Files and links (2)
PDFOpen Access (License Unspecified), Open Access
Thesis
Toward an AI representation of Moses Williams at Peale's Philadelphia Museum
Master of Science (M.S.), Drexel University
Sep 2025
DOI:
https://doi.org/10.17918/00011168
Abstract
This thesis project investigates the potential and boundaries of employing artificial intelligence (AI) for the ethical recreation of the life and voice of Moses Williams (c. 1775-1825), an African American silhouette artist whose biography survives primarily through Charles Willson Peale's Museum records. Taking inspiration from works that draw on critical traditions which rethink history and methodologies that blur the line between scholarship and creative practice, the study situates itself within ongoing conversations about race, history, and representation. It examines how a conversational AI can center marginalized voices while acknowledging the gaps and prejudices of the archive. A specially assembled dataset was constructed from authenticated historical documents, museum reports, and ethically framed speculative analyses. This dataset was used as the basis for training and developing the Moses Williams conversational AI on the Convai platform. The system was first tested with a subject matter expert and later judged by a panel of five scholars and practitioners in African American history, museum studies, education, and cultural heritage interpretation. The panel found the conversational AI effective in highlighting archival bias and sparking engagement. However, they also raised concerns about oversimplification, modern phrasing, and limited context on Williams' early life. After revisions, the panel expressed satisfaction with the improvements, noting the model's potential as an educational tool when paired with transparency about speculation and archival gaps. The outcome illustrates that AI can serve as an effective process for recovering and transmitting marginalized histories--but only within the limits defined by the quality, scope, and bias of its source material. Without active counter-archival strategies, such systems risk reinforcing prevailing narratives. This research proposes an interdisciplinary approach for building ethical AI heritage projects, emphasizing transparency, clear marking of speculative content, and continuous expert consultation. The model can be adapted for other AI reconstructions in cultural heritage while encouraging critical reflection on both the technology and the histories it attempts to represent.
Metrics
2 File views/ downloads
14 Record Views
Details
- Title
- Toward an AI representation of Moses Williams at Peale's Philadelphia Museum
- Creators
- Aman Chandre
- Contributors
- Glen Muschio (Advisor)
- Awarding Institution
- Drexel University
- Degree Awarded
- Master of Science (M.S.)
- Publisher
- Drexel University; Philadelphia, Pennsylvania
- Number of pages
- xi, 105 pages
- Resource Type
- Thesis
- Language
- English
- Academic Unit
- Digital Media; Drexel University; Antoinette Westphal College of Media Arts and Design
- Other Identifier
- 991022093155604721