Research philosophy

Analyzing the impacts of South African apartheid using computer vision techniques and satellite imagery. Read our NeurIPS paper and this MIT Tech Review article, and check out our dataset and visualizations.

Analyzing the history of Black protests in the US and Canada using machine learning methods. Read our papers on Sociological Science and Mobilization to learn more.

Creating natural language processing tools, and using qualitative methods to analyze the impacts of social media platforms on neglected countries and languages (more soon). Read Dylan's op-ed in Tech Policy Press and his guide to social media data collection for researchers.

Centering the experiences of data workers in machine learning by using the Workers' Inquiry approach to uncover opportunities for collaborative knowledge exchange and collective organizing. Go to https://data-workers.org to find zines, reports, documentaries, podcasts and more created by 16 data workers from 4 continents.

Creating a wage-theft calculator to estimate how much workers are owed in lost wages that are stolen with the assistance of surveillance technologies. Watch Adrienne's talk.

Showing the extent to which vulnerable populations like refugees are both exploited laborers fueling “AI” systems and those most harmed by them. Read our papers and articles to learn more.

Educating the public on the harmful ideologies driving the current race to build Artificial General Intelligence (AGI). Read our paper on the topic, and find more resources and articles on our TESCREAL page. Thumbnail is Clarote & AI4Media / Better Images of AI / Power/Profit / CC-BY 4.0.

Using satire to expose the hype surrounding AI systems and the harm this causes in many sectors (see Mystery AI Hype Theater 3000 Series, Scientific American Op-ed).

Examining who benefits from "AI" systems and who is harmed (e.g. read AI for Whom?, AI Art and Its Impact on Artists, AI and Inequality in Hiring).

Guidelines for dataset annotation, documentation and release practices. Read Datasheets for Datasets, Dataset Accountability, Participatory Data Work, The Subjects and Stages of AI Dataset Development: A Framework for Dataset Accountability, to learn more.

Read Algorithmic Tools in Public Employment Services and Recommendations for Appropriate Gender-Aware Information Access, to learn more.

What does it mean to conduct AI research which works for communities? Read DAIR's Research Philosophy to learn more.

Broad, near-future speculative pieces designed to de-center dominant narratives about technology and challenge us all to realize that things can be different. These are not alternative realities, they are possible futures. Read current entry.

We have partnered with lesan.ai and Whose Knowledge? to work on language tech in a manner that does not steal data from communities and supports locally relevant content creation (read how Lesan outperforms Google, and this Deutsche Welle article on the AI language gap).

Counteracting the "one giant model for everything" approach by BigTech, by empowering small, community-rooted organizations around the world building task specific models serving their needs. Read this interview and our paper.