The dramatic rise of AI capabilities during the past decade is a watershed event for humanity and the world economy.  It is also sure to transform research and teaching in every academic discipline. With our stellar interdisciplinary team, Princeton Language and Intelligence seeks to greatly enhance fundamental understanding of AI, enable its use in academic disciplines, as well as examine AI’s safety, policy, and ethical implications. We are grateful to Princeton University for funding our activities —including a group of postdocs and research staff, seed funding, and a state-of-the-art computational infrastructure with 300 Nvidia H100 GPUs. Our seed funding and technical know-how will enable us to impact AI-related research and teaching across Princeton. Our governance structure includes colleagues representing over a dozen disciplines.

Our resources will let us focus on research questions that will not get addressed by corporate labs. What are efficient ways to train large models—even better, repurpose existing ones—to obtain inexpensive models with high performance to cost ratio? How do we adapt large AI models —especially models that can be instructed in plain English— to enable transformational research methodologies in humanities and social sciences? Can we adapt AI to create new foundation models that enable better human comprehension of scientific data, e.g., brain imaging; DNA sequences;  chemical reactions etc.? Can we develop new methods for design and use of AI models that align better with human needs, assure safety, and prevent societal harms that this technology might otherwise cause?

We are very aware of the important role that organizations such as Princeton Language and Intelligence can play today. AI has swiftly become a battlefield, and is in danger of developing primarily inside private labs, with little opportunity for the rest of the world to know how it works or be assured that it is responsibly deployed with full safeguards. We are committed to keeping AI expertise and know-how in the public sphere and are eager to work with institutional partners and companies who share our commitment to open research. Initiatives such as PLI can democratize AI knowledge, enhance transparency and accountability, and help ensure that benefits of AI accrue to all of humanity.

I invite you to join us on this journey!
 

Sanjeev Arora
Charles C. Fitzmorris Professor in Computer Science
Director of Princeton Language and Intelligence
Princeton University