In a recent site visit to Stanford University, the WashU Digital Transformation team, along with University Libraries’ staff and leadership from the Transdisciplinary Institute in Applied Data Sciences (TRIADS), gained insights from a wealth of similar initiatives on the Bay Area campus, centered around healthcare, AI, digital infrastructure, and cross-disciplinary collaboration.
The WashU team visited with the groups below, which each play an important role in fostering a culture of data-intensive research at Stanford, from the infrastructure, to support, to the faculty on the front lines.
HAI (Institute for Human-Centered Artificial Intelligence): HAI works at the intersection of humans and AI to enhance human performance. The institute primarily fosters interdisciplinary work through pilot funding, industry relations, convening events, and graduate fellowships. HAI’s impact is attributed to its emphasis on outcomes, effective internal and external communication, policy work, and industry partnerships.
RAISE Health: RAISE Health is a joint initiative between Stanford Medicine and the Stanford Institute for Human-Centered Artificial Intelligence (HAI). Institutes such as HAI operate outside of school structures which allows them to play a vital role in fostering cross-campus, project-based collaborations. RAISE aims to ensure a common trajectory by conducting internal environmental scans and organizing convening events to build a library of relevant work at Stanford.
Literacy Lab: The Stanford Literacy Lab emphasized a bottom-up collaboration model led by faculty and students, focusing on data science and human-centered AI as cross-cutting themes. The challenges of connecting departments across campus were evident, and efforts were made to educate faculty about opportunities to apply for grants and provide grant support from personnel with experience in the research area.
Educational Technology (EdTech): Considering the post-COVID landscape, Stanford has emphasized a hybrid/high-flex curriculum model. Stanford takes a highly structured approach to select and prioritize content creation projects, which are expected to not only produce a course but also generate academic output. EdTech has also adopted a self-service approach, offering both in-person and virtual solutions for smaller projects, accompanied by virtual live support and an accessibility assessment.
Chief Data Scientist, Stanford Medicine: Stanford University’s IT leadership and infrastructure is distributed across the School of Medicine and the rest of the university. Investments that have facilitated campus-wide data science collaboration at Stanford include the establishment of on-premises GPU clusters. The visit highlighted the importance of communicating success stories internally and hosting small gatherings to foster collaboration.
Stanford Data Science: One leg of the visit was dedicated to Stanford’s Data Science initiatives, including four current centers. Notable projects include understanding human smartphone usage, mapping conversational dynamics, and the interdisciplinary use of data for varied research purposes.
Business Library: The Business Library at Stanford acts as a research hub, providing research compute capabilities, analytics, end-user support, and a data team. It licenses a wide range of data sets and offers a research operations team for grant support and communication. Efforts to build a comprehensive data index and support personnel like data concierge are in place.
Research Compute: Stanford’s Research Compute infrastructure shares similarities with WashU, focusing on providing technical support, on-prem vs. cloud computing decisions, and expanding AI capabilities through investments in NVIDIA infrastructure. The cost-benefit analysis between on-prem and cloud computing considers the scale and the potential impact of mistakes, particularly in training environments.