Creating guidelines for designing AI systems that center the needs of specific communities.
Our Research
We provide a number of guidelines and frameworks informing AI researchers, practitioners and adopters on how to center the needs of various communities while designing, developing and deploying AI systems.
Instead of building something akin to an everything machine, we advocate for building small, task-specific AI systems that center the needs of specific communities. Those aiming to build or adopt any type of technology should not do so before conducting thorough needfinding activities and ascertaining that what they aim to build/deploy addresses the identified needs.
Image: Nidia Dias & Google DeepMind / https://betterimagesofai.org / https://creativecommons.org/licenses/by/4.0/
How to avoid harms perpetuated by computer vision systems and build community-rooted models instead. Tailored to researchers and practitioners in computer vision.
Three design considerations for building data-driven and algorithmic public employment services that center the needs of jobseekers.
Recommendations for ethical and well-grounded use of gender in building and researching information access systems.
Pioneered in the 1990s, value sensitive design seeks to provide theory and method to account for human values in a principled and systematic manner throughout the design process. Central to this approach is engaging our moral and technical imaginations.
An exploration of how design might be led by marginalized communities, dismantle structural inequality, and advance collective liberation and ecological survival.
"The Community Partnerships Playbook aims to provide some answers to such questions as: How can technologists and scientists engage communities in a spirit of partnership, without such extractive practices? How can community organizations work with researchers in ways that benefit their communities and expand their capacity, rather than burdening their staff?"
