The AI Lab at the University of Rhode Island (URI) has been open only since September 2018, but it is already sparking a potential wave of similar facilities across the US and the world.
The lab is located within the university’s Robert L. Carothers Library and Learning Commons, making it the first lab of its kind to be housed in a college library. The location provides easy access for students, faculty, and even the local community to learn about both the technical and nontechnical aspects of artificial intelligence (AI). Some of the equipment available includes an AI supercomputer, six specialized laptops for running large data sets and open-source software, 3D printers, and virtual reality technology.
“Why in a library?” asks Karim Boughida, dean of libraries at URI. He promptly continues: “Because the library is the common space for people to explore or create new things.”
According to Boughida, the mission of the lab is threefold. Currently, the first priority is education, followed by research and then community building. The lab encompasses 600 square feet of space on the first floor of the library and has three “zones.” Zone 1 is equipped with the supercomputer, where individuals can test out basic exercises and prepare to transition to more complex projects. Zone 2 provides hands-on opportunities for students to develop lab projects based on robotics, the Internet of Things, “smart” cities, wearable technology, and big data analysis. Zone 3 is dedicated as a brainstorming hub where groups are encouraged to collaborate and discuss challenges or potential opportunities with AI technology. In addition to the more freeform style of Zone 3, part of the educational program of the AI Lab involves discussions around the ethics and social justice of AI technologies. The lab hosts a series of workshops and “AI Meetup” events, where attendees are invited to join discussions and presentations on a variety of topics surrounding AI development, ethics, data bias, and fairness.
“I’m interested in providing a space where student-centered learning can go on. I want the students to drive the progress and determine what the programs or activities will be,” says Harrison Dekker, interim director of the AI Lab. “There’s a whole set of skills that are emerging that aren’t necessarily provided in the curriculum. So by having these sorts of training spaces, it helps fill the gaps in the curriculum, and also helps students take what they’re learning in the classroom on a theoretical level and then apply it to material things, code, or an actual project,” he adds.
A nearly $150,000 grant from the Rhode Island-based Champlin Foundation provided much of the support to jump-start the lab. An interdisciplinary effort from faculty within the library and the computer and biomedical engineering and philosophy departments, as well as computer science statistics and URI’s Big Data Initiative, enabled the lab to get up and running at the start of the fall 2018 semester. Getting representation from multiple fields and departments from the start of the lab launch drives home the message that the lab’s resources are available to anyone with an idea or concept that they want to carry out, and aren’t restricted to those in a specific field or major.
Finding a home in the library
Incorporating this type of lab into URI’s library had been a long-time dream of both Boughida and Dekker. As AI has become more of a trending “hot topic” in the news and among researchers, Boughida thought it was the perfect time to bring the vision to reality. The students at URI agreed. A recent survey that asked URI students about the topics they would like to see in their curriculum showed that AI was among the top requests. Housing the lab in the library—a central, welcoming location for all students and faculty—was a natural fit.
“People have a stereotypical view of what librarians do or what goes on in a library. It’s important for more libraries to be involved with these new, emerging conceptual spaces,” says Dekker.
Some examples of the groups working, and the work being done, in the AI Lab include engineering students applying machine and deep learning algorithms to enhance designed wearable items that collect data on health; a group of students using the lab’s processing power to gain a deeper understanding of using brain electrical activity to control robots; and a philosophy class conducting programming exercises to engage in discussions related to relationships between humans and machines. The lab will also serve as a generator of brand-new courses that can explore and incorporate AI.
“When you’re the first, some people will question why,” says Boughida. “We believe people aren’t [fully] aware of the impact AI will have on our daily lives, and since we see ourselves as a service to the community, we need to do something about it in terms of educating students, the public, and faculty.” He also notes that AI has become more of an “umbrella” term now and a lot of different technologies fall under the AI label, such as machine learning, natural language processing, and robotics, among others, which is why it is important to learn the various applications and their impact. “Other universities and research institutions are looking at us to lead this trend,” Boughida says.
When asked what the future vision of the AI Lab is, Boughida says that he is “thinking big.” He hopes to develop a long-term plan to allocate and train more individuals to join the lab, expand its services, establish more relationships with industry partners, and work with other libraries on data projects. Dekker echoed Boughida’s vision, adding that he would like to leverage the existence of the lab for fundraising efforts, to earn grants, and create more opportunities for students to get paid work experience. “I’d also like to have more of an applied research focus—where we can do more library-centric applications of artificial intelligence.”