Natalie Buda Smith on AI in libraries, human at the center, deeper storytelling, and language recreation (AC Ep47)

2026-06-17 · Source: Humans + AI · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Data Science & Analytics, Cultural Heritage & Libraries · Depth: Intermediate, extended

Summary

Natalie Buda Smith, Director of AI at the Library of Congress, discusses how digital interfaces and AI are transforming access to human knowledge and cultural memory. She highlights the shift from direct primary source access to AI-intermediated information, emphasizing the importance of preserving historical context through multiple digitization versions, such as keeping 1970s scans alongside higher-quality modern ones. Smith addresses challenges like navigating proprietary data versus open APIs and the Library's approach of empowering staff with AI tools for tasks like language translation and transcribing handwritten historical texts. Collaborative, AI-powered projects, including the "Revolution Crossroads" initiative with the Smithsonian and National Archives, are surfacing new connections and enabling creative storytelling across diverse collections, like recreating a lost tribal language from an old Bible. The Library manages over 179 million physical items and provides public access to nearly 250 years of legislative history via APIs.

Key takeaway

For research scientists or AI/ML directors managing large historical or cultural datasets, you should prioritize human-centered AI deployment to empower staff and enhance public access. Consider making your digitized collections available via well-documented APIs to foster external innovation and deeper storytelling. This approach allows for personalized information delivery and the discovery of new connections, ensuring historical context is preserved even with AI intermediation.

Key insights

AI and digital interfaces are revolutionizing access to cultural memory by enabling personalized delivery and deeper contextual understanding.

Principles

• Preserve multiple digitization versions for historical context.
• Prioritize human-centered design in AI tool deployment.
• Public data accessibility fosters creative and research applications.

Method

The Library of Congress empowers staff with AI tools for translation and OCR, and collaborates externally to develop AI applications for large, diverse datasets, making them publicly available via APIs and platforms like Hugging Face.

In practice

• Use AI services to translate historical documents in unfamiliar languages.
• Apply AI for OCR on handwritten historical materials.
• Explore public APIs for large-scale data analysis and creative projects.

Topics

• AI in Libraries
• Digital Humanities
• Cultural Heritage Preservation
• Data Intermediation
• Public Sector AI
• API Accessibility

Best for: NLP Engineer, Director of AI/ML, Research Scientist, Domain Expert

Related on AIssential

• This Startup Is Using AI To Keep Oral History Alive - Black Enterprise — AI-powered platforms can effectively preserve personal and community oral histories, transforming loss into legacy.
• Office workers of the world unite: it’s time to revive the three-martini lunch | Andrea Javor — Authentic workplace connections, crucial for human interaction, are eroding amidst AI-driven productivity demands and digital communication.
• Chicago Public Media on Building AI Tools That Serve the Audience: In Conversation with Melissa Bell, Aron Pilhofer, Mark Chonofsky & David Chivers — Public media organizations are developing AI tools to directly serve audiences, prioritizing ethical implementation and engagement.
• AI is reorganizing how authority, workflow, attention, and distribution power flow through scholarly communication, and most publishers are responding with the tools... — AI's impact on scholarly publishing is structural, shifting value to inference-time context and audience ownership.
• How Preply combines AI and human tutors to personalize learning — AI can augment human expertise by automating repetitive tasks, enabling deeper personalization and continuous learning experiences.
• A Need for Nuance: The Economist’s Andrew Palmer — Successful AI adoption requires open-minded experimentation, human oversight, and a nuanced understanding of its organizational impact.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Humans + AI.