Meaningful work in the AI revolution: Duty of Employers
Today I find myself reading a recent article published in the Journal of Ethics investigating whether employers have a duty to design meaningful work in the context of innovation driven by AI. This is a fascinating read that help me think much more deeply about what the responsibilities of employers are, as they make these changes and technological innovations, and what it means to create jobs that are designed to be meaningful.
The authors distinguish carefully between decent work and meaningful work. Decent work is having a job that pays you reasonably well, gives you reasonable vacation and leave time, has reasonable work hours, and adequate job security. Many of these conditions are the object of labor regulation, and thus they differ widely depending on which country you work. Yet, there is a global consensus that employers are to be paying attention to these conditions, and that violating these conditions is a form of labor exploitation.
Contrast this concept of decent work with meaningful work. Meaningful work is something we all aspire to, but it seems quite subjective, harder to define. Yet, it is important that it can be defined objectively because, as the authors point out, if the meaningfulness of a job is not defined objectively then you really cannot ask an employer to design the job with meaningfulness in mind.
The authors provide five dimensions of meaningfulness, so a job could be meaningful in one of these areas, but not in another. Ideally, the job should be meaningful in all of them.
Here are the five dimensions. It allows the employee to pursue a purpose. This is usually a shared purpose with the employer: satisfying the needs of customers, providing excellent care for patients, or designing innovative technology that is going to help follow citizens.
The job also must promote social relationships to be meaningful work. This is so important because most people need to work. They do not have a choice on whether they should or should not work because they’re not financially independent. Thus, so much of our lives are spent at work that if work does not provide meaningful social interaction, we are going to have difficulty becoming full human beings.
The job also has to allow us to exercise the skills we have and learn new ones. Not just technical skills, nor only social skills but also moral skills as one learns to make good decisions that are fair to both employer and employees.
In some ways, if we demand that employers design meaningful work, then we are saying that somebody who is employed by that firm has as part of their job to design meaningful work for him or herself and for others. Definitely, such a person would be improving and learning new skills.
Not surprisingly, the article mentions autonomy as this goes against the historical scars of Taylorism, and the current scars being produced by digital Taylorism in which the AI micromanages your work in minute-by-minute detail.
The last dimension is that human beings have a need for recognition and self-esteem. We need to be acknowledged when our work has been done well.
You may want to think about other possible dimensions that make work meaningful. This is such an interesting question because we are so diverse. There are some people who would go to work and say I don’t need to have a great degree of autonomy, but it’s really important to me that I will be learning skills. Other people may say social relationships are key to me, as is finding purpose, but learning new skills may not be as important.
The authors philosophically prove this duty of the employer to provide meaningful work from three ethical traditions: Kantian ethics, consequentialism and virtue ethics.
Finally, the authors look carefully at this issue of AI automation in warehouses. This is an example where digital Taylorism has become a concern. The most interesting aspect of this portion of the article for me is that the authors discern that while the introduction of algorithms to manage workers may make the work of some employees less meaningful, it also makes the work of other employees, namely the managers, more meaningful. This happened because while AI innovation robs autonomy, purpose, social relations and other meaningful aspects of the job from the former, it requires increased skills, renewed purpose, different social relations and so on for the latter.
The notion that AI automation may make the job of some people more meaningful while it makes the job of other people less meaningful is powerful because it reveals that AI automation may be conceptualized as a transfer of meaningfulness from one group to another.
Of course, we would not be surprised if those who benefit in terms of meaningfulness happen to also be the ones that benefit in terms of salary, promotions and other work conditions. While AI may make the firm more productive, it probably does so by trading in equity among workers.
The article did not discuss how to enact this duty in practice. How do you set the firm up so that as innovation happens it happens in a meaningful way? How about if we made it reciprocal? How about if the people whose job you design, design your job? Would that not set the right incentives to get it right? How about if the financial benefit for the innovation is shared primarily with those affected and only shared with the designers and owners after it has been deemed meaningful by the workers? If we are going to move from ideas to actions, someone is going to design incentive-compatible decision rights and structures to make it happen.
Overall, what a fascinating topic to research and how many applications it may have. We have barely scracthed its surface.
Ideas for dissertation projects
I always like to write my ideas for dissertation projects in case they help someone out there. In this case, the field is completely open. For example, a nationally representative survey could find what workers in a particular country think about each aspect of meaningful work. The respondents could rate their current jobs along each dimension. Does this vary by experience, type of field, gender, race, immigration status or other characteristics?
Experiments could be done in the behavioral economics tradition to determine what aspects of meaningful work individuals would be willing to trade for each other and why. How do individuals assess these trade offs?
Proposed policies such as whether workers should have a right to petition for redesign towards meaningfulness could be scientifically polled to see if they could gather enough support among citizens.
If a worker knows a way to redesign the job to be more meaningful that does not systematically cause undue burden on the employer, why would the employer not accommodate the request? This and many other policy ideas could be carefully formulated and tested.
Phenomenological studies and grounded theory studies could provide much needed information about how different people within the organization have or are experiencing the introduction of AI on their jobs with a focus on the inequities in terms of decent work and meaningful work.
Reference
Smids, J., Berkers, H., Le Blanc, P. et al. Employers have a Duty of Beneficence to Design for Meaningful Work: A General Argument and Logistics Warehouses as a Case Study. J Ethics (2023). https://doi.org/10.1007/s10892-023-09442-9
Disclaimer: Please read here.
