Mentoring Partners Explore Racial Justice—and Artificial Intelligence
Devyn Brown, an AIR Opportunity Fund Pipeline Partnership Program (P3) Fellow, is completing her Ph.D. at Howard University and, until recently, worked as a research fellow at The Sentencing Project. Her academic and professional lives intertwine: She is laser-focused on life reentry among formerly incarcerated people, particularly African American adult males.
AIR Scholar Odis Johnson Jr., Bloomberg Distinguished Professor of social policy and STEM equity at Johns Hopkins University, is a national authority on education and racial justice who uses artificial intelligence (AI) to conduct research. At the same time, he recognizes that algorithmic bias has profound, potentially life-threatening implications. He is leading efforts to address fundamental questions about how AI works and to advance AI “design justice.”
Through P3, Devyn and Odis have developed camaraderie in their mentoring relationship. In this conversation, they discuss the power of mentorship and their commitment to racial justice, including in the realm of AI.
Q. Devyn, why did you decide that your dissertation topic and career pursuit would be African American males returning from prison?
Devyn: I am extremely ambitious and passionate about this population of individuals. I’m from California, where we have a lot of gang presence, activity, and violence in impoverished neighborhoods. Many of my friends who grew up in family households with limited financial resources got trapped in that cycle of gang activity. Some of my male friends were incarcerated during our teenage years.
At the time, any individual who was suspected of being an affiliated gang member could receive a sentence three to five times as long as the regular sentencing requirements. As a result, a lot of my friends were looking at 20-plus years. Some received life sentences. Then the laws for youthful offenders changed in California, meaning that an individual under age 25 at the time of their offense could receive reduced sentences. Even before I applied to the Ph.D. program, I started looking into reentry models and different ways that we could work to support them. I’m so happy that of the three of my friends who sparked this motivation, two have been released from prison and are doing very well for themselves in the community.
I’ve been very intentional about applying for opportunities that aid me at this stage of my educational career. At The Sentencing Project, I was working alongside researchers whose work I’ve been citing for years.
Q. Odis, why is it important for you to make time for a sustained mentoring relationship?
Odis: In part, it’s because I’ve benefited from great sponsorship and advice. With sponsorship, I had professors who were there to write that letter of recommendation and talk with me about opportunities. You need that someone who can open doors. But sponsorship and advice are much different than mentorship, which I would define as technical assistance, skill application, and actual knowledge related to how people do their work. A lot of times that happens through training, right? Your doctoral students are there with you on your research project. They observe how you conduct your research. They learn about particular practices, procedures, and methods. Their horizon about how to conduct themselves professionally is expanded.
I did not have that as a doctoral student. I was not on teams led by faculty. I was never invited to coauthor with my professors who were on my dissertation committee. So, I seek to be that presence, especially in the lives of minoritized scholars, because I did not have this. I had the “sink or swim” model. I think that scuttles a lot of human potential, and it is no good for any of us. That’s why I do this, especially for scholars who are, in turn, likely to do the same. And I think Devyn is one of those people.
Q. Devyn, how has this mentoring relationship helped you make progress toward achieving your goals?
Devyn: This mentoring relationship has been so significant for me, not only in the knowledge and skills I’ve acquired, but in the humanistic aspect of our relationship. Like Dr. Odis mentioned, oftentimes doctoral students, especially minority and female doctoral students, can feel like our passions, our interests, are not really valued by our committee members. It’s more of a formal process of “let’s get them through.” With Dr. Odis, he’s very easy to talk to. He’s always available to me and he’s extremely knowledgeable. This makes me comfortable seeking advice, talking with him about challenges and successes, and learning skills that he is very well versed in and that I so desperately want to learn.
It’s been extremely rewarding and necessary for me, especially where I am with my degree. Imposter syndrome is real. I’m a first-generation college student and graduate out of hundreds of family members. I certainly can’t look within my family networks to help with, say, regression analysis. Dr. Odis just naturally, genuinely, and authentically allows me to show up as I am and provides the coaching, mentorship, and support for me as a human, as an individual.
Sponsorship and advice are much different than mentorship, which I would define as technical assistance, skill application, and actual knowledge related to how people do their work.
- Odis Johnson
Odis: I, too, am a first-generation college-going graduate. We share that. And I would say the work Devyn is doing is groundbreaking, timely, and important. We have had, in our past, mass incarceration that disproportionately impacted people of color. Compounding that crisis is the fact that young men of color are having those experiences far too soon. This work also aligns with my thinking about how conflated our social worlds are. We don’t experience just schooling, neighborhoods, families, or other institutions, as well as mental health and wellness. They are all interrelated, and anyone who can think in an interdisciplinary or perhaps transdisciplinary way on how to address these issues is on the right side of the historical bend when it comes to science. It requires a different type of training that I’m so happy AIR is providing to people like Devyn.
Pursuing Racial Justice with AI
Odis Johnson believes that AI will continue to be “an innovative and generative space to advance human capacity and technology.” His own research, however, lays bare AI’s biases and potential consequences. Most recently, in an AIR Opportunity Fund Distinguished Lecture Series presentation, he concluded that the increasing use of AI in school surveillance systems may amplify the probability that Black youth will enter the school-to-prison pipeline, due to racial biases in these technologies.
Johnson is on a mission for design justice, which he defines as “vigilant commitment to algorithmic fairness and equitable exposure to AI” and “human-centered approaches” to the design of AI, and related technologies, such as machine learning, infrared technologies, and robotics. He cited as examples pulse oximeters that measure blood oxygenation and AI detection systems in autonomous vehicles that are meant to recognize pedestrians.
“We’ve known for a long time that such devices do not interact with or recognize melanated skin in the way that they do nonmelanated skin, and therefore the actual biometrics will be off for people of color,” Johnson said. “This can have a disparate impact with profound, life-threatening and wellness implications.”
Some work on design justice is underway now through the Institute in Critical Quantitative, Computational, & Mixed Methodologies (ICQCM) at Johns Hopkins, which Odis founded in 2019 with support from the National Science Foundation. “ICQCM trains the next generation of scholars, particularly those in minoritized institutions, in the appropriate use of transformative quantitative and computational methodologies and their integration into mixed methods,” he said.
For her part, Devyn Brown expects to use AI when she begins her dissertation, but she’s not sure yet exactly how. With Odis as her mentor, she’ll be prepared.
