Everyone's debating what AI will do to the job market. Paul Osterman thinks we're asking the wrong question. The MIT Sloan professor and author argues that 35% of the workforce is already treated as expendable - and AI didn't cause that. Employers did.
His new book, Disposable Workers: The Transformation of Employment, is a wake-up call for anyone who thinks the problem started with ChatGPT.
Mindstream: You place employer behaviour at the centre of how work is changing, rather than technology. Does that hold up in an era when AI genuinely seems to be reshaping work?
Paul Osterman: There is great uncertainty about the impact of AI on employment levels as well as the mix of skills needed.
Some expect a massive impact on employment levels, whereas others see it like the internet: changing work but not destroying it. Firms will therefore want to use more disposable workers because they are uncertain about their future staffing needs.
Your research found that around 35% of the workforce is in some form of disposable status. Is it inevitable that AI will increase the use of disposable workers?
AI will certainly impact employment, but the nature of that impact is not at all clear. What I feel very confident about is that 35% of workers are treated as disposable today. My best prediction is that AI will exacerbate this.
Which human skill do you think is becoming more valuable in the AI era?
AI cannot replace human interaction, and it is not at all clear that AI can come up with new ideas.
With this in mind, creative thinking skills and skills related to interacting with other people, ranging from management to child care and health, will remain valuable and become even more valuable in the AI era.
You're cynical about CEOs who blame AI for layoffs that have more conventional explanations. How do you tell the difference between genuine AI disruption and a convenient cover story?
I’m not cynical. I just think that firms are profit-maximising organisations. This is what they should be: they innovate and create jobs.
But they also want to squeeze their costs, and this leads to problems for the workforce. It is also known that firms use recessions as an opportunity to restructure their workforce.
With this in mind, certainly some layoffs today are justified via AI but driven by other considerations. Without looking deeply into staffing patterns on a firm-by-firm basis, it is hard to characterise specific restructuring events.
Is the disposable workforce strategy actually good business, or are companies paying a price they haven't noticed yet?
There is a long body of research that shows that what is termed “high road” employment practices–better wages, more training, and more respect - lead to lower turnover and higher product or service quality.
In the case of disposable workers, my research shows that companies are paying a price because disposable workers are less willing to put in extra effort for the success of the organisation for whom they work.
Do you think we’re in an AI bubble right now?
The incredible amount of money going into AI investments and the lack of profitability of AI firms are certainly reasons to be nervous.
I have no research-based prediction on this, but that said, individual investors should be cautious.
AI is making it easier than ever to manage, monitor and replace workers. Does that change the maths of the disposable workforce strategy?
The use of disposable workers is attractive to firms for a variety of reasons: ease of staff reductions, savings on benefit costs, and driving down wages.
AI will not significantly affect these calculations.
Which recent AI development stopped you in your tracks?
I submitted the draft of an academic article I am writing to an AI agent, and it gave me amazingly useful and insightful comments and critiques.
What advice would you give to someone who finds themselves in disposable work and wants out?
There is clear evidence that high-quality skill training pays off in the job market.
But for those people who are not in a position to have personal economic power due to the level or uniqueness of their skills, they should think about working with a community organization to push firms to raise their employment quality and convince policymakers to take this issue seriously.
Finish this sentence: In five years, AI will make people feel ___ about their work. Why?
Better. The spread of disposable workers will become a salient political issue, and policymakers will respond.
There are recent examples such as New York City proposing to regulate the quality of Amazon driver jobs or Massachusetts opening the door to bargaining for gig workers.
Paul Osterman is Professor Emeritus of Human Resources and Management at the MIT Sloan School of Management. His most recent book is Disposable Workers: The Transformation of Employment. His Other books include Good Jobs America, Who Will Care for Us?, and The Truth about Middle Managers.
You can get Paul’s new book, Disposable Workers: The Transformation of Employment, here!