Zhang's work seeks to address this challenge: I design tools to let people critically reflect, deliberate, and contextualize datasets underpinning AI, and then through case studies, examine how these can surface corresponding technological features, applications, and outcomes to (not) pursue. In this talk, Zhang will present three such tools to support human-centered AI with corresponding case studies. The first, “Data Probes”, are interactive visualizations using workers’ own data to help them contextualize data patterns to inform AI design. Rideshare driver-participants used these to suggest AI features in response to well-being trade-offs and AI behavior they identified from their data probes. The second, “Deliberating with AI”, allows users to create ML models from historical data. Using their models built on past master’s admissions data, participants discussed personal experiences and deliberated over how to fairly respect diverse values when designing future technologies for organizations. The third, “Contextualizing Datasets," guides users in critically exploring a civic dataset for opportunities and limitations. Students used the tool to reflect how contexts, such as data collection methods, can embed biases that data scientists should be aware of to mitigate unintended consequences of data-driven tools. Zhang will conclude with opportunities and challenges for future research to ensure critical and beneficial human-AI tools and collaboration. Overall, Zhang's work advances a vision of AI design that centers stakeholders’ realities and ideas towards responsible societal outcomes.

Angie Zhang is a PhD candidate and HCI scholar in the School of Information at the University of Texas at Austin. Her research seeks to advance human-centered AI design for inclusive outcomes, with a focus on workplaces and civic settings. Towards this, her work contributes empirical insights, tools, and methods to support stakeholder-centered AI design and inform tech policy. Her work has been published at top peer-reviewed venues such as ACM CHI and CSCW, and recognized by awards including three Best Paper Honorable Mentions and UT Austin Provost’s Fellowship. She holds a BS in Industrial Engineering from Georgia Tech and worked at Accenture prior to her doctoral studies.