AI incubator makes understanding practical applications as easy as P-ai

Drawing of students working with computers
(Brynne Barnard-Bahn • The Student Life)

You can call them a club, organization or community, but P-ai — pronounced “pie” — is best described as an incubator for artificial intelligence (AI) projects and ideas. P-ai takes students’ open-ended research questions, startup ideas, hobbies that warrant collaboration and everything in between, then turns them into projects lasting a semester or longer. Started in May 2019, P-ai has quickly grown as the only AI club at the 5Cs.

While computer science is a growing field at the Claremont Colleges, some students have felt a gap in knowledge in practical applications, especially regarding AI, according to chairman and former president Alex Ker PO ’23. Ker described how the limited number of AI courses offered cannot cover the niches of the field.

“I think [P-ai] does a better job … addressing some of those specific interests that students have,” Ker said. “[Students] really like learning from their peers, with their peers or teaching their peers … Ultimately, it’s a lot of curiosity-driven learning versus that prescribed learning.”

Current president Varun Bopardikar PZ ’22 added that P-ai gives students a chance to work with real-world applications to data science, as opposed to the often theoretical work done in classes. By “democratizing the interdisciplinary applications of artificial intelligence,” P-ai aims to increase awareness of AI among all curious students, according to its website.

“Not everyone’s going to be interested in how it’s technically implemented, all the nitty-gritty behind all the algorithms and the statistics,” Ker said. “But we hope to at least allow students to be able to talk about it and be able to … understand its impact on our world.”

Working with AI might seem daunting, but students don’t need to have experience with it to join P-ai — the only requirement for project team applicants is a computer science course at the 5Cs or some coding experience. Even those that haven’t met this requirement can still apply as general members to access speakers, workshops, alumni panels and other events.

This past Saturday, P-ai held the first of a series of four workshops to introduce new project members to AI, with senior team members teaching some basics of Python and machine learning along with their connections to and importance in AI. This semester, the project team is working on six projects, each led by a project manager, ranging from one on extracting deep-fake images to one on neuron signaling in seizures.

Many of the fields students are exploring are interdisciplinary and require consultation from professors and students of various majors. Director of projects Andy Liu HM ’23 is also the manager for a project called “Applying GPT-3 to Interactive Fiction,” in which team members are generating code through language modeling to create interactive fiction stories.

“Students in our club have really diverse sort of interests and majors — not everybody likes just doing [computer science],” Liu said. “We have neuroscience majors and philosophy majors and environmental analysis majors. The beauty of AI is that it’s applicable to so many different fields. And, for example, if you look at our project list for this semester, we have a bunch of diverse products.”

At the end of each semester, project teams participate in a demo day to showcase what they have learned in a way that people with no AI experience can understand. While project timelines are flexible, P-ai expects all teams to have a deliverable by the end of the semester.

“They can present maybe a finding or maybe a result or maybe some places where they failed,” Bopardikar said. 

Online due to the pandemic, their last demo day allowed for projects to be shared more widely. When campuses shut down, P-ai was able to adapt more easily than most due to the nature of the organization. They moved all project meetings to Zoom and had co-working sessions to facilitate community. However, they still faced community challenges.

“I definitely felt looking at the project teams that there was a little bit less motivation to do projects,” Bopardikar said. “But I think we really tried our best to have as much club camaraderie as possible, like we hosted events where members could get to know one another … We persevered well, and all of our projects still turned out very, very well … We did grow a ton as a club, even during the pandemic.”

Furthermore, as many summer internships around the country were canceled, P-ai matched around a dozen students to various opportunities to advance their work in data science and machine learning at places like Samba TV and various research labs at Keck Graduate Institute, Harvey Mudd College and Pomona College.

They have now continued to partner with startups and professors interested in hiring students to do research or work.

“We like to help one another when we’re struggling,” Bopardikar said. “We want to be there for one another and want to help each other learn as much as possible.” 

Facebook Comments