Instead, the principal obstacle to computerizing creativity, is stating our creative values sufficiently clearly that they can be encoded in a program. Moreover, human values change over time and vary across cultures. Because creativity, by definition, involves not only novelty but value, and because values are highly variable, it follows that many arguments about creativity are rooted in arguments about value. Thus, even if we could identify and encode our creative values, there would still be disagreement about whether the computer appeared to be creative. In the absence of engineering solutions to overcome this problem, it seems unlikely that occupations requiring a high degree of creative intelligence will be automated in the next decades.”

I hear many breathing a sigh of relief. And it is somewhat justified. But there is a lot to unpack here. With ChatGPT and Midjourney, the notion that computers are not creative is no longer a given. In fact, by dint of sheer combinatorial explosiveness, a computer can frequently outdo a group of humans.

All told, we are left with an uneasy feeling that maybe this time around Ricardo’s Principle may not hold, and beyond jobs the technological imbalance could lead to massive sociological dislocations. In this case, the earliest days of the industrial revolution might provide some insight.

In 1589, the inventor William Lee created a machine that replaced the hand-knitting of stockings. On paper it was a manufacturing tour-de-force, but the Queen thought it was a societal nightmare waiting to happen.?

“Thou aimest high, Master Lee. Consider thou what the invention could do to my poor subjects. It would assuredly bring to them ruin by depriving them of employment, thus making them beggars.”

We are again aiming high. And there is work to do. As people come to grips with the power of AI, it raises questions as to whether or not we are prepared to answer some of its biggest questions. One of them being, how will society deal with the windfalls AI provides if and when they happen?

“The balance between job conservation and technological progress therefor, to a large extent, reflects the balance of power in society, and how gains from technological progress are being distributed.”

How AI-created wealth is fairly and equitably distributed will be one of the biggest issues of time. There are no easy answers. But until that day comes, how we all stay relevant seems to be the most pressing issue.

And for that, we should double down on what we've always known and lean into the superpower all creative people possess:

“Our findings thus imply that as technology races ahead, low-skill workers will reallocate to tasks that are non-susceptible to computerization – i.e., tasks requiring creative and social intelligence. For workers to win the race, however, they will have to acquire creative and social skills.”

Native soft skills will be a mighty asset in the Age of AI.