Which Jobs will AI Take?
My speculations, sector by sector
Below are some speculations about which jobs will be affected by AI over the next ten years or so.
I start with two beliefs regarding AI and jobs.
The new machine learning tools have their biggest impact in improving the human-computer interface. This means, for example, that robots become much easier to program. It means that your customers can have conversations with your systems, rather than try to deal with preset menus on phone calls, apps, or web sites.
AI will be held back most by organized political activity. Doctors are better than truck drivers at getting politicians to protect their incomes.
With that as a preface, let me go through industries using the government’s high-level classifications.
Agriculture, Forestry, Fishing, and Hunting
Mining, Quarrying, and Oil and Gas Extraction
Utilities
I don’t think that chatbots will be difference-makers here. But productivity has been going up for a long time in agriculture, and that will continue, with some of the gains coming from machine learning. Dangerous jobs in mining and oil and gas extraction will increasingly be done by robots. Utilities may find robots useful for installation and repair.
Construction
Productivity has lagged in this sector, in part due to regulation. But some of the challenge is that scale economies are difficult to realize, given the way that topography and other conditions vary. I do not foresee robots making rapid progress here, because of the absence of scale economies. I am willing to be corrected.
Manufacturing
For decades, our manufacturing output has increased while the proportion of the labor force doing production work has declined. This trend should continue as robots replace more workers.
Wholesale trade
For this sector, I don’t know what determines the boundary between what humans do and what computers do. I would assume that computers already do a lot, and there may not be much more that they can do.
Retail trade
Online shopping has already reduced the need for salespersons. Chatbots and robots should accelerate that process.
Transportation and Warehousing
Major job-killing will happen in this sector. Truck-driving is perhaps today’s biggest employer of blue-collar workers. How long do you think it will be until the majority of packages we order on line get delivered by driverless trucks and robots? Or maybe drones? I want to say five to seven years, but without wanting to make a large bet on that.
Information
A lot of jobs done today in this sector will soon be obsolete. I would say don’t learn to code. And what’s to stop an “influencer” from being replaced by a bot?
Actually, I think overall employment in this sector will go up. It’s just that we don’t know what the jobs will be. Nobody in 1990 could have predicted “web site developer.” Nobody in 2000 could have predicted “social media marketer.”
Finance and Insurance
I picture a lot of the jobs in that sector as involving sales and/or customer support. My guess is that these jobs will get taken over by chatbots. If you have such a job in 2030, I’ll be shocked.
Right now, Wall Street executives are less imaginative than I am. They are still thinking of AI as a gofer/research assistant. Mark McNeilly reports,
Major Wall Street banks are expected to slash up to 200,000 jobs over the next three to five years due to AI adoption, according to Bloomberg Intelligence. This significant reduction in workforce is primarily attributed to AI's ability to perform tasks traditionally carried out by human workers more efficiently and accurately.
The impending job cuts are expected to primarily affect back-office, middle-office and operational departments, where routine and repetitive tasks are prevalent. Positions involving data analysis, financial trend assessment and risk evaluation are particularly vulnerable, as AI systems can process vast amounts of information and generate insights at speeds far surpassing human capabilities.
“Routine and repetitive” tells me that they are still thinking in terms of 20th-century computing. The new machine learning can do much more.
Real Estate and Rental and Leasing
A lot of these folks are used to enjoying regulatory protection. The potential efficiency gains from rationalizing this sector are not going to be held back forever. Five years from now, there could be many fewer humans working here.
Professional, Scientific, and Technical Services
Public Administration
These are your lawyers, accountants, and consultants of various stripes. I think that AI ought to reduce employment in this sector, but I am pessimistic that it will do so. The thing is, they tend to make work for one another—lawyers especially. So if an AI allows lawyer X to handle twice as many cases, does that just mean more work for lawyers Y and Z? Plus, lawyers are very powerful politically, so taking away income from them tends to be hard. And they tend to make more work for accountants and consultants, too.
Management of Companies and Enterprises
In a typical large organization, there are people with management titles who are not really doing management work. And there are people doing management work without having a title. I expect that we’ll see AI’s in both situations. But there will still be plenty of legit management jobs around for humans.
Administrative and Support and Waste Management and Remediation Services
Robots and surveillance systems are going to take away a lot of jobs here.
Educational Services
Like real estate, this is a sector that is overdue for disruption. I think that there will be a trend away from legacy schools and toward smaller, more specialized schools. There will be AI mixed in to both types, but I don’t think that the overall number of jobs will decline.
Health Care and Social Assistance
As Moses Sternstein keeps pointing out, this is where all the American job growth has been the last couple of years. It’s where all the demand increases are happening, as the population ages. It seems reasonable to predict robust growth in employment in this sector for both humans and AI.
Arts, Entertainment, and Recreation
I am much more inclined than most people to view AI as an aid to creativity, and even creative itself. As with health care, the demand is high enough that there is room for gains in both human employment and the utilization of AI. But I forecast a big decline in jobs tied to the major media of the 20th century.
Accommodation and Food Services
I think that productivity will increase faster than demand. The net effect on jobs will be negative.
Other Services
To give you an idea of how broad this category is, it includes automobile repair and social advocacy organizations, among many other services. This is too much of a grab bag for me to make a general prediction.
Conclusion
Technological change can find some workers unable to adjust. An interesting take on this is Alexander Field’s book A Great Leap Forward, which looks at the 1930s as a period of innovation.
We are told that the Great Depression consisted of some shortfall in “aggregate demand,” but some of the employment decline might have come from technological innovation. I forget whether it was there or in a different book that I saw the anecdote that in the 1920s cigars were rolled and glass lightbulbs were blown by humans, but machines did this work in the 1930s. Maybe we should call the 1930s, with a nod to Schumpeter, the Great Creative Destruction.
The full-employment economy of the late 1940s was very different from the economy twenty years prior. The pre-Depression jobs did not return. The postwar jobs involved new products and processes. And the work force was better educated and included more women.
As an aside, I think that the means of providing protection against violence, both locally and internationally, are going to change. There are going to be fewer humans risking their lives confronting bad guys. There is going to be more surveillance, more robots, and a lot more focus on intelligence-gathering. Of course, AI will play a part in this, and the roles that humans play will be quite different from what they are now.
Looking ahead, it is easier to speculate on which jobs can be displaced than it is to identify the new jobs that will emerge. If I knew exactly how innovation would shake out, I could make a killing by investing in the right products and processes. I do not have that precise level of insight. But just because it is easier to spot the jobs that are in jeopardy than the jobs that are on the horizon does not mean that mass unemployment is around the corner.
Like most economists, I believe that the net effect of technological innovation is to restructure the labor force. I think that there will be plenty of jobs, especially in information services, health care, social assistance, and entertainment. I just cannot tell you exactly what those new jobs will be.
substacks referenced above:
@
In my space AI is mostly used to help doctors upcode patients and increase Medicare revenue.
There is also an arms race with AI re-submitting claims and other AI denying claims. Like your example of lawyers creating work for each other, I could easily see competing AI creating work for each other.
There is a lot of dumb Red Queen Race Bullshit out there that people forego because it would be "too expensive" to do. If you lower the price, they can increase the quantity. This is more or less the thesis behind why we don't have Keynes 15 hour workweek.
In a longer reply to John Hall I note that if olds and poors prefer humans CMS has a lot of ways to incent using humans at the taxpayers expense. CMS would much rather have sales brokers fielding member complaints then having to do it themselves.
However much creative destruction went on in the '30's there was no reason for the Fed to allow
so many markets not to clear.