Fears mount that AI will replace humans in all kinds of work. As with any technological revolution, people in general think the technology will cause job losses rather than the other way around. Many economists, who have studied the role and impact of technological revolutions, disagree. After all, the vast majority of the population in all Western countries used to be employed in agriculture and were, ultimately, replaced by tractors and other novel technology. But this did not cause mass unemployment. In fact, the population today is much greater than back then and unemployment is quite low, so it should be obvious that we have more jobs than was the case in agrarian times. And they earn higher salaries too.
Technological revolutions—albeit at the time considered unique—have always had the same result: increased productivity, greater prosperity, and more and higher-paying jobs. The high standard of living we currently enjoy is due to these revolutions. What is to say the same will not be the case this time around?
But, assume the doomsayers are correct, that AI will cause massive job losses. Which jobs should we expect will disappear? Not the ones that create the most value. These jobs should be the last to go. The first to go are the ones that can easily be replaced or that make only marginal contributions to a firm’s bottom line.
There are two categories of jobs that tick both those boxes, which means they are both easy to replace and create comparatively little value. One of them is the bureaucratic control function in service of management. These are jobs that focus on creating and enforcing procedures, that produce statistics, create reports, and calculate performance metrics that support management decision-making but do little else. Simply put, these are the jobs that create (and make up) forms and see to it that they are filled out and signed, that “push papers” and collect data on how different functions in the business are doing. Their only or main contribution is to support management’s aim to control, supervise, and direct production processes that make up the core of the business.
The other category is what David Graeber called “bullshit jobs.” These are paid positions in organizations that are so pointless, unnecessary, or harmful to the value of the business that even the persons carrying them out cannot justify the job’s existence. Unlike the bureaucratic control function, these jobs do not even indirectly contribute to the firm’s bottom line. Instead, they are created for other reasons. For example, unnecessary assistants that managers can easily do without may be hired as a perk or simply for the vanity of the manager. Similarly, there are jobs that contribute only to the status or ego of persons in management, but that the business could just as easily do without. In other words, such jobs are a form of consumption rather than—as one would expect within a for-profit business—production.
Graeber—not an economist—did not recognize this difference, but would instead call much of the bureaucratic control function “bullshit.” This is perhaps more telling of his political views than anything else, since much of the bureaucracy is necessary to facilitate management. But he is not entirely wrong—some of it is indeed often “bullshit,” as jobs can exist as a remnant of previous procedures, bureaucracy that overshoots what is necessary or advisable “just in case,” or that result from path dependence instead of a focus on value creation. Some of these jobs are also the result of agency problems, as management is not, and often cannot be, sufficiently monitored by the principals. Consequently, management can use company funds for their own consumption rather than as contributions to the bottom line, which fits well with Graeber’s concept.
Why would we expect AI to displace workers in both of these categories? For the simple reason that AI will simply, quickly, and almost costlessly provide the necessary data to support management. No forms are necessary if processes and communication are computerized, the work place passively surveilled by AI interpreting a video stream to determine who does what and when, or that keeps track of both input and output as well as automatically calculates performance metrics. The bureaucratic control function will be no more, and management will have instantaneous access to the metrics needed to make effective and correct decisions.
This also applies to the monitoring of management itself, however, which means the end of costly bullshit jobs: management’s ability to use company funds for their own consumption will be greatly diminished; managers will not be able to employ people for their own vanity’s sake. Here too, then, AI will help streamline the business.
Certainly, a lot of people are currently employed in bureaucratic but largely non-value-adding functions like HR, as administrators, secretaries, etc. Many of them will likely lose their jobs to AI agents that can do what the employees do at a fraction of the cost—and much faster and more accurately. In fact, this change is already underway.
The results of this change are two. First, by replacing bureaucracies and bullshit jobs with AI, we get more effective and efficient businesses, better management, and lower cost overall. This increases competition, lowers prices for consumers, and increases productivity. The result is higher economic growth and, consequently, higher standard of living. Elon Musk has predicted that AI will double the rate of growth.
Second, it means the people in these soon-abolished jobs will need to find other jobs. Whereas the transition may be painful, history teaches that there is no lack of things that can be done—and that create value. It is simply a matter of imagining how these people’s skills can be used to the benefit of consumers—in an economy that is growing faster. This is the role of entrepreneurs, who—thanks to AI streamlining existing businesses—will have greater access to resources and, therefore, to opportunities.
The future, in other words, looks bright.
