Roger Grosse and Marzyeh Ghassemi awarded AI2050 fellowships to advance research on beneficial AI
On December 10, 2024, Schmidt Sciences announced its 2024 list of AI2050 fellows, awarding USD $12 million to advance research on beneficial artificial intelligence (AI) and recognize scholars working to create AI systems for a better world.
Among this year’s recipients of these prestigious awards are two members of the Schwartz Reisman Institute for Technology and Society (SRI) research community: Schwartz Reisman Chair in Technology and Society Roger Grosse, who was named an AI2050 Senior Fellow, and Faculty Affiliate Marzyeh Ghassemi, who was named an AI2050 Early Career Fellow.
The AI2050 program asks researchers to imagine the year 2050, where AI has been hugely beneficial, and to pursue projects that help society realize this vision. The program funds senior researchers and early career scholars for two years to address a wide range of global challenges in AI.
Previous AI2050 fellows include inaugural Schwartz Reisman Chair in Technology and Society Gillian Hadfield, who was named an AI2050 Senior Fellow in 2023, as well as SRI Research Lead Karina Vold and Faculty Affiliate Nicolas Papernot, who were awarded AI2050 Early Career Fellowships in 2023 and 2024 respectively. This year’s announcement also named the University of Toronto’s Florian Shkruti, an assistant professor in the Department of Computer Science, as an AI2050 Early Career Fellow.
“Artificial intelligence has the potential to transform our lives for the better, particularly by accelerating scientific progress in profound ways,” said Eric Schmidt, co-chair of the AI2050 program and co-founder with his wife Wendy of Schmidt Sciences. “While the benefits of AI can far outweigh the challenges, realizing this potential requires thoughtful action today. We look forward to seeing how the 2024 AI2050 fellows help shape a future where AI serves the greater good.”
Grosse, an associate professor in the University of Toronto’s Department of Computer Science and a founding member of the Vector Institute where he holds a Canada CIFAR AI Chair, focuses his research on improving fundamental understandings of machine learning models to improve the safety and robustness of advanced AI systems. His AI2050 project will focus on algorithmic techniques that will help make transparent safety cases for next-generation AI assistants. This will include methods for detecting unintended consequences of training data or methodologies, as well as methods for estimating probabilities of rare catastrophic failures such as the AI model carrying out a harmful plan.
"AI has reached a point where its capabilities are extraordinary, but so are the risks. The goal of my research is to ensure that next-generation AI systems are developed safely in ways that are aligned with human values,” said Grosse. “I'm grateful and honoured to have the support of Schmidt Sciences through this fellowship, and to be part of a remarkable cohort of scholars."
Ghassemi, an associate professor at MIT in the Department of Electrical Engineering and Computer Science (EECS) and Institute for Medical Engineering & Science (IMES), focuses her research on the design and implementation of robust and fair AI models in healthcare settings. Ghassemi’s AI2050 project aims to make healthcare AI more equitable and reliable by cleaning and improving the data used to train AI systems in order to mitigate biases against women, minorities, and other underrepresented groups. Ghassemi’s project also will develop tools to monitor and maintain AI performance as medical practices evolve, keeping the AI safe and effective in real-world use.
The five senior fellows and 20 early career fellows in the 2024 AI2050 cohort join a community that now includes 71 researchers from 15 different academic disciplines and 37 institutions in seven countries. The researchers are working on projects as varied as using AI to transform drug discovery in Africa, reducing maternal mortality, creating a global “genome” of human values, unlocking mysteries in particle physics, and even completing household chores—all with an emphasis on safety and effectiveness of AI systems.
“The AI2050 program gathers experts across disciplines, continents and career stages to imagine a future that is brighter because of AI,” said James Manyika, co-chair of AI2050 and a senior vice president at Google. “I am excited about the work these fellows will undertake, motivated by our AI2050 challenge to tackle the hard problems associated with achieving beneficial outcomes for society from AI.”
Schmidt Sciences is a philanthropic organization that accelerates scientific knowledge and breakthroughs to support a thriving world. Founded in 2024, the organization prioritizes research in six focus areas: AI and advanced computing, astrophysics, biosciences, climate, science systems and space.
With files from Schmidt Sciences.