Hi, I'm Shaomei Wu. I study, design, and build technologies with marginalized communities to dismantle barriers and maximize public interests. My work takes place at the intersection of community, technology, and advocacy. I am particularly interested in understanding and challenging structural marginalization materialized through pervasive technologies such as AI and social media.
I am a fan of Afsaneh Rigot‘s Design From the Margins approach and frequently shift my focus between community research, intervention development, and social advocacy, depending on the needs of the community and the stage of the engagement. My current work looks into speech diversity and communication technologies, investigating the experiences and challenges of people who stutter with telecommunication technologies such as videoconferencing tools and speech AI technologies, and co-developing community-centered solutions to address these challenges.
My work has been supported by NSF's Responsible Design, Development, and Deployment of Technologies (ReDDDoT) program (Award #2427710), Patrick J. McGovern Foundation, Borealis Philanthropy’s Disability Inclusion Fund (DIF), Mozilla Data Collective, as well as small, individual donations from the community.
To learn more about my prior and ongoing projects, please check out my CV, my list of selected publications, my work at AImpower.org, and my Google Scholar Page.
Bio:
Shaomei Wu is a person who stutters and the founder and CEO of AImpower.org, a tech nonprofit that researches and co-creates empowering technologies for, with, and by marginalized communities. As a technologist, researcher, and disability advocate, she brings together lived experience, technical expertise, and community knowledge to advance more accessible, equitable, and empowering AI systems, with a particular focus on disability-affirming speech and communication technologies.
Previously, Shaomei was a staff research scientist and Tech Lead at Facebook and Instagram, where she led research and product initiatives serving underserved communities, including automatic alt-text, an AI-powered service that describes billions of images on Facebook and Instagram to screen reader users each day. Shaomei holds a Ph.D in Information Science from Cornell University, where she studied the spread of information and behaviors in large social networks. Her current research interests lie in AI ethics, accessibility, human computer interaction (HCI), and speech science & technology. Shaomei frequently publishes at leading academic conferences including CHI, CSCW, FAccT, ASSETS, and InterSpeech, and serves on program committees for NeurIPS, CHI, CSCW, FAccT, and AIES. As a stuttering advocate, Shaomei regularly organizes and speaks at technical and community conventions on disfluency-friendly speech and communication technologies, and advises stuttering community organizations on technical issues and strategies.