Does artificial intelligence steal our jobs?

Does artificial intelligence steal our jobs?

Artificial intelligence (AI) has reaped an upsurge in interest in recent decades with so-called “generative AI” drawing much attention. These advancements inevitably led people to question their potential impact on employment. In this article, I will dissect some of the possible ramifications of these technologies.

When assessing the present, let alone when trying to predict the future, one should look at history. A tacit yet somehow strong assumption is that our jobs are going to be ultimately supplanted by AI-driven agents, thereby exacerbating the unemployment rates, provided that we keep on developing these tools. I beg to differ.

First of all, let me emphasize that I do agree with the imminent impact on the job market. Nevertheless, instead of focusing on just losing jobs, I would venture to claim that AI will affect our jobs overall. True, certain jobs are going to be transferred from some people to others. Thus, some will lose while some will gain, but I reject the idea of ending up with fewer jobs in the end.

Historically speaking, we have been automating for the past 200 years or so. Despite the often inexorable but still temporary transitional friction, the number of jobs has never gone down. There is the “Lump of Labor Paradox” that highlights the inaccurate perception of people that there is a fixed number of jobs available, and if some of them are taken, subsequently there are fewer available. As a matter of fact, the number of jobs has been growing steadily on a global scale, with local fluctuations. The reason why people worry about not having sufficient job opportunities is because many of them do not exist yet and they are difficult to predict.

This is perhaps our “cavemen thinking” when resources were limited and accessing them before our tribal competitors was paramount for survival. The economic prosperity of today’s age essentially means that the cake that we all share is growing (despite being cut in the meantime). We should by no means take anything for granted and our incessant exploitation of the planet’s resources is staggering (but that is for another discussion).

Notwithstanding the possible economic growth, blindly assuming that history is going to repeat itself might lead to blatantly wrong conclusions. But I feel that “Jevons Paradox” is another idea to take into account. Briefly speaking, increasing the efficiency with which a specific resource is used (e.g., coal), does not result in using less of it, but the exact opposite happens, the demand increases. Likewise, if a certain task takes a worker 10 hours to accomplish, automating part of it means the person will be able to do more of such tasks. This has been the case historically (with existing exceptions). Economic growth is the driving force behind our labor market shifts.

Consider the character of our jobs, in particular the ones that have been or are being affected by the adoption of AI the most. Many of them are regularly described as already being mechanized and still executed by people. It happens to both blue-collar as well as white-collar jobs. For instance, even software programmers or medical doctors might operate by habit on a daily basis unless they escape the comfort zone of complacency and deliberately seek improvement, regardless of how cognitively demanding their jobs were initially.

It’s important to note that AI-based solution needs to be deployed into production, integrated with existing systems, and maintained to retain its value. Anyone who has at least partially dealt with these issues realizes how cumbersome it is to do it properly. Furthermore, it is never just the “AI engine” that is being offered. Frequently it is part of a more generic service. We are merely adding one more item to our toolbox, outsourcing certain tasks to AI instead of hard-coded systems. Broadscale utilization of AI in conjunction with rapid alternation in work demands will take years, not months.

We should exploit AI to our advantage, to make it serve us. We need to care about possible consequences for these tools can be dangerous, especially in the wrong hands, but simultaneously they can be of great value.

In the end, it is the work itself that calls for mechanization. In 1913, Henry Ford vastly revolutionized the world of mass production with his conveyor belt assembly line. This is what capitalism is about, among other things. But these are the offshoots of our endeavor for efficiency and prosperity. The development of AI is the consequence, not the cause. It is not the AI itself that steals your job, but instead the process of attaining wealth through the means of prosperity founded upon efficiency and productivity. If a job can be easily automated, then it probably deserves to be. Ultimately, it might feel like a relief, as your job has already been performed by a human-robot anyway.