Format: Invited Talk
Speaker: Melissa Terras
This talk describes how READ-COOP SCE built and runs Transkribus, the AI platform for extracting information from historical texts. Transkribus now has 279 cooperative members and 550,000 users, and it has read more than 210 million historical documents. READ-COOP is a working, full-stack AI cooperative. Its members own the platform, software, and data and have a say in how it develops. That makes it a concrete example of Solidarity AI in practice. The talk will cover how the cooperative formed, how members govern a technically demanding AI service, and how it sustains itself financially. It will also reflect on what other communities can learn from READ-COOP if they want to build and own AI infrastructure themselves rather than rent it from large tech firms.
Reporter 1: Kanikka Sersia
Reporter 2: Cindy Nguyen
Speakers
-
Melissa Terras Professor of Digital Cultural Heritage at the University of Edinburgh