Back to Annual Report

The AI Experts At Texas A&M University Libraries

Universities everywhere are trying to define the role of artificial intelligence (AI) in research and academics. 

Since higher education has served as the foundation for the research that fuels Fortune 500 companies, it stands to reason that it will also be the driver of the AI age. 

Students and staff on campuses everywhere are asking the same questions. What is the future role of AI? How will it affect me? Texas A&M University Libraries are quickly becoming experts at delivering answers.

“It's very much our responsibility to stay on top of the new tools that are coming out and the claims that they're making, testing them whenever we can,” said Esmeralda Rodgers, evidence synthesis librarian. 

Rodgers is part of a team of librarians focused on using AI to process large amounts of information from multiple sources to answer specific questions.  

As part of her exploration, she is experimenting with creating an AI agent (a “digital worker” or “brain”) with a large knowledge base that guides users. Rodgers’ primary goal is training the AI agent to focus on systematic reviews, a common research step that requires sifting through and summarizing vast amounts of data.

The Libraries also help professors, students and staff productively incorporate AI into their work and studies. 

For example, Medical Sciences Library staff worked with Texas A&M pharmacy students to utilize a large language model AI to role play difficult patient conversations. This new tool allows them to practice discussing diagnoses with a variety of character simulations before they need to do so in the workforce. 

"It was just a great way to try out some different Large Language Models (LLMs), said Amy Martin-Klumpp, research and education librarian. “As well as to evaluate all the different LLMs, both the ones that are supported by Texas A&M, as well as some other ones that are open freely."

Considerations need to be made when using new and rapidly-developing technology, such as AI. 

Rodgers and Martin-Klumpp advise patrons to: 

• Experiment with a variety of AI apps because they produce different results;
• Validate, validate and validate again all findings;
• Take into consideration data privacy and what the AI is learning from what is being input. 

“You can use AI to summarize literature, to kind of give you ideas of what you want to search for,” Rodgers said. “But it is not a replacement for critical judgment and subject expertise.” 

Additional AI assistance resources from the Libraries include an AI Literacy tutorial series and AI-related workshops.