News Exploring Structural Oppression in Digital Spaces Ph.D. student Avriel Epps studies how bias in the digital world impacts users across diverse backgrounds Posted August 1, 2023 By Ryan Nagelhout Diversity, Equity, and Inclusion Technology and Media Photo: Courtesy of Avriel Epps Finding your purpose in life is a powerful thing. Getting there isn’t always easy but, once a person discovers what truly motivates them, the most driven among us find every possible avenue to fulfill that purpose.For Avriel Epps, being uniquely aware of how physical and digital environments impact our lives has inspired a unique educational journey. A doctoral candidate at Harvard Graduate School of Education, her research has explored how the often unseen world of machine learning, artificial intelligence, and algorithms impacts the people who use them online. Especially when many of those human-created technologies carry biases and structural oppression from the real world into digital spaces.“What motivates me is I really want to figure out this complex issue: How do we combat or dismantle structural oppression in technological tools?” says Epps, who is working toward her Ph.D. in Human Development, Learning, and Teaching. “You have to look at it from a million different perspectives and you have to take on a million different skills because the issue is just so complex.”Gaining the skills to identify and combat those systemic biases means experiencing them firsthand. Epps grew up in Los Angeles, attended public schools and spent her childhood building a career in entertainment. She started acting at age 2, modeled, and worked as a voice actor, bringing Timberly Johanssen to life on the 90's Nickelodeon cartoon Hey Arnold! As an R&B artist, she was making music as King Avriel.Epps remembers her interest in education started with a presentation from education professor Tyrone Howard while attending a summer program at UCLA for community college transfers. When Howard presented some sobering numbers about how few Black and Brown students matriculate from community colleges into the University of California system, things clicked into place.“That was the first time anyone had explained educational inequality based on race to me, and it provided a coherent explanation for my experiences," says Epps, who despite having taken music theory classes and working as a recording artist failed to get attention from more established music programs. “I now had a name and a lens for understanding for what I had experienced — structural racism — and it really blew my mind.”Epps went on to earn her bachelor’s in communications at UCLA, studying racism and sexism in the music industry. Her work, for example, revealed that hip hop music promoted by record labels often was more violent than the songs that gained popularity organically online. Much of what she learned helped her further reflect on her own experiences, and she discovered a passion for battling educational inequality and helping young people, especially people of color, develop as artists and students. Her goal now, Epps says, is improving the online experience for everyone by giving those most vulnerable the tools to overcome systems not designed with them in mind.“I went through adolescence experiencing racism and sexism but didn’t necessarily have the framework, the tools to understand and heal from my experiences,” Epps says. “How can I prevent that from happening to other kids of color?”It’s an instinct that’s served her well as both a researcher and educator, studying machine learning while also designing and teaching courses based on her own work. That also means working past the established rhetoric about how to diversify tech spaces. It’s not just enough to say more women or Black people should be involved in STEM, for example, if the structures that create the internet and the systems that rely upon it are built using the same status quo that’s created other forms of structuralized racism.“The experiences of marginalized people using these products isn’t centered so there are all kinds of problems when we use them that may or may not be taken seriously by the people who develop the tools,” explains Epps. “I think trying to implement all these proposed solutions developed by people who don’t have to actually experience the barriers has been unsuccessful because it’s clear that the issues run deeper. The narrative that marginalized people are not striving hard enough is false.”Much of Epps’ career has seen her move between worlds. While her music streams on Spotify she’s also worked there, taking on a visiting research fellowship to study how the company’s algorithmic recommendation system selects artists based on gender. She’s researched how people of color are impacted by digital environments and are shaped by algorithms and discovered the real-world challenges behind the scenes with funding and developing projects as a woman in tech. Epps characterizes herself as both an insider and an outsider, taking knowledge from one world and applying it to another. “I think it goes back to being a biracial kid and being the darkest kid in a family of mostly white-passing people,” said Epps. “I always kind of jump in and out of being the subject and the object, the insider and the outsider. And so, it’s always been in my nature to try to understand things from both perspectives. I’m just drawn to doing that.” News The latest research, perspectives, and highlights from the Harvard Graduate School of Education Explore All Articles Related Articles News Computing with Creativity Paulina Haduong, Ph.D.'23, on understanding the challenges that K–12 teachers and students face in learning computing together News The Human Element of Data and AI Gahyun Callie Sung's journey to HGSE and the LIT Lab is reflected in her research into data and using AI to improve student outcomes Ed. Magazine Students: AI is Part of Your World Alum helps young people understand how artificial intelligence is changing everything they know