إطلاق مبادرة البيانات المؤسسية (IDI) في جامعة هارفارد: سد الفجوة بين المؤسسات المعرفية والذكاء الاصطناعي

The Harvard Law School Library has launched the Institutional Data Initiative (IDI), a research initiative dedicated to supporting knowledge institutions in making their data collections accessible for artificial intelligence (AI) development. This initiative aims to bridge the gap between libraries, universities, government agencies, and the AI community by structuring, analyzing, and publishing institutional collections as data. The IDI believes that increased access to data from these institutions will lead to more diverse, inclusive, and safer AI systems.

The Core Mission of IDI

The IDI’s primary goal is to empower knowledge institutions to provide broad access to their vast and unique collections, much of which remains untapped. By collaborating with the AI community, IDI seeks to make this information accessible for traditional patron access and AI training, recognizing the expertise that knowledge institutions possess in areas such as information classification, cultural context, and privacy frameworks. This collaborative effort aims to align the interests of AI model developers with the goals of knowledge institutions.

Key Initiatives and Projects

The IDI is embarking on several key initiatives to achieve its goals:

• Public Domain Book Corpus: Refining and releasing nearly one million public domain books scanned at Harvard Library as part of the Google Books project.
• Historical Newspaper Collaboration: Partnering with the Boston Public Library to make millions of pages from hard-to-find historical newspapers accessible. This involves employing new methods to extract text accurately from newspaper layouts, which are notoriously difficult to process.
• Symposium: Hosting a spring symposium at Harvard Law School to connect institutional and AI communities, fostering collaboration and exploring alignment.
• Expanding Data Types: Actively seeking collaborations across all forms of data, including scientific and biomedical data.

Benefits for the AI Community

The IDI anticipates significant benefits for the AI community, including:

• Lower Barrier to Entry: Increased access to data will reduce the barriers to entry for model creation, allowing more diverse groups to participate in AI development.
• Improved Representation: Greater language and cultural representation in AI models, enabling them to serve a broader range of users.
• New Capabilities: Opening doors to new scientific and medical discoveries through access to specialized knowledge.
• Safer and Transparent AI: Contributing to the development of safer and more transparent AI systems through access to curated and stewarded knowledge.

Community Engagement and Support

The IDI actively seeks participation from various stakeholders:

• Knowledge Institutions: Encouraging institutions to share their challenges and needs in making their data accessible.
• AI-ML Researchers: Inviting contributions to ongoing projects and research.
• Academics: Welcoming integration of work and perspectives from digital humanists and other scholars.

The initiative is generously supported by gifts from Microsoft and OpenAI, with ongoing efforts to build a diverse base of philanthropic and industry support.

Conclusion

The Institutional Data Initiative represents a crucial step in bridging the gap between knowledge institutions and the rapidly evolving world of AI. By fostering collaboration and increasing access to valuable data collections, the IDI aims to ensure that AI development is grounded in the depth, diversity, and complexity of human knowledge. This initiative has the potential to unlock new opportunities for innovation, promote inclusivity, and shape the future of AI for the benefit of all.

Source: Harvard Law School Library