Interviewer: Could you each please introduce yourself?
Laurel Smith-Doerr: I'm Professor Laurel Smith-Doerr. I'm a sociologist, and I've been at UMass for 10 years now. I came here as the inaugural director of the Institute for Social Science Research (ISSR), and now I'm directing the UMass ADVANCE program, which explores knowledge-driven solutions for faculty equity based on research. Since I'm a sociologist of science and technology, a lot of my work—as well as my personal interests—are in thinking about technology decisions in ways that are nuanced and responsible.
Henry Renski: My name is Henry Renski. I'm a professor of Regional Planning in the Department of Landscape Architecture and Regional Planning in the College of Social & Behavioral Sciences. I am halfway through my sixteenth year here at UMass. I have a number of different research strands; in regional planning I mainly focus on state and local economic development. In particular, I've always been interested in how technological change shifts the interaction between people and the built environment, as well as the types of work that people do. I’m really interested in how technology will qualitatively change the types of jobs that are available and the types of people that are likely to get those jobs or lose their jobs.
Shannon Roberts: I'm Professor Shannon Roberts, and I’m in the Mechanical and Industrial Engineering department. My research interests are focused on human factors in transportation safety. More specifically, how do we take information about human abilities and limitations and use that to design transportation systems and technology used in transportation systems to make sure that people are safe, that the systems themselves are efficient, etc.
Interviewer: Tell us about this research project and how it came together.
Laurel Smith-Doerr: I've always looked at inequalities, especially gendered and racialized inequalities, in studying socio-technical systems. And I'm especially interested in organizations, because race and gender are embedded in organizations and organizational practices, and technologies are also, of course, part and parcel of organizations. There are some pretty stunning examples of racialized bias in AI and automation that I learned about from colleagues of mine in sociology like Ruha Benjamin and Alondra Nelson. These are issues that need to be studied, AND we need an organizational sociologist to do this work, because it's not just about individual people's decisions, but how these systems get created and take on lives of their own. I’m interested in questions like: Who's designing the system? What are the imaginaries these designers have of users? How can we understand how the inequalities get built into the system?
This project idea came out of an NSF funded workshop where we had assembled roundtables of social and computer scientists and engineers. Everybody brought different perspectives to these questions about automation and it just kind of came together organically. From there, Shannon took the lead on the current NSF research grant because she had a really clear vision about the research that she wanted to pursue in automation and transportation, especially with long haul trucking, and I could see all of the things that I was interested in playing out in that space.
Shannon Roberts: I'd been at UMass for a year and was trying to start a research program. I knew I wanted to do something related to vehicles, and I knew I wanted to be more inclusive in my research. I try to look at the biggest problems in the field of automation and specifically automation in transportation. We know that some industries will get the technology first and some will get it last. Specifically, trucking will get it first and passenger vehicles will get it last. There's always been a shortage of truck drivers, but the COVID-19 pandemic made that unavoidably clear. The movement of goods has increased but the number of truckers has not. I think the pandemic kind of propelled companies to try to push automation technology faster in order to solve all these problems.
But at the same time, trucking is a job, and more importantly, for a good chunk of people, trucking is a way to move to a middle class lifestyle. So while automation is intended to solve some problems, it’s also creating new problems.
Henry Renski: My dad was a truck driver, and I come from a very working class background. So part of my interest in automation and the future of work comes from that very personal perspective of just wanting to understand how technology impacts people's ability to make a living.
First we got the NSF conference grant, and in that initial project I saw how different occupations were going to be impacted by automation and technological change, with a focus on racial and spatial impacts. But this analysis was at more of an overview level. So when Shannon identified this NSF Future of Work grant opportunity, for me it was an opportunity to take some of this high-level work that I've done and do a deep dive into one particular sector – trucking.
Working with Shannon, Laurel, and other members of the research team was an opportunity to not only think about this problem from my perspective of workforce development—but also to learn how other disciplines study the impacts of technological change. For example, Shannon is working with drivers in her simulations to understand how these drivers respond to different technologies. That's a different type of research than I've ever done. I really see this project as an opportunity for me to expand as a scholar—not just look at a new area, but new ways of doing research, and that was particularly appealing.
Laurel Smith-Doerr: Thanks to my colleagues in the Labor Center here at UMass, we were able to connect with the Teamsters, which is the major truck driver union. Using a participatory photo elicitation research method called PhotoVoice, the truck drivers we are working with will be able to show us what they think automation is going to be like. The data we collect will be their narratives and their photos, and we’ll be able to explore: What do they understand to be the future of their work? What does automation mean for them? And then we're going to do the same thing with the AI automation designers and engineers, and basically ask: What's in their imagination? Do they think automation will create more access to the work of trucking?
Shannon Roberts: Do they think about that at all?
Laurel Smith-Doerr: Right! There's a lot of discussion about regulation right now but I don't think there's as much of the kind of interdisciplinary data that we are collecting in this project. With this project we are putting together qualitative and quantitative research—and I'm hoping that we'll be able to have some informed conversations about the elements we need to be thinking about in decision making around these particular socio-technical systems. In this way, we're hoping that this project will have some impact on policymakers who are thinking about automation.
We're just at the beginning stages of this project, but I do hope that when I'm collecting information, whether it's from Teamsters, truck drivers, or engineers, with Henry’s insights I will be able to understand more of the regional economic impacts, and with Shannon’s insights I'll be able to understand more about the technical possibilities that are on the horizon. So I think each of us, in our own way, will have a more well-rounded understanding of the problem than we could ever have on our own.
