Big Tech, You Need Academia. Speak Up!

The current U.S. administration has launched a war^a on academia. On Feb. 7, 2025, the U.S. National Institute of Health announced^b a 15% limit on indirect research costs for new and existing grants. Indirect costs, or, more accurately, facility and administration expenses, support research but cannot be directly attributed to a specific project, such as lab infrastructure, utilities, and administrative support. These are real costs; the limit, which has since been suspended by courts, is a severe blow to biomedical research in the U.S.

Beyond expanding this limit to other agencies, such as the National Science Foundation (NSF), the administration is also reportedly considering slashing NSF’s annual budget from approximately US$9 billion down to about US$3–$4 billion. This would deal a devastating blow to academic U.S. research, especially computing research. As stated^c by the Computing Research Association (CRA), “NSF budget cuts would put the future of U.S. innovation and security at risk.”

But U.S. academic computing research has been in trouble for several years. Enrollments of computing majors have been climbing up^d since 2010, which triggered a wave of faculty hiring. According to the CRA Taulbee Survey,^e faculty size in research-intensive departments grew about 60% since 2010. But NSF’s budget in constant dollar seems^f to have stagnated since 2010. And NSF plays a critical role in computing research. As stated by CRA: “While NSF funds over 50% of basic research at U.S. institutions, NSF funds nearly 80% of fundamental computing research at U.S. institutions, making it the single most important agency for sustaining advances in computing—the foundational technology behind artificial intelligence, cybersecurity, quantum computing, and other areas critical to economic growth and national defense.”

Big Tech may dismiss these concerns regarding the future of U.S. academic computing research. After all, Big Tech consists of six corporations with more than US$1 trillion in market capitalization each; their research budgets dwarf governments research budgets in computing. Furthermore, industrial researchers have access to large-scale data and computing that academic researchers can only dream of. While the NSF’s Directorate for Computing and Information Science and Engineering has an annual budget of about US$1 billion, DeepMind, a subsidiary of Alphabet focusing on AI research, has an annual budget of around US$2 billion.

But dismissing academic computing research as irrelevant, given the size of Big Tech, would be a grave mistake. As I argued^g in 2019, research is a long game. Consider the current wave of generative AI. While this wave, going back to the mid-2010s, is driven almost exclusively by industrial research, it builds on a 70-year academic effort in neural computing. We must remind ourselves of the concerns of Abraham Flexner, founder of the Institute for Advanced Study in Princeton. In The Usefulness of Useless Knowledge, Flexner explored the dangerous tendency to forgo pure curiosity in favor of alleged pragmatism.

In the previous century, monopolistic corporations such as AT&T and IBM were able to sustain research laboratories where curiosity-driven research was encouraged, leading even to Nobel prizes in physics. But already in 1995, Andrew Odlyzko, then at AT&T Bell Labs, wrote about the “Decline of Unfettered Research.”^h Only academic research is keeping curiosity-driven research alive today.

As evidence for the value of curiosity-driven research in computing, I offer the Tire-Tracks Diagram,ⁱ a graphical illustration of how federally funded university research and industrial research and development precede the emergence of large IT industries by decades. The diagram shows eight technologies whose values have surpassed USD$1B. From the eight technologies, only one clearly traces its birth to industrial research (relational databases.)

Research-intensive academic departments, however, produce more than just original ideas. Generally, academic research is carried out by doctoral students, supervised by faculty members, and supported mostly by federal funding. According to the CRA Taulbee Survey, in 2023 U.S. computer science (CS) departments graduated 1,883 doctorate holders, from which 870 went to industry. Obviously, industry values people who can formulate and solve problems. Doctorate holders play a key role in tech; case in point, Amazon’s Automated Reasoning Group.^j The most notable example, however, is Google, which was founded by Sergey Brin and Larry Page—two CS doctoral students at Stanford University.

It is unlikely that Elon Musk reads Communications. Even if I send him a copy of this column, will it change his mind? I am not sure. Perhaps, however, he would listen to his fellow Big Tech CEOs. Big Tech, you need academia. Please speak up!