The Stanford University Center for Human-Centered Ai has just published an important paper titled “Societal Implications of Open Foundation Models” which covers a number of topics but also helps to clarify questions about what constitute “open” foundation (or large language) models, which is a very vexing question; the intertwining of code, process and data leads to very different ideas about what it means to be open when compared to historical open source community practices. Strongly recommended.
Announcement of paper
https://hai.stanford.edu/news/societal-impact-open-foundation-models
Summary of paper
https://crfm.stanford.edu/open-fms/
and the paper proper
https://crfm.stanford.edu/open-fms/paper.pdf
Clifford Lynch
Director, CNI