Stanford HAI Fuels AI Breakthroughs Amid Funding Challenges and Uncertainty

Stanford AI Scholars Find Support for Innovation in a Time of Uncertainty

Date: July 01, 2025
Topics: Machine Learning, Foundation Models, Education, Skills

Federal funding cuts to key scientific organizations like the National Institutes of Health (NIH) and the National Science Foundation (NSF) have slowed progress in medicine and technology across the United States. This reduction in financial support threatens the country's ability to attract top scientific talent and slows advancements in critical research areas.

Amid this challenging environment, the Stanford Institute for Human-Centered AI (HAI) continues to provide essential resources to faculty and students, encouraging innovation throughout the AI landscape. Now in its sixth year, HAI draws on a diverse mix of funding sources including individual donors, philanthropic organizations, and industry partners alongside government grants.

Vanessa Parli, director of research at HAI, emphasizes the importance of partnerships. “Many researchers are losing resources needed to test ambitious ideas, making HAI’s grant programs more critical than ever,” she says.

Hoffman-Yee Research Grants

The Hoffman-Yee Research Grants, HAI’s largest funding initiative, support bold, interdisciplinary projects addressing scientific, technical, or societal challenges. Funded by Reid Hoffman and Michelle Yee, these grants provide up to $500,000 in the first year with the potential for an additional $2 million over two years. Successful proposals must align with HAI’s core research areas: human and societal impact of AI, augmenting human capabilities, and developing AI inspired by human intelligence.

One notable project led by HAI Senior Fellow Michael Bernstein explores how societal values can be integrated into social media algorithms without compromising engagement. Their work, supported by Hoffman-Yee funding, has allowed them to test new ranking methods on over 1,000 Twitter (X) users in real time. The team is now building tools that empower users to customize their social media experience. Bernstein highlights the value of cross-departmental collaboration fostered by HAI’s support, which is rarely achievable elsewhere.

Seed Research Grants

HAI’s Seed Research Grants help researchers explore speculative AI ideas in early stages. These grants support a wide variety of projects, from discrete studies to system development and speaker series, aiming to generate initial results that can lead to larger efforts.

For example, Jonathan H. Chen, assistant professor of medicine and HAI Faculty Affiliate, credits these grants with advancing work at the HealthRex Lab. His team received an HAI and AIMI Partnership Grant to study machine learning models that predict oncology diagnostics, a project recently published in Nature Digital Medicine. Chen points to the interdisciplinary collaboration enabled by HAI as crucial to their progress.

Graduate and Postdoctoral Fellowships

HAI also supports interdisciplinary research through fellowships for graduate students and postdocs. These programs encourage collaboration across fields related to AI’s applications and societal effects. Postdoctoral Fellow Joba Adisa works with Victor Lee from the School of Education to promote AI literacy among high school teachers via the CRAFT project. Adisa describes HAI as a “diverse community offering mentorship, events, and collaboration opportunities” that help shape AI research to augment human capabilities.

Faculty Fellowships

Faculty fellows at HAI focus on AI’s intersection with other disciplines. Johannes Eichstaedt, the first HAI faculty fellow, studies AI and psychology to better understand AI’s mental health implications. He praises the university’s support through seed grants, startup funds, and cloud resources that foster interdisciplinary research. Eichstaedt is currently working on a book examining AI’s social and psychological effects.

Another fellow, Hari Subramonyam, combines human-computer interaction and learning sciences to explore AI that enhances human creativity and learning. HAI also jointly supports fellows with Stanford’s seven schools, such as Erik Brynjolfsson from the Graduate School of Business and Yejin Choi from the School of Engineering.

Google Cloud Credit Grants

Partnering with Google, HAI offers Cloud Credit Grants to provide researchers with up to $100,000 in Google Cloud computing resources. These credits enable projects requiring significant computational power. Recent recipients have developed a new social reasoning benchmark for large language models and trained discrete diffusion models that outperformed GPT-2, winning best paper at the International Conference on Machine Learning (ICML) 2024.

HAI Workshops

Workshops funded by HAI bring together experts to explore shared interests. Upcoming is the Nightingale Workshop, hosted by assistant professor Judith Fan, focusing on data and visualization literacy—a key area for improving public scientific understanding. Fan notes HAI’s support is vital given uncertainties in future NSF funding.

Recent workshops have addressed topics like Interactive AI Systems for Live Audiovisual Performance and Public AI Assistants for global knowledge access.

Student Affinity Groups

To foster the next generation of AI scholars, HAI funds cross-disciplinary student affinity groups. These teams, which include undergraduates to postdocs across Stanford’s seven schools, receive support for one academic year to investigate human-centered AI topics.

PhD candidate Julia Di focused on the future of embodied AI, while postdoc Kristina Gligorić formed a group exploring intersections between natural language processing and computational social science.

Driving Collaboration and Impact

HAI has supported over 500 projects, with many recipients valuing the collaborative environment as much as the financial backing. By uniting teams across departments, HAI helps address AI challenges that require diverse perspectives.

As faculty fellow Johannes Eichstaedt puts it, “It’s in the spaces between disciplines where larger impacts are more likely.”

For more examples of recent AI projects and details on fellowship and grant applications, visit Complete AI Training.