Skip to Main Content

Artificial Intelligence

This subject guide provides information and suggested readings and resources on the topic of Artificial Intelligence

AI and Libraries

Artificial intelligence technologies present both opportunities and challenges to libraries. The American Library Association's Core Values of Access, Equity, Intellectual Freedom and Privacy, Public Good, and Sustainability are all implicated in the integration of AI in society. Libraries exist for the people and have evolved alongside the technologies that affect the ways information is created, distributed, and obtained. Like the printing press, the computer, or Wikipedia, AI is yet another technology for libraries to adopt, teach, and provide to the communities they serve. Libraries must play a central role in cultivating the benefits of AI while also mitigating its harms. Such efforts are vital to the mission of ensuring that people can access the information they need – regardless of age, education, ethnicity, language, income, physical limitations, or geographic barriers.

Databases and Discovery tools

AI technologies have the potential to enhance and facilitate information discovery and retrieval. AI models can bridge the gap between a user's natural language queries and a database's formal query language or interface. Algorithms can analyze user intent, preferences, browsing history, and context to provide more relevant and personalized results. 

Although not search engines, generative AI programs like ChatGPT can aid in crafting search strings and terms, enabling the discovery of more relevant results. 

Collections and Accessibility

AI has the potential to make library and museum collections more accessible and engaging. Current workflows for collection digitization and description are labor-intensive and time-consuming. Integrating AI tools into these tasks could increase the pace and efficiency of the universal design process. A responsible approach to AI integration, involving experts and end users, adherence to ethical practices, and human oversight will help ensure AI's efficacy in accessibility. Potential applications include:
 

Transcription and Narration
AI speech-to-text and text-to-speech models could be used to generate transcripts and audio narrations for multimedia content like videos, audio recordings, and oral histories. 

Translation
Machine translation models could translate text descriptions, labels, and other collection metadata into multiple languages, increasing accessibility for non-native speakers and promoting cultural understanding.

Alt Text Generation
Computer vision models could automatically generate alternative text descriptions for images, artworks, and artifacts. This alt text provides crucial context for those with visual impairments using screen readers.

Content Summarization
Natural language processing models could automatically summarize long-form texts like manuscripts, making the key information more accessible and digestible.

Accessible Interfaces
Conversational AI assistants could provide voice interfaces and natural language interactions to help users discover and navigate digital collections in an accessible manner.