Reframing "Frugal AI"

Dr. Timnit Gebru is the founder and executive director of The Distributed AI Research (DAIR) Institute, an independent organization conducting community-rooted research. Prior to that she was fired by Google in December 2020 for raising issues of discrimination in the workplace.

‍

In this interview, Dr. Timnit Gebru critiques the dominant "one giant model" paradigm in AI, which prioritizes massive, ill-defined models and creates new problems in systems that previously served more specific purposes. Gebru argues that this approach stifles innovation around potential resource-efficient approaches and results in subpar tools due to poorly defined tasks. The approach is in direct opposition with the engineering principle of building specific tools for specific contexts. She advocates for stitching together “frugal AI” efforts, highlighting localized, community-rooted organizations – like those focused on low-resource languages – that curate data and use smaller models. She proposes that these organizations federate their resources and tools to collectively challenge the monopolistic power and resource-intensive practices of Big Tech.