Jane Warren at the Lincoln Center campus. Photos by Kelly Prinz
When Jane Warren was looking at colleges, she knew she wanted a school with a range of strong options across the sciences and humanities.
As a double major in English and math and computer and information sciences, she says Fordham’s wide-ranging strengths—and the ability to connect her studies— have been at the heart of her college experience.
“I think my English classes have given me such great tools to ground my science and math work, just because we think so much about social justice and how stories are told, and rhetoric and things like that,” Warren says.
Questioning Assumptions Through Research
Warren has used those lessons particularly in her research. She examined bias in AI-generated images and how stereotypes about body shape and weight get translated into those images. Warren presented her findings at the Nations of the Americas Chapter of the Association for Computational Linguistics.
She’s also exploring how gender and racial biases come up in AI-generated alternative text, which provides descriptions of images for people who can’t see.
“Our most interesting finding in that paper was that if we ask an AI model to describe an image of a person with lighter skin, they don't usually mention race, but if we ask them to describe an image of someone with darker skin, they will assign a racial label to them or point out their skin tone,” Warren says, adding the paper is currently under review to be published. “Hence that means they’re treating whiteness as a default identity.”
It’s her liberal arts background that helps her use math and computer science to tackle these complex issues.
“One thing my Fordham humanities classes have given me is a lot of appreciation for critical investigation of ideas that might be widely adopted,” she says. “These principles of social justice and equity have really been reiterated in all my courses, and that’s something I value taking into research as well. It’s really important to me to do research that, even if it’s in a technological field, it still stays grounded in, ‘How would humans think about this, or how would humans experience this?’”
At Fordham, Warren pursued a double major in English and math and computer science.
Taking a Critical Lens to AI and Beyond
Warren brings this critical lens to her internships as well, including one at Together AI, where she’s a research intern examining benchmarks used to evaluate AI’s mathematical abilities.
“That’s just another space where I'm really being pushed to develop my own identity as a researcher and thinker,” she says.
Warren plans to pursue a doctorate to continue researching how “humans are affected by technological change.”
“A lot of my work has been very influenced by people who might be seen as ‘radicals’ in the AI research community, and who think about algorithmic harms,” she says. “In this case, it’s AI as an idea that is being pushed pretty ferociously in schools, in corporate and business settings, and in people’s everyday lives, but it’s accelerating so quickly and not very many people are questioning, like, ‘Oh no, what are ways this might be doing harm?’”