Generative AI and Worker Voice

Much has been written about the potential impact of generative AI in the workplace, but the perspective of one set of key stakeholders has often been left out: workers themselves.  What’s more, research has found that incorporating the perspective of end users such as workers into the development and implementation of new technologies tends to produce better results than strictly top-down technology development approaches.

So argues a new working paper coauthored by an interdisciplinary team of  MIT researchers. The new paper, “Bringing Worker Voice into Generative AI,” was recently published online as part of a new collection called An MIT Exploration of Generative AI. This collection consists of 25 papers on the societal impact of generative AI that were authored by members of the MIT faculty, with support from seed grants from MIT.  “Bringing Worker Voice Into Generative AI” was coauthored by Thomas A. Kochan, the George M. Bunker Professor Emeritus at the MIT Sloan School of Management and a member of the faculty of the MIT Institute for Work and Employment Research (IWER); Ben Armstrong, Executive Director and Research Scientist at the MIT Industrial Performance Center; Julie Shah, the H.N. Slater Professor of Aeronautics and Astronautics at MIT and Director of the Interactive Robotics Group at the MIT Computer Science and Artificial Intelligence Laboratory (CSAIL); Emilio J. Castilla, the NTU Professor of Work and Organization Studies at MIT Sloan and Co-Director of IWER; Ben Likis, 2024 MBA Candidate at MIT Sloan; and Martha E. Mangelsdorf, Director of Strategic Communications at IWER.

The researchers conducted more than 50 interviews with a cross-section of stakeholders that included AI developers, business leaders, labor leaders, academic experts, and policymakers. In their interviews with AI developers, the authors learned that some factors weigh against incorporating worker input into the design of AI tools—including the pressure technology vendors feel to bring products to market quickly, without taking the time to incorporate worker input. “A number of our interviewees, along with the published works by scholars, suggest that there is a bias toward labor-displacing/replacing humans in the minds of many inventors or technology vendors,” Kochan, Armstrong, Shah, Castilla, Likis, and Mangelsdorf write.

But the authors also found examples of other approaches. They note, for instance, that Microsoft Corp. recently formed a partnership with the AFL-CIO labor federation to share information about how AI is impacting the workplace and to incorporate worker voice and experiences into the development of AI technology. Kochan et al. see this as an approach that might be a model.

In general, “the broader the set of stakeholders involved in defining the problems and opportunities that generative AI technologies can address, the more likely it is that these tools will be used to augment how workers do their jobs rather than displace them,” Kochan, Armstrong, Shah, Castilla, Likis, and Mangelsdorf write.

The authors also argue that, more than many previous new technologies, generative AI is particularly conducive to “bottom-up” innovation processes by workers and teams who use the tools. “This ‘democratization’ feature of generative AI opens up possibilities for the workforce to exert greater influence in how the technology evolves than was possible during many prior technological changes,” Kochan et al. observe. Moreover, the authors’ interviews with business leaders revealed that some companies are indeed incorporating “bottom-up” input from employees as one part of generating promising use cases for generative AI.

The research team also interviewed U.S. labor leaders, and the authors observe that unions have both an increasing interest and capacity to engage proactively with other stakeholders about the impact of AI. Examples include the partnership with Microsoft and collective bargaining agreements reached in 2023 between the Alliance of Motion Picture and Television Producers and the Writers Guild of America (WGA) and the Screen Actors Guild-American Federation of Television and Radio Artists (SAG-AFTRA); these new collective bargaining agreements incorporate protections for members of those unions regarding the use of generative AI in their professions.

The authors conclude with set of recommendations designed to encourage the incorporation of the perspective of the workforce into four phases of AI design and development: 1) defining the problems and opportunities to be addressed; 2) designing the technical and work process features that need to be integrated; 3) educating and training the workforce in the skills needed; and 4) ensuring a fair transition and compensation for those whose jobs are affected.

“We see this report as a first step in encouraging ongoing research, teaching, and outreach by MIT faculty and students to help bring the voices of the workforce into generative AI,” Kochan, Armstrong, Shah, Castilla, Likis, and Mangelsdorf conclude. “We hope this report serves as a catalyst for taking up this critical challenge and opportunity.”

 Read “Bringing Worker Voice into Generative AI.”