In today’s column, I address an increasingly voiced concern that our universities are losing ground on inventing the leading edge of AI advances due to a lack of needed computing resources and the luring of top AI talent out of academia and into the private sector. The consequences are dismal and disturbing if this trend continues unabated. A vital engine that drives forward innovative approaches to AI will indubitably run out of gas and sputter to a low point, undercutting the vaunted and necessary pursuit of pure foundational knowledge that drives the future of AI.
Let’s talk about it.
This analysis of the advent of innovative AI breakthroughs is part of my ongoing Forbes column coverage on the latest trends underlying AI advancements (see the link here).
Noteworthy Wake-Up Call That Must Be Heard
I recently attended an excellent seminar featuring Russell Wald, Executive Director of the Stanford Institute for Human-Centered Artificial Intelligence (HAI), and I will be noting various points mentioned in his invigorating talk as part of my discussion below. The HAI seminar took place on January 15, 2025, occurring at the Stanford University campus and covered the emerging qualms of an industry-academic divide in AI (for information about HAI’s events, activities, and programs, see the link here).
There is also a helpful and wholly pertinent white paper by HAI entitled “Expanding Academia’s Role in Public Sector AI” that provides a close look at the growing issue (see the link here), which is co-authored by Kevin Klyman, Aaron Bao, Caroline Meinhardt, Daniel Zhang, Elena Cryst, Russell Wald, and Fei-Fei Li (posted online December 4, 2024).
Here are some salient points from the HAI research paper (excerpts):
- “Building and deploying AI systems has become hugely resource-intensive, often requiring billions of dollars in investment, custom supercomputing clusters, and enormous datasets containing much of the available data on the internet.”
- “This shift has created a significant power imbalance, where academic talent and government support flows to private companies that now produce the vast majority of the world’s most powerful AI systems.”
- “This disparity undermines not only the future of academic research but also the potential for a public sector AI ecosystem that serves the public interest.”
- “Academia must play a leading role in developing frontier AI to ensure that we can understand and safely deploy the technology.”
The Triad That Is Silently Eroding AI Academia Pursuits
As a former computer science professor in AI who also served as director of an AI lab, I can directly and personally attest to the rising phenomenon that is sorely impacting our universities when it comes to the vital pursuit of AI advancements. I might add that as a founder of AI startups, I undoubtedly have participated in the very acts now recognized as a kind of unintended brain drain on AI academia. Yes, I ended up shifting into the private sector where computing resources were more readily plentiful, and yes, I hired many of my top graduate students upon the completion of their degrees. I sheepishly admit this.
It is an easy trap to fall into.
Amid the many factors impacting academic AI pursuits, I tend to concisely shorten the list to a triad of these three:
- (1) Insufficient computing. The advancement of AI at the leading frontier edge involves massive-scale GPU computing that is typically not available on campuses and is exorbitantly expensive to utilize via commercial cloud services. Academic AI researchers and their students are unable to move the needle on AI since they are hamstrung by a lack of iron to run their AI endeavors.
- (2) Faculty exiting to industry. Private industry is eager to snap up AI faculty, offering big pay packages, seemingly infinite computing resources, and appeal to the gut-felt desire to make demonstrative progress in AI. A downside is that the AI breakthroughs often then land in the hands of the employer and become a proprietary asset. University-driven AI research tends to be open source and envisioned to widely spread innovations publicly and openly.
- (3) Graduates lured away from academia. Any student who slogs their way through several years to get a computer science degree with an AI emphasis is going to be fully tempted to say goodbye to the academic world once they complete their degree. This means that, for example, Ph.D. graduates are less likely to seek faculty positions. Their eyes are attracted to private industry instead.
A vicious cycle consumes the triad and leads to a downward spiral.
Here’s what I mean.
Faculty that can’t get needed computing will exit to industry. Exiting to industry leads to less faculty able to cover AI classes, plus not enough expertise in the halls of academia to sufficiently guide students on cutting-edge AI research. Students who do manage to complete their AI-related degrees instantly jump into industry efforts. This lessens the pool of teaching assistants and research assistants. And hopes of bringing in freshly minted AI doctorates are dashed as the pipeline is spewing directly into private industry and not back into academia.
Rinse and repeat.
The Problem Has Societal Repercussions
A thin glance might suggest that this is just something of a narrow consideration and doesn’t have larger ramifications. Sorry to say, it does have humongous consequences.
I’ve covered extensively the belief that AI constitutes a crucial element to the future of our economy and that countries are going to be waging a geopolitical superpower dominance based on the AI they wield, see my analyses at the link here and the link here. In the United States, the rinse-and-repeat cycle is taking our country in a direction that none of us are going to like. Commercial humdrum use of AI will be the mainstay, while breathtaking breakthroughs in AI will be discovered in other countries but not here. That’s bad for us.
A retort by some is that AI is moving toward smaller models anyway, shifting from large language models (LLMs) to small language models (SLMs), as I’ve detailed at the link here. Or that the LLMs are being compressed so that they don’t require large-scale computing, such as the advent of 1-bit LLMs (see my analysis at the link here).
Let me set the record straight, we aren’t going to somehow satiate university AI research pursuits by squeezing down the computing needs. Sure, more can be done as good progress evolves in slimming the resource requirements for AI. But if you are playing at the topmost of AI pinnacle innovations, tiny iron isn’t going to get you there.
Play big or go home.
Smart Ways To Swing Upward
The good news is that we are still in the early days of this semblance of decay, thus, the ship can in fact be set upright. Do not let despair or doom and gloom put you into a catatonic funk.
Some elbow grease and astute prudence can get things back in proper gear.
As indicated by Russell Wald during his presentation, now is the time to transform the emerging industry-academia AI divide toward becoming a society-saving industry-academia partnership on AI. The private sector and the public sector have a substantial role in shoring up the gap in computational resources that are desperately needed in our universities.
In addition, since it takes a village to truly advance AI, the idea of “team science” suggests that university AI efforts should be collaboratively intermixed with commercial and governmental AI leading-edge projects. This could shift the brain drain in a manner that moves us toward a brain gain.
A final thought for now.
I recently noted in my column that Generation Beta is getting underway — it’s the name of the generation for youngsters born in 2025 through 2039, so it started this month of January 2025. My eye-opening assertion is that Generation Beta is going to go beyond being a digital native generation, they will be what I refer to as an AI-natural generation. They will grow up with a natural and intrinsic familiarity with AI, occurring fully throughout their lives, since AI is going to be pervasive and ubiquitous (see my predictions on this AI-naturalism phenomena at the link here).
We must get our act together to first stop and then valiantly turn around the inching forward