Visions of the future: four ways in which generative AI could be integrated into libraries

What are the challenges of artificial intelligence (AI) for libraries, and how could it be leveraged to improve services and users' search for information? When used responsibly, generative AI offers new opportunities for libraries, and its potential applications are many and very varied. Here are four aimed at meeting people's needs.

^{Click on the infographic to zoom in.}

Improved user experience

• A new generation of search engines. AI promises to make it easier to search for information with more efficient and intuitive search engines that help find accurate results through the use of natural language and in any language.

There are already several AI-based search engines, with varying levels of accuracy, on the internet. Some publishers are working to integrate AI-assisted search tools into their databases: Elsevier has launched Scopus AI, whose features include an automatic summary of documents, and in December, Clarivate began beta testing an AI assistant for the Web of Science.

• Tailored recommendations: by analysing data using AI, libraries could anticipate user preferences and make personalized suggestions, helping users discover new contents and encouraging the use of collections. Examples include the WorldCat.org's AI-based book recommendation tool and Helsinki Central Library's mobile app Oodi, where a virtual assistant makes personalized reading recommendations.

"With necessary preparations – and regard for ethical concerns and current limitations – libraries can responsibly use AI technologies to advance their social mission."

^{Source: IFLA Statement on Libraries and Artificial Intelligence, 2020}

AI literacy: an opportunity for libraries

Some libraries are gradually adapting their information literacy programmes to guide users in the ethical and critical use of AI tools and search engines. The key issues addressed include:

• The responsible use of generative AI tools: recommendations on data privacy, intellectual property and confidentiality;
• Critical reasoning to analyse the potential of AI tools as well as their limitations;
• Assessing the results obtained and checking the information with academic sources;
• The basics of generative AI and its impact on society;
• The possible uses of AI tools and how to interrogate them to obtain good results.

An example of AI literacy is ExperimentAI from the Universitat Autònoma de Barcelona's Computer Vision Center (CVC-UAB), a citizen science project to introduce AI and computer vision to a general and non-specialized audience. Another initiative is The Laughing Room, as mentioned in this article, an installation at the Cambridge Public Library that encourages reflection on the impact of technology on our lives. Another case is that of the University of Rhode Island, which in 2018 launched an artificial intelligence (AI) laboratory with equipment, workshops, and courses.

These types of initiatives involve training library professionals in generative AI and the use of its tools so they can support users when it comes to giving well-defined and precise instructions and prompts. This is why training in prompt engineering is becoming more and more important.

Chatbot support

Libraries use generative AI conversational assistants to answer the simplest queries received from users so that human staff can attend to more complex ones and provide more personalized support. 

There are currently several libraries that use chatbots, such as the library of the University of Oklahoma with Bizzy or the Lehman College library.

Data management and analysis

AI provides new opportunities for analysing large datasets, improving analytics and helping make informed decisions. This is key for developing collections based on usage patterns and predicting future demand for resources.

Helping library staff

Apart from these four potential uses, generative AI can also help with tasks carried out at libraries. The IFLA's Artificial Intelligence Special Interest Group, for example, highlighted the importance descriptive AI could have in managing documents, cataloguing and creating metadata to make collections easier to find. They also pointed out how generative AI could be used as a support tool for marketing, creating and curating content.

Although the applications of artificial intelligence in libraries are promising, they will always have to be analysed critically and used responsibly.

An AI project at the UOC Library

The UOC Library has created a working group to analyse the possibilities of AI and assess its use. One of the projects is currently studying how to integrate this technology into the Library website to provide customized recommendations based on user preferences and offer a search engine that returns more accurate results and personalized suggestions.

The UOC Library team presents the "Personalització de l’experiència d’usuari amb IA" ("Customizing the User Experience with AI") project on the #UOC2TheFuture 2023 website

 

In order to facilitate the learning and understanding of AI within the UOC community, the UOC Library team has created two library guides in collaboration with the group of experts on artificial intelligence (AI) from the eLearning Innovation Center (eLinC).

To learn more

Experts

• 

  Natàlia Castellana

  Operative subgroup: Coordinator. Communication and website Operative group: Library Development