MCN 2024

The recent advancement and integration of AI’s advanced language models (LLM) in higher education, culture, and art institutions offer transformative potential for museums and galleries with permanent holdings of art and material culture items. This session explores an innovative project undertaken by Mead Art Museum at Amherst College, MA, to develop and record audio descriptions for its online collection using ChatGPT 4, when it became available in 2023.

The project required a structured methodology that combined AI's impressive efficiency with diligent human oversight. The museum envisioned this initiative to primarily enhance its digitized collection's accessibility for visually impaired visitors, but the benefits and impact have proven to be multi-dimensional. The Mead uniquely engaged undergraduate students in the roles of reviewers, editors, researchers, and narrators, empowering them to refine AI-generated content, check fact accuracy, and adjust the composition and style of the descriptions according to the pre-set guidelines. By increasing online discoverability and providing experiential learning opportunities, the outcome proved to be beneficial for both the museum and the students and faculty involved.

By presenting this case study, I aim to highlight the innovative uses of AI in museum collections and provide valuable insights into the collaborative possibilities between museums and institutions of higher education.

I hope to inspire other museums and educational institutions to explore similar collaborative initiatives with far-reaching impact, enhancing learning and accessibility across their campuses, constituents, and communities.