The artificial intelligence cooperative: READ-COOP, Transkribus, and the benefits of shared community infrastructure for automated text recognition

The research article titled "The artificial intelligence cooperative: READ-COOP, Transkribus, and the benefits of shared community infrastructure for automated text recognition" is published by Open Research Europe, a platform that supports open access to research. The authors include Melissa Terras, Bettina Anzinger, Paul Gooding, and several others from various institutions across Europe, focusing on the development and governance of artificial intelligence tools in the cultural heritage sector.

READ-COOP is a platform cooperative that emerged from European Commission-funded projects, aimed at creating and maintaining automated text recognition tools like Transkribus. As of October 2024, READ-COOP has 227 members from 30 countries and 235,000 registered user accounts. The cooperative model promotes democratic decision-making and community involvement, leading to sustainable growth and stakeholder engagement.

Transkribus has processed over 90 million digital images of historical texts, showcasing its effective use of AI technology in cultural heritage. The platform has won the European Union’s Horizon Impact Award in 2020 for its contributions to the field. The cooperative approach allows for equitable ownership and ensures that all earnings are reinvested into improving the infrastructure and services provided.

The cooperative structure facilitates high levels of member satisfaction and engagement. Qualitative feedback indicates that members appreciate the integrity and utility of the AI infrastructure, which supports innovation and technology development. Members have the opportunity to participate actively in decision-making processes, emphasizing the importance of community input in the cooperative's governance.

READ-COOP employs environmentally friendly practices by sourcing all electrical power for its administrative activities and computational processing from renewable resources. This commitment aligns with the cooperative principle of caring for the community and addressing the environmental impact associated with AI technologies.

The article suggests that cooperative models may be particularly suitable for developing AI infrastructures initially funded through public grants. This approach can provide a sustainable transition from public development to long-term community ownership. The authors advocate for the broader application of cooperative models in AI and machine learning technologies to ensure responsible creation, governance, and use.

READ-COOP serves as a pioneering example of how cooperative governance can facilitate innovation and sustainability in the AI sector, particularly within the cultural heritage domain. By embedding community engagement and democratic governance into its operations, READ-COOP exemplifies an alternative framework for building trustworthy and responsible AI technologies that meet specific societal needs.