A new paradigm for partnership between industry and academia in the age of AI

In the global pursuit of top AI talent, the world’s leading scientists and researchers are often faced with a difficult choice between academia and industry. But this choice has potentially debilitating long-term consequences because today’s university professors are the stewards of tomorrow’s talent. When they depart from academia, the pipeline of AI talent available to all sectors — industry, academia, and government — shrinks over time. To avoid this tragedy of the commons, there is a critical need for new paradigms of multisector partnership between academia, government, and industry that enable a better balance of talent across sectors.

Historically, industry has looked to academia for specialized talent across many fields of expertise. And over the years, various models of collaboration have been explored, including academic consulting, private-sector sabbaticals, industry-funded research at universities, and even venture funding that allows university faculty to pursue an entrepreneurial idea for a year or two.

But starting in the mid-2010s, with the arrival of deep learning, the need for artificial-intelligence, machine learning, and data science talent grew exponentially and introduced a fundamental change to these longstanding historical mechanisms for industry-academic partnership. Because AI development requires a rare combination of multidisciplinary skillsets, large datasets, and computational capacity, many long-established faculty members with special expertise in AI and machine learning left their academic positions to take full-time roles at tech companies.

This one-way migratory pattern can have lasting and undesirable impact on our top universities. Especially when entire research departments are hired away in bulk, as was the case when a top-ranking US university lost its entire cadre of world-class robotics researchers to a venture-funded enterprise.

At Amazon, we had a need for world-class AI talent, and we challenged ourselves to come up with a solution that would meet that need without causing enduring disruption to the vitality and success of the world’s best universities. We recognized that top AI and machine learning talent is drawn to industry by the sheer scale of the opportunity — to work on big, complex, real-world problems that can change the way we live and work. But in the absence of a more accommodating arrangement, many have to choose to walk away from their passion for nurturing the next generation of great AI talent.

Recognizing the need for new models of partnership that allow industry to tap into academic expertise while strengthening, rather than weakening, our universities, we applied an Amazonian approach by first developing a few tenets:

• Partnership models should strengthen all participants — industry, academia, and the government.
• Partnership models should be sustainable and accelerate scientific discovery and the advancement of each sector.
• Partnership models should provide demonstrable value to all constituents — society at large, faculty, students, university administrations, end customers, government entities, and industrial labs.

These tenets guided the development of the Amazon Scholars program, which allows academics to join Amazon in a flexible capacity, through part-time arrangements and sabbaticals. The program is designed for academics from universities around the globe who want to apply emerging and established research methods in practice, work with real-world systems, and deliver value to end users and society at large. Overcoming some of the hardest technical challenges requires the best expertise in the world to come together, and the Amazon Scholars program allows university faculty to help address these technical challenges at scale while maintaining their professional homes in academia.

To help universities develop the next generation of academic leaders, we expanded the Scholars program in 2020 by adding the Amazon Visiting Academics position, which gives pre-tenure-to-newly-tenured academics an opportunity to stay connected with emergent challenges and opportunities. Amazon is a unique place to measure the impact of new scientific ideas, given the diversity of our customers’ needs and expectations.

The value of the Amazon Scholars program is evidenced by the outstanding quality of academic talent who have joined Amazon as part of this program and its global appeal: we launched the Scholars program in 2019, and today we host more than 250 Scholars and Visiting Academics in 24 cities across the globe, from Los Angeles to Bangalore to Tel Aviv.

In 2019, we launched the University Hubs program to expand the access to academic partnerships beyond the faculty who are Amazon Scholars. The University Hubs program established multiyear commitments of gift and sponsored research funding, PhD fellowships, and community events to promote technology collaborations and community.

Having started with the first Hub at Columbia University, the program now spans five universities. Activity at the Hubs is comanaged by university administrators and Amazon program managers. In 2021, in partnership with Columbia University, we launched the Columbia-Amazon SURE Intern Fellowship program with the goal of providing students from underrepresented backgrounds an opportunity to contribute to cutting-edge research at some of the world’s top universities.

Following the resounding success of the pilot instance of the SURE program at Columbia in the summer of 2021, the SURE program quadrupled the number of supported interns in 2022 and expanded the program to two universities — Georgia Tech and the University of Southern California (USC). This year, the program expanded further, with the additions of Carnegie Mellon University and the University of California, Los Angeles (UCLA).

The impact of SURE

The alumna of the 2021 Columbia SURE Amazon cohort becomes the first Amazon MS Fellow at Columbia.

“The university hub in collaboration with Amazon provides a truly innovative framework for faculty and scholars to continue foundational research in academia or collaborate with industry practitioners in scaling technology breakthroughs to demonstrate real-world impacts,” said Shih-Fu Chang, dean of Columbia Engineering and Morris A. and Alma Schapiro Professor in the Departments of Electrical Engineering and Computer Science. “We are also extremely excited about the success of the SURE program in broadening access to the STEM pipeline by providing the combined resources from both industry and academia to training of students from diverse populations.”

Also in 2019, Amazon launched the Program on Fairness in AI in partnership with the US National Science Foundation (NSF), a $21 million initiative that, since its inception, has provided funding to 34 university-led teams and has produced a growing body of knowledge in the increasingly important area of fair and responsible AI. Amazon also contributes to the NSF’s National AI R&D Institutes program.

The complementary initiatives described above seek to broaden access to emerging technologies by strengthening partnerships across academia, industry, and government. These initiatives enable academia-based talent to harmonize their passions, life stations, and career ambitions by recognizing that academia and industry offer distinct paths to intellectual fulfillment. The success of these initiatives shows that achieving such harmony can enable industrial needs to be met while preserving — and even strengthening — the crucial pipeline of future talent.