Tag: archives
-
XML Coding with AI, Episode 3: Return of the AI
This blog post discusses the author’s integration of AI, specifically a custom GPT termed the EAD XML Architect, into XML coding for cataloguing. The author emphasizes the advantages of AI in ensuring accuracy, identifying links in data, suggesting improvements in written work, and serving as an effective final draft editor despite its inconsistencies.
-
XML Coding with AI, Episode 2: The Hallucinations Strike Back
This blog post discusses the challenges faced while using generative AI for XML cataloguing, particularly the issue of ‘hallucinations’ where AI fabricates information. Despite initial success, the AI generated errors and unreliable outputs when processing large datasets. The author shares strategies to improve output accuracy and emphasizes the need for vigilance in verifying AI-generated data.
-
XML Coding with AI, Episode 1: A New Hope
At the University of Manchester Library, the process of creating XML files for archives is labor-intensive and context-dependent. The author explores using AI tools, particularly ChatGPT, to automate and improve this task. Initial tests show promise, significantly reducing the time to generate usable XML files, though challenges remain.