Shannon Roberts: This project requires me to step out of my comfort zone of being engineering-centric. Working with Laurel, Henry, and our other collaborators, we are looking at this issue—which we all care deeply about—from our different perspectives, and are trying to help each other look at the balance of these issues together.
I think one of the things that we hope will come out of this project is an event which brings everyone together to discuss these findings. I think what we are seeing now, as Laurel said, is that there are silos. Sociologists, economists, and engineers are all working on this problem, but not talking to each other, and we see the same thing happening with truckers, developers, and regulators. One of the things we want to do is to bring all of these people together to discuss our findings and hopefully form a better collective understanding as a result. Because the grant funding for this NSF Future of Work project lasts four years, we of course will see changes in this space across the life of the grant. One of the impacts we want to see is just more collaboration and conversation across the different stakeholders that exist in this space. Because as far as we know, that is not happening.
Henry Renski: I work in a space with a lot of economists and planners, and initially there were a lot of doomsday projections about how everybody was going to lose their job due to automation. Over time, as people do more research, those predictions almost inevitably become a lot more nuanced as we develop a deeper understanding from actually talking to truck drivers and understanding what the technology companies are working on.
And so now I’m asking: What will be the different demands for truck drivers? What kind of workforce development programs do we need in terms of skills? Are companies going to combine with truck driving some of the tasks that used to be taken care of within different areas of the organization? Or are these tasks so fundamentally incompatible that those will always be separate things? But then, also thinking a little bit more broadly, if we have a lot more automated truck driving, that's going to change not only the fundamental nature of the way that we move freight, but also where we put the distribution centers, which may change where people decide to live and work. These are all important questions for policy.
I also have some specific aspirations for how this work will help advance my discipline. City and regional planning is supposed to be a very forward thinking, public interest-oriented discipline and profession. Within my discipline there's a lot of planning and discussion around very important topics like climate change, for example, but relatively few of us actually talk much about technological change. So this project provides an opportunity for me to advance a very important area of planning in the future, which is thinking about automation. The automation of driving in particular is going to change the way that people live and work in so many different ways.
Interviewer: What steps are you taking to try and ensure the ethical and equitable development of this project?
Henry Renski: I think that as social scientists we do have to be very conscientious about the process of collecting information, making sure that we're not only representing a lot of different viewpoints in our data but that we treat our subjects with dignity and respect. We're going to be talking to a lot of truck drivers, right? So we don't want to come in there and be like, “Oh, we're the smart academics.” That's not how you actually do research when you respect the people that you are collecting knowledge from.
Laurel Smith-Doerr: Our interlocutors are co-producers of knowledge, and that's why I'm so interested in the PhotoVoice method, which comes from anthropology and is more community engaged. It's about empowering the participants to make decisions about what is going to be collected, how it's collected, and what they share. All of those decisions are made together in an attempt to have these more power-sharing principles guide our own knowledge-building processes.
Interviewer: How is this project aligned with your personal understandings of “public interest technology”?
Henry Renski: In my domain, public interest is not an afterthought: it's baked into our DNA. As a planner, we basically do the things that create the background of everybody's everyday lives. We're also trying to think forward into the future to anticipate and offset the unintended consequences and side effects of technological changes. Technology has a major impact on people in their everyday lives: how people get to work, how stuff gets from one place to another, how people order stuff. Almost all of the questions that planners deal with are fundamentally of the public interest, so there's not as much of a kind of traditional separation between practice and theory as you see in some other disciplines.
I think that there's been a pretty seismic shift in my discipline within the last 10 to 12 years. In that time, we have taken on the challenge of being much more intentional about questions of distributional impacts. Historically in economic development and regional science, these questions were always thought of as somebody else's research question. We were just more interested in jobs: How many jobs and what type of jobs? But these questions about who technology is going to benefit and who it's going to hurt, these are distributional questions have really become central to my discipline. It's not that people didn't ask these questions before, but now it's a predominant theme of most of the research in my discipline. It's not enough to just say, “Well, we're going to create X number of jobs with development.” Now we also ask, “Who are the beneficiaries of this? How is this going to impact different groups of people?” And this, I think, is pretty central to the public interest.
Laurel Smith-Doerr: One way to think about public interest technology is a focus on more democratic distributions of power. And something I've studied for a long time is different kinds of organizations, and whether their structures or cultures are more or less conducive to equity and to democratic processes. Hierarchies are not good for equity, but horizontally connected kinds of organizations can be more conducive to democratic processes. But those horizontal structures are not very good for efficiency. There are pluses and minuses and I'm interested in trying to understand questions such as: Who is the organization developing this technology for? Who benefits, who loses, and who decides? So that's one way to think about public interest technology. But like the intersection of society and technology, PIT cannot be separated from trying to understand power and inequalities.
Shannon Roberts: In my field of human factors engineering, one of the tenets behind what we do is that people will use whatever it is that is being developed. So we need to consider the person that will use the technology. In that sense, it always has this public interest focus in that there will be a person using this, and so we need to think about the person and ask ourselves how do we do this equitably? Also, I feel similarly to Henry that within the past decade or so there's been a greater appreciation for DEI-focused efforts. While the focus has been on the person, the person, the person, more recently, we're saying, “OK, so there are actually different types of people.” And like Laurel was saying, some people benefit, and some people do not benefit when we make certain decisions. We need to consider all voices when we develop technology, not just convenience sampling.