Introduction

Universities aid in national artificial intelligence (AI) innovation and competitiveness in several ways. Most obviously, they are where much AI research and technological innovation gets done. Countless major AI discoveries have been the direct result of university research; in fact, the field of AI research itself was founded at a workshop held on the campus of Dartmouth College in 1956. Universities and colleges also educate and train the entire range of workers that will one day enter the AI labor market, including those who go into occupations directly related to developing AI systems, such as engineers, scientists, and product developers, and those who go into occupations that use AI systems. Finally, universities foster technological change indirectly via their links with industry and government. A 2022 report finds that collaborations in AI publications between universities and industry, universities and government, and universities and nonprofit actors have more than doubled over the past two decades.

The problem is university-based AI knowledge and talent production in many countries is concentrated to small groups of elite universities, and a chief cause of the divergence between the “haves” and “have nots” is access to AI infrastructure. Limited access to AI infrastructure, which includes hardware, software, data, and expertise, not only hampers the ability of academic AI researchers to solve important and pressing challenges, but also restrains the types of educational opportunities universities can offer students. Hyper-concentration of AI resources to well-funded universities therefore represents a serious productivity, competitiveness, and equity problem.

The U.S. government has rightly recognized it has an important role to play in increasing access to AI resources for academia and through government-university collaborations. The National Science Foundation (NSF) is currently leading an effort to create a national network of AI research centers by facilitating partnerships between government agencies and academic institutions.

There is also another type of collaboration that could create hubs of AI research and education that support national AI goals: industry-university partnerships. A recent and comprehensive example of such a partnership is between the University of Florida (UF), a public land-grant research university in Gainesville, Florida, and NVIDIA, a U.S. company that develops graphics chips. Anchored by gifts from UF alumnus and founder of NVIDIA, Chris Malachowsky, and hardware, software, training, and services from NVIDIA, UF has launched an initiative to become an “AI University,” which has significantly increased its ability to conduct impactful AI research, educate the Floridian workforce, and help ensure all individuals have equal opportunity to succeed in becoming the next generation of AI researchers by sharing its AI resources with groups that are traditionally underrepresented in science and engineering.

The UF-NVIDIA partnership to create an AI university is the first of its kind in the United States—a nation uniquely positioned for such partnerships, as, compared with many other countries, it has a highly developed and successful industry-university collaboration system. Among the many reasons why is its culture. A long tradition of pragmatism has dominated U.S. universities, leading them to view collaboration with industry as something that is useful and can advance knowledge. In addition, the U.S. system, with a diverse assortment of universities and ownership has created a competitive environment wherein universities innovate and compete to work with industry. Finally, in many states, public colleges and universities are encouraged and supported by state and local governments in their efforts to work more closely with industry.

Drawing on interviews with key stakeholders from industry, academia, and government, this report presents the UF-NVIDIA partnership as a case study to provide an understanding of how this partnership was constructed and implemented; highlight the impact of the initial outcomes from this partnership in supporting national goals for AI research, AI education, and equitable access AI to resources; and draw five lessons for effective industry-university partnerships to foster AI:

1. Focus industry-university partnerships on increasing AI computing capacity at universities.
2. Articulate realistic goals for partnerships based on a clear understanding of institutional capacity.
3. Encourage institutional buy-in within a partnering university by spreading the benefits of the partnership across disciplines.
4. Encourage government to stimulate demand for the AI research and skills industry-university partnerships create.
5. Develop an appropriate and ongoing assessment program.

Read the full report. (PDF)