Stanford HAI Announces 2022 Seed Grant and Wu Tsai Neuro Grant Recipients
37 research teams receive a total of $3 million for innovative AI projects
The Stanford Institute for Human-Centered AI (HAI) has awarded seed grants to 33 AI research teams representing all seven schools within the university. The funding is intended to bring faculty together across disciplines to share knowledge and pursue new ideas in the field. This year’s research themes range from healthcare and culture to mental health and robotics.
Stanford HAI’s research focus falls into three main areas: Human Impact, Augmenting Human Capabilities, and Intelligence. Grants are awarded in each focus area to ensure that humanity benefits from AI technology and that the benefits are shared broadly by all. The seed grant program launched when Stanford HAI was founded in 2018. To date, more than $10 million has been committed to early-stage AI research.
HAI & Wu Tsai Neuro Partnership
Four additional research teams received grants through a new partnership with the Wu Tsai Neurosciences Institute at Stanford. The two organizations each contributed $250,000 to support researchers who aim to transform brain health, counter disease, and develop AI technologies inspired by the versatility and depth of human intelligence.
For example, one of the grantees under this program is working on a new approach to at-home stroke rehabilitation therapy, using robotics, brain-computer interface, virtual reality, and wireless technology. By introducing real-time feedback, the researchers believe the system will be able to engage more neural circuits in the patient’s brain and enhance physical therapy.
“HAI and Wu Tsai Neuro share a commitment to funding proposals that make a persuasive case for how initial results will catalyze further support from internal and external stakeholders,” said James Landay, Stanford HAI Vice Director and Faculty Director of Research.
2022 HAI Seed Grant Projects
Many of this year’s awardees focused on AI advances in healthcare, including areas like patient monitoring, cancer treatment and retinal implants. Others plan to examine the impact of technology on our mental health, the role of AI in climate change and sustainability, advancing robot learning, and AI applications in art and education.
Below, a few examples of this year’s winners:
Augmenting Physician Capabilities with AI-Powered Patient Monitoring in the Emergency Department
In most hospital emergency departments, vitals are taken via sophisticated equipment, but only shared with clinicians manually when a nurse writes down the info, which can range from every 10 minutes to every few hours.
"The scarcest resource in the emergency department is the time and attention of clinicians and staff,” says project lead David Kim. “Our patients are connected to monitors that continuously record multiple dimensions of their physiology, but without placing those signals in the context of other clinical data, and distilling them to something interpretable to humans, most of this information goes to waste.”
To address the problem, this team has developed a first-of-its-kind dataset and deep learning models to monitor vital signs of 50,000 emergency department patients continuously. “We want to empower physicians to make the best decisions for their patients by harmonizing the relevant data streams in real time and allowing the physician to focus on personalized diagnosis and management, instead of data collation and chart review."
Funding from HAI promises to further their efforts to deploy these models, enhancing the emergency department physician’s ability to understand patients’ individual physiologies, clinical risks, and likely responses to therapeutic interventions.
Identifying Cultural Values in Language Models
In a human impact study of large language models, a team proposes to explore the values and biases that are present in generative models trained on large amounts of textual data. The team has identified three topics to study: cultural representation, characterization, and impact. Using psychological surveys and other tools, they hope to go beyond stereotypes related to demographics and identify more subtle – and potentially harmful – imbalances in training data.
Optical Convolutional Neural Networks for Enhanced Human Vision
What if we could see through fog or rain and have clear, daytime vision at night? This team believes that current state-of-the-art approaches to enhancing human vision and imaging capabilities has barely scratched the surface of what’s possible. For this project, researchers will harness recent advances in the design of flat optical elements and computational imaging to take vision and imaging systems to previously unimaginable levels of sophistication.
Creative Physical Problem Solving in Humans and Robots
Another HAI-funded team aims to help robots mimic the human ability to apply learned manipulation skills to new objects, scenarios, or goals. The team is developing new computational models to allow more creative problem solving in robots, while also looking at how well these models explain human reasoning.
"This funding is enormously important,” says project lead Jeannette Bohg. “It allows me to explore an initial idea for a year with a student and other faculty. It's also specifically interdisciplinary and thereby fosters interaction across campus, which can lead to novel insights on a problem."
Aside from these defining trends, additional grants were awarded to further the study of AI in the areas of hardware, data privacy, and misinformation. “We couldn’t be more excited about the potential of this year’s cohort,” says HAI Research Director Vanessa Parli. “All of these teams are working on early, innovative, and interdisciplinary AI research with the goal of making sure the technology benefits humanity.”
As with last year’s grant recipients, each winning team for 2022 completed a rigorous Ethics and Society Review process to consider any negative impact before the project was approved.
The full list of 2022 HAI Seed Grant recipients is available here. And projects funded through this year’s HAI & Wu Tsai Neuro Partnership are listed here.
Learn more about Stanford HAI grant programs.
Related News
The AI Sovereignty Paradox: Should Countries Buy, Build, or Lease to Maintain Strategic Control of Their AI?
As nations invest billions to reduce reliance on foreign AI providers, a new Stanford HAI report surveys commercial sovereignty solutions and assesses the extent to which they meaningfully reduce dependencies on U.S. tech giants.
As nations invest billions to reduce reliance on foreign AI providers, a new Stanford HAI report surveys commercial sovereignty solutions and assesses the extent to which they meaningfully reduce dependencies on U.S. tech giants.
With increased use of chatbots in mental health contexts, AI developers now rely on human experts to evaluate AI’s responses for “safety” – but experts rarely agree on what’s safe.
With increased use of chatbots in mental health contexts, AI developers now rely on human experts to evaluate AI’s responses for “safety” – but experts rarely agree on what’s safe.
Biomni can analyze mountains of medical data, spot patterns humans might miss, and even design experiments—helping researchers make discoveries faster in the race to cure disease.
Biomni can analyze mountains of medical data, spot patterns humans might miss, and even design experiments—helping researchers make discoveries faster in the race to cure disease.