Skip to main content
Download our Annual Report (PDF)

This Annual Report reflects the research, activities, and impact of Stanford HAI between September 2021 to August 2022


An Interdisciplinary Approach

HAI’s research efforts gained significant momentum in the 2021-22 academic year. The team was excited to return to in-person events and to establish the Ethics and Society Review process as part of HAI’s grant programs.

HAI Research Focus

HAI focuses its research initiatives around three main priorities, each with its own vision and mission:

Human Impact

To develop technology that is equitable, researchers must understand how AI interacts with humans, as well as how it interacts with vital social structures and institutions.


Civics Education for a Just and Sustainable Future

Understanding and changing textbook content is one important lever in the multifaceted process required to redesign history and civics education. Patricia Bromley in the Graduate School of Education has designed a study that uses Natural Language Processing (NLP) to assess textbook content and document depictions of diversity, equity, and sustainability.

Augmenting Human Capabilities

HAI believes that breakthroughs in human-centered design methods will further progress in healthcare, education, sustainability, automation, and countless other domains.


Rating Systems and the Future of Algorithmic Worker Evaluation

A team from the Schools of Engineering and Humanities and Sciences is developing a forward-looking digital resumé for workers in online labor platforms. The project brings together statistical methodology, algorithm design, human-computer interaction, and political economy to enable the same career pathways for online workers that society takes for granted in offline work.


To create a machine-assisted—yet human-centered—world, the AI community must develop the next generation of techniques that overcomes the limitations of current algorithms, expands the class of problems that can be addressed, and complements human cognitive and analytic styles.


Toward Machine Models of Structured Human Reasoning through Explanation-based Meta-learning

Some of the most impressive achievements of the human mind have led to deep scientific understanding and powerful technologies. Two researchers from the Schools of Engineering and Humanities and Sciences seek to combine human and machine learning research to build artificially intelligent systems that could potentially achieve these abilities.

Grant Programs for AI Research

Since its founding, HAI’s grant programs have supported 248 faculty members from all seven Stanford schools. Many of these projects span multiple departments, in keeping with the commitment to support interdisciplinary AI research.

How HAI Research Grants Connect Departments

HAI Seed Grants

HAI seed research grants are designed to bring faculty together across disciplines to share knowledge and pursue new ideas. Three projects received notable awards to continue their research:

Hoffman-Yee Grants

Hoffman-Yee research grants are made possible by a gift from philanthropists Reid Hoffman and Michelle Yee. In August 2022, HAI announced its second cohort of Hoffman-Yee Grant recipients. Six research teams were each awarded a total of $2.75 million to solve some of the most challenging problems in the field of AI. Each of the proposals went through a rigorous review with faculty, as well as an Ethics and Society review.

Cloud Credit Grants

HAI saw record demand for cloud credit grants during AY 2021-22. This program provides advanced computational resources that researchers need to conduct rigorous AI projects. HAI allocates these credits to studies showing promising, novel, or emerging ideas.

In the second year of the Google Cloud Credit Grants, HAI awarded a total of $2.5 million in GCP credits to 48 faculty across 59 projects. The institute also launched the Microsoft Azure Cloud Credit Grants and awarded $330,000 in Azure credits to 7 faculty across 11 projects.

Featured Cloud Grants

“Foundation Models Homogenize Outcomes,” Percy Liang, Director, Center for Research on Foundation Models, and Associate Professor of Computer Science Human Impact—Azure

“Online Surgical Analytics through Computer Vision via Multi-task Learning,” Serena Yeung, Assistant Professor of Biomedical Data Science Augment Human Capabilities—Google

“Large-scale Evaluation of Visual Intuitive Physics,” Dan Yamins, Assistant Professor of Psychology and of Computer Science Intelligence—Google

Research Centers

Center for Research on Foundation Models

Center for Research on Foundation Models (CRFM) celebrated its one-year anniversary in August 2022 and now counts more than 300 students, faculty, and postdocs working on relevant research projects.

This center was founded to drive fundamental advances in the responsible study, development, and deployment of foundation models. Foundation models are resource-intensive AI systems, such as GPT-3, Codex, StableDiffusion, Imagen, and AlphaFold, which are trained on broad data at scale. Foundation models

provide valuable representations for many downstream applications; however, the technology is advancing at a rapid pace and there are unknown risks that must be considered.

Stanford established CRFM to focus on three pillars:


The center conducts interdisciplinary research that lays groundwork for how foundation models should be built to make them more efficient, robust, interpretable, multimodal, and just. During the past academic year, CRFM scholars authored dozens of papers, including:


The center builds and releases foundation models, code, and tools to improve the ecosystem.


CRFM engages with universities, companies, policymakers, and civil society to develop professional norms that govern the responsible development and deployment of foundation models.

Percy Liang

Percy Liang, Director of the Center for Research on Foundation Models

Percy Liang is an associate professor of computer science at Stanford University and the director of the Center for Research on Foundation Models. His research spans many topics in machine learning and natural language processing, including robustness, interpretability, semantics, and reasoning. Percy’s awards include the Presidential Early Career Award for Scientists and Engineers (2019), IJCAI Computers and Thought Award (2016), an NSF CAREER Award (2016), a Sloan Research Fellowship (2015), and a Microsoft Research Faculty Fellowship (2014).

As the director of CRFM, he is leading Stanford’s efforts to make fundamental advances in the study, development, and deployment of foundation models through an interdisciplinary group of faculty, students, postdocs, and researchers.

Research Centers

Stanford Digital Economy Lab

Over the last academic year, the Digital Economy Lab has been focused on conducting research and hosting major events. The lab currently has more than 50 research projects in its portfolio.

Erik Brynjolfsson and Christie Ko

Erik Brynjolfsson, Director of the Stanford Digital Economy Lab

Jerry Yang and Akiko Yamazaki Professor and Senior Fellow at HAI, Ralph Landau Senior Fellow at the Stanford Institute for Economic Policy Research (SIEPR), and professor, by courtesy, at the Stanford Graduate School of Business and Stanford Department of Economics.

Christie Ko, Executive Director

Research Highlights

Areas of Research

Measuring the Digital Economy

Creating better methods of measuring the health of an increasingly digital economy

AI & The Future of Work

Understanding the future of the workforce in a rapidly changing global economy

Adoption of Advanced Technologies and Management Practices

Measuring, assessing, and predicting how data-driven decision-making and management practices are impacting the global economy and workforce

Digital Platforms and Society

Exploring how digital technologies can transform platforms and social media infrastructure to benefit society

Community Engagement and Public Outreach

To gather high-level feedback on its research areas, the lab hosted two advisory group meetings, which pointed the organization toward some key directions for growth and the expansion of its work. To build community at Stanford, the lab hosted two Stanford faculty roundtable discussions as well as a summer community BBQ in collaboration with HAI, the Golub Capital Social Impact Lab, and Stanford’s RegLab.

Research Centers

AI Index

Measuring trends in artificial intelligence.

HAI and the AI Index Steering Committee published the 2022 AI Index report in March 2022. The latest edition included data from a broad set of academic, private, and nonprofit organizations as well as more self-collected data and original analysis than any previous edition. The authors included a variety of new data points to cover technical performance in reinforcement learning, the pricing of robotic arms, and artificial intelligence legislation.

The AI Index received positive press coverage, both around the time of the launch and throughout the rest of the academic year: highlights included pieces from publishers like IEEE Spectrum, Fast Company and Morning Brew. A number of prominent publications cited information contained in the report—TechCrunch, Fortune, Forbes and the Harvard Business Review among them.

Chess Software Engines: Elo Score

Progress in reinforcement learning, an important AI skill capability, can be captured by the performance of the world’s top chess software engines. Computers surpassed human performance in chess a long time ago, and since then have not stopped improving.
Source: Swedish Computer Chess Association, 2021. Chart: 2022 AI Index Report