We are living as data bodies, producing continuous streams of information that shape how AI systems model the world. Yet most machine learning pipelines are trained on global data that is unevenly distributed, linguistically imbalanced, and structured by historical relations of power. As a result, these systems often flatten the cultural, linguistic, and historical conditions that give data meaning.

In this context, questions of representation take on new urgency. What forms of knowledge become legible within AI systems, and what remains obscured? How do computational models register difference across languages, histories, and unequal relations of power? This talk approaches these questions by developing ways of reading computational systems against the grain—attending to what is excluded, misrepresented, or rendered invisible within dominant data infrastructures.

Drawing on the concept of the contrapuntal—understood here as a way of holding multiple, uneven histories and perspectives in relation without resolving their differences—Dr. Shereen Sakr traces a research trajectory from R-Shief, a multilingual social media archive spanning more than seventy languages, to a new AI platform, Contrapuntal, alongside an expanding body of publications and creative scholarship.

The result is a reframing of AI: from prediction over data to an infrastructure for interpreting relations within uneven global knowledge.

Laila Shereen Sakr (VJ Um Amel) is an artist and scholar whose work merges digital arts and media, feminist decolonial thought, and critical AI to interrogate power, memory, and technoculture. Through her long-running R-Shief media system and exhibitions such as Capital Glitch  (Qualcomm Institute, 2021), Dis-Assembling the Cloud (Lagos Biennial, 2024), and Rosetta Stones Resurrected (Diriyah Art Futures, 2025), she explores data as material and metaphor, using algorithmic disruption to reveal global entanglements. Her book Arabic Glitch: Technoculture, Data Bodies, and Archives (Stanford University Press, 2023) situates Arab digital cultures at the vanguard of media theory. Sakr is Associate Professor of Film & Media Studies at UC Santa Barbara, where she co-directs Wireframe Studio and co-founded the Creative Critical AI Undercommons. She is also a faculty affiliate in the Art Department, Feminist Studies Department, Media Arts and Technology Department, Center for Responsible Machine Learning, Center for Information Technology and Society, and the Center for Middle East Studies, and has co-founded the Network of Arab Women in AI (2023), Autonomous Futures (2024), and the D.C. Guerrilla Poetry Insurgency (2003).

Her art and research cultivate transnational, feminist, and community-authored approaches to technology and culture. As VJ Um Amel (moniker or video jockey “Mother of Hope”), she combines artistic innovation with critical inquiry, using data like clay to transform how we understand our contemporary worlds and future possibilities. As both an artist and a scholar of emergent media, she writes prolifically in venues such as Minnesota University Press’ Debates in Digital Humanities series and Middle East Critique, develops machine learning (ML) software and natural language processing (NLP) analytics for social media, and is developing an Arab futuristic video game about liberation. Her work has been shown in venues such as the SF MoMA, National Gallery of Art in Jordan, Camera Austria, Cultura Digital in Brazil, Kirchner Cultural Centre in Argentina, Tahrir Cultural Center in Egypt, Lagos Biennial in Nigeria, Qualcomm Institute in San Diego, and Diriyah Art Futures in Saudi Arabia.