As artificial intelligence (AI) transforms society at a breakneck pace, are we asking the right questions about what it’s for and whom it serves? Rice University’s Kirsten Ostherr said the answer depends on when and how humanists are brought into the process.
Ostherr, director of Rice’s Medical Humanities Research Institute (MHRI) and the medical humanities program, recently participated in “Doing AI Differently,” a two-day workshop hosted by the Alan Turing Institute in London that brought together global leaders in computer science, the arts and the humanities to reimagine the foundations of AI.
“This was an international convening by a prominent institution saying, ‘Yes, the work you’ve been doing between medical humanities and engineering at Rice is truly cutting edge, and we need your insights to lead this international effort forward,’” Ostherr said.
The workshop emphasized that humanities disciplines, especially medical humanities, shouldn’t just be consulted at the end of the development pipeline when systems are being evaluated for bias or misuse. Instead, humanities scholars should be present from the beginning, helping to define the problems AI is being built to solve.
“Up to this point, the main way that humanities have been brought into conversations about AI is through the lens of ethical assessment after the AI is already in use, which is of course very important,” Ostherr said. “But the premise of this group is that they should be involved at a much earlier starting point.”
For Ostherr, that earlier point begins with defining what counts as a problem in the first place.
“Before you start to say, ‘Let’s develop an AI tool to solve X problem,’ there’s a question of what is the problem and how are we defining it, which is actually something that the humanities brings a lot of richness and depth of understanding to,” Ostherr said.
This approach mirrors the work of MHRI, which places human flourishing at the center of health care and technological innovation. By uniting humanists, artists, scientists and clinicians, the institute tackles major global challenges that can’t be solved by science or technology alone.
“Participating in this workshop made it clear to me that the Turing Institute was truly endorsing the kind of work we’ve been doing at the Medical Humanities Research Institute,” Ostherr said. “It was a powerful signal that leaders in computer science recognize how essential the humanities are to shaping AI.”
That recognition is happening in the classrooms at Rice as well. For example, the Ostherr- taught course Responsible AI for Health attracts students from the George R. Brown School of Engineering and Computing and also contributes to the medical humanities minor.
“The engineering students in that class really appreciate having space to talk about the kinds of issues we were exploring,” Ostherr said. “They are able to work on some creative projects, which helps them think differently about how they would develop AI tools and why.”
The Turing Institute workshop’s collaborative, interdisciplinary design points to a future where AI systems are better aligned with societal values and more adept at capturing cultural and contextual nuance. This aligns with Rice’s broader push to become a global leader in responsible AI development, an initiative that emphasizes ethical advancement, cross-disciplinary research and innovative applications of machine learning across sectors.
As funding bodies across the United Kingdom, United States and Canada prepare new support mechanisms for interdisciplinary AI projects, Ostherr said Rice is well positioned to lead.
“For Rice and for the future of AI, the message is clear: The humanities are not a luxury,” Ostherr said. “They’re an essential part of building technologies that serve people.