Faculty Reflections: Augustin Chaintreau, PhD

October 29, 2020

This July, faculty and administrators participated in the Mini-Institute on Addressing Anti-Black Racism entitled, “Moving beyond Symbols and Performance: A Radical Reimagining of Antiracist Selves and Institutions.” The three 90-minute sessions were moderated by School of Social Work faculty Dr. Courtney Cogburn, Dr. Jalana Harris, Dr. Zuleka Henderson, and Dr. Ovita Williams.

Augustin Chaintreau, Associate Professor of Computer Science, was a participant in the summer pilot, and Jennifer Leach, Assistant Director for Faculty Advancement, spoke with him about his experience. Their conversation is below.

Tell me about yourself and your research. 

I joined the Computer Science department about 10 years ago. My research was in understanding how personal information is used online to improve commercial or public services. Since the beginning, I have been interested in understanding models of human behavior, because that's part of what the algorithm needs to be effective. My class on social networks describes algorithms and mechanisms that are at the core of this data revolution.

In some domains of computer science you keep the same problem formulation for decades, but the methodology, the algorithms, and the hardware changes every year. Here, the opposite holds: the algorithms and the mathematical foundation have stayed relatively the same, accelerating information flow between individuals. But they are completely rediscovered in a new light as we formulate new questions. And my class initially had a broad discussion of how algorithms mimic and learn from human behaviors, but without intervening with social science theory. Now, this theory has actually become an increasingly important part of the class: we have come to understand that hypotheses that originate in social science force us to challenge the simplicity of our assumptions. Algorithms have become interventions that shift long-standing social theory challenges on how power is distributed.

Augustin Chaintreau

Can you describe how you came to this place of engagement with #shutdownstem? How did that connection evolve?

It began in two ways. When I joined Columbia, the primary focus of my research community was the problem of privacy, the potential dangerous consequences of people and organizations knowing sensitive information about us.

For a while, my colleagues at Columbia and elsewhere worked on anticipating future threats. We asked which resources individuals have in our world where privacy is almost always compromised. You could be aware of what’s being done with your data: the current harm being perpetrated by algorithms. We saw that we needed to help people who are being marginalized today: women who do not receive ads for the highest-earning jobs with the same frequency as men, and Black, Latinx, and indigenous populations are systematically targeted by predatory use of keywords in search engines and in ad systems.

Our new mandate was to help the people who are fighting these abuses: journalists, nonprofit watchdogs, students, and activists. We want to prepare our students to do this work; it became a significant part of the design. Issues that were absolutely beyond my expertise and training had become integrated and embedded in my research.

How did you get from there, to now where you are with #shutdownstem? Can you trace the evolution?

Since I joined Columbia, I became increasingly aware that the field of computer science needed to broaden in terms of diversity. The first time I taught a social network class in the US, I had a tiny group with a majority of female students. I just assumed that that was the norm; then I realized it was actually something unusual. My lab is 50% female and male, if we use binary gender categories, and we have been helping to grow an online community called Mechanism Design for Social Good, a student-led organization funded by five female PhD students. There is a very strong female leadership in that community, so that has always been something that was very present in my lab.

And to be frank, that was pretty much it for diversity until this June. The only other work that I've been involved in that took a broader view was three years ago, when I had a great opportunity to meet Desmond Patton from the Data Science Institute. He was working on increasing diversity beyond gender, especially to Black, Latinx and Indigenous populations, and other people of color, to be represented.

We started a summer school with the help of the nonprofit association AI4All, and I was providing computer science expertise and access to the wonderful PhD students and undergrads that I knew. And that's how we got started, with this summer school following Desmond’s previous effort, but sort of making it more integrated between Computer Science and the School of Social Work.

Then in June, everything changed. Beforehand, our broad mission was to make tools that would provide better equity in how people are treated online. We had tested and found biased advertisements in different web search engines before, but it was not the main focus. And in June, I was just one member in a group of faculty who were receiving calls for action from the Black in Computing Association. The #shutdownstem movement made it clear: if you care, you can actually stop business as usual, and move beyond a sort of general sense that diversity will slowly increase with time.

Our first meeting was on June 10. It started as a simple Zoom call; I was setting aside the day to participate and thought that others might want to co-locate. Since that meeting, we have tried to translate that moment into actual actions. I am encouraged by what we heard from our colleagues at the School of Social Work at the Mini Institute: this is a once-in-a-lifetime opportunity, a time with a real potential for change. We have to take this very seriously, and commit to action.

People like me, white males who had all of the possible privilege in developing an interest and confidence in STEM, have not done our share in solving the problem.

Can you describe the format of the June 10 meeting?

We followed the examples of Black in Computing and #shutdownstem. There was a need for a new format. We didn't want to reproduce the traditional power dynamics of organizations with a real interest and dedication to the cause, but also demonstrate invisible bias in the way they approach the problem. We didn't want that to overwhelm the experience for everybody. We encouraged silence and a different form of interaction: simultaneously working on a Google Doc. Those who might not feel comfortable speaking could still contribute.

People like me, white males who had all of the possible privilege in developing an interest and confidence in STEM, have not done our share in solving the problem. When I attend a diversity-enhancing program, I see women overrepresented. I see people of color overrepresented. We have a special mandate: to participate more without reproducing the dynamics that give us more visibility, more influence, or more say about how things get done. This is the paradox that I'm trying to resolve. There is no easy solution; it has to come from being aware of both extremes: silence or listening, even in an active way, has the potential to stop a movement. And if you formulate recommendations and push things with a sense of urgency for results to make your institution or your lab look good, you might make progress, but at a cost.

What was the feedback from the people who participated in the June 10 meeting?

Many have been energized to act. Some faculty see this as a decade-long initiative: we have to put a committee in place and change our practice. We have to renew our commitment and listen more. There is a sense that we should be careful in how we proceed so that we sustain the effort, because news cycles are going to change and people are going to have other priorities.

I remember our colleagues from the Social Work Mini-Institute making it clear that this work is going to be a marathon, not a sprint, and we have to be prepared. We also have students who want to act more immediately, and we really cannot disappoint them by losing our momentum.

We are in the moment where the collision is at its maximum, and committees have been put in place across the university. At the same time, the natural law of diminishing attention makes many students feel like this moment is over, and they have not seen any change. This semester we need to critically demonstrate that there's a before and after; there are immediate effects, but we can also reflect and improve our toolkit to address structural racism.

What was your plan for the semester for the students that you have now? How was your first meeting with them?

On September 11, the first Friday of the semester, we organized a three-hour meeting, which was essentially a repeat of the June meeting. What emerged from our discussion is that we need more diversity among our faculty, but we also need to become good allies and listeners and supporters. These are not things that you typically put in an accountability system and we believe that's not going to happen from the top down.

This is why the organization of #shutdownstem is very loose. Anyone can join and we want to hear everybody. The email list includes colleagues from other countries and people at all career stages. It was very important to create a space where we could ground ourselves in listening to everyone, as opposed to the usual channels of power. We heard from students, postdocs, and others that we are late; it had already been three months, and they had been meeting. I read books, and I am sure that many other faculty also took time to educate themselves to feel better prepared. That’s good for us, but not for our students who feel like they have no sense of belonging to our field.

We need to change that message, and the time is now. This is a new semester and people are eagerly waiting to hear what we are doing. I've heard of internal discussions between the faculty, which are important, but these efforts didn't really reach anybody else. People don’t know if we are taking action. We will be judged this semester by how much we accomplish.

I think the most powerful experience was seeing faculty in the School of Social Work combining real, embodied experience, speaking in the first person, with a body of literature and theories that were very structured.

What is one first step that you feel doesn't require too much heavy lifting, that anyone. in any department, could do? What is a first step for someone who is considering this work?

I have heard from female students of color that bringing attention to injustice is important. One example was from a computer vision class. For a long time, there was a canonical female portrait used as the benchmark for testing an algorithm. That particular image perpetrated sexism and colorism and racism, and this was being discussed in the field. In this case, a professor just explained the story of the image, and said that they would not use it any longer, and they had adapted the curriculum accordingly.

This didn’t demand more additional effort, and the students in the class felt recognized and validated. I would imagine that many faculty could find an occasion to do that. Because our primary mission is teaching, this can have an important effect on our students. Actions like this started a movement; the field has begun to listen to the students.

One more example is that, on the first day of class, I always have a slide where I acknowledge the #metoo movement in academia and how every faculty member is actually mandated to help our students against sexual harassment. Having one slide to recap that reminds students that this is part of our job. You're not asking us a favor. We are not incredibly supportive and wonderful people because we say this. It is part of our job description, and we have to do that well. To some, it sounds obvious because we’ve been to faculty meetings where people discuss sexual harassment and express support, but it’s not obvious to see that support every day when you come to class. It doesn't take more than three minute, and it makes a difference to a student entering your class. Those are just examples to prove you can start small and of course, they are many other ways to be more involved.

What was your experience of the Mini Institute, and what are you taking away from that experience?

I think the most powerful experience was seeing faculty in the School of Social Work combining real, embodied experience, speaking in the first person, with a body of literature and theories that were very structured. We would have a conversation that wasn't intellectualized, or trying to put things in the abstract, in a forum that would allow everybody to relate personally to it, without ignoring the merit of translating lessons between domains, as most theory does. We can now cite a rigorous body of literature that helps us understand.

I saw this when one of the instructors was explaining the difference between positionality and identity and one person had a question. The instructor followed up with a remark by asking something very personal. That was exactly the right shift to make us reflect more deeply on the concept.

My main personal experience was a constant watching for when it was my time to speak. I have to acknowledge the fact that there is a body of evidence that males and white people statistically speak more often and are interrupted less. So that was a consideration during each 90-minute session, and, going forward, I will make sure this dynamic does not hurt people around me.

Yes. Does it seem like the sessions brought the implicit to the forefront, and the things that you would do in the back of your mind were now the topic of discussion?

Yes. I tried everything. I tried participating through the chat so that I would not interrupt the discussion. I emailed some people after the sessions. I wanted to convey my support, and not let silence be eloquent as it can often be. I realized that I cannot remain comfortable. This realization led me to read a lot more after the Mini-Institute. As scholars, we read articles all the time. We want the message compressed, and the articles that were selected were dense and thought-provoking.

This helped me to appreciate a broader literature, renewing my dedication to invest my time, technical skills and personal commitments to meaningful action. Unsurprisingly, as a white person I was unprepared, but my learning can help other people like me. I'm sure I'm like many other white people who realize, this is an historical moment and I need to do something. We need and can do more.

How did the Mini-institute help you transform your engagement?

It was helpful to connect this commitment to something personal. As a teenager in France, I learned from history class and from family stories about my ancestors who fought oppression. Broadening that lens made me more resolute that I could not just contemplate the question “What would I have done in the 1940s?” I am part of a research field that often exacerbates our society’s oppressions. Sometimes it is discouraging. We know that our ancestors, who fought structural racism in the courtroom, in the labor market, in the public discourse, all made a contribution. Those reflections probably seem self-evident, they are not original, and I had even personally heard them before, but it was a different experience to spend time up close with them. I admit, like most people working in research, I have been trained to be rewarded for “novel ideas,” and I need to counteract that when I work on any structural, long-standing challenge.

How can others engage with the work that you're doing, If they're not in your orbit?

We just discussed the Mini-Institute, which is a wonderful place to learn from the breadth of expertise of the Social Work faculty. I'm very inspired by maintaining connections with people from the Institute and hearing from people working on more racial equity in art, in the nursing school, in the medical school, and in other parts of the university.

Since computing has an important role in a lot of fields, I certainly hope that, in addition to making computing more equitable, true data science education reaches so many people, at least during one class of their training. I hope that if we are active, if we are vocal, students may be surprised to explore racial equity in our classes and find how relevant it is to understand what’s at stake. We need to do that in our computing class because the painful lessons learned from one domain can translate to another.

For faculty in other fields: If you are using data, and you feel there's something in your data that relates to equity and race, talk to me. Your inputs and shared practice can be how abstract principles about the importance of equity become real and unforgettable. Beyond data, there is an amazing new generation of students who are extremely prepared to surprise us.

Since we are in higher education. I do think that the ability to open conversations where you know you can be vulnerable, to talk to students about uncomfortable things, we need to make those conversations happen or create the conditions that will allow them to happen. As educators, we have this incredible opportunity to be quite independent, we can really present more than just content, we can pose questions that continue to resonate. We can do this in our domain of expertise, so that our message about equity comes from a credible place. Since the Mini-Institute, I have been able to start these conversations with students and that has been eye-opening.

We need to help students by connecting what they know to the body of work that we know.  Our main mandate is to be able to tell a student, “You can do that!” That student doesn't believe this yet because they have never done so before. It’s intimidating. But as a professor, I can guide them. Teaching is not reproducing ourselves in our students. It’s not even inspiring them by being good models. It's actually telling students that we have seen people exactly in their seat, who have made a difference and have made an impact. When we tell them “You’re next,” they can believe us.

This profile is part of a series entitled “Addressing Racism: Faculty Reflections,” which chronicles the narratives of faculty who are engaged in this work. Click here to learn more about the Addressing Racism initiative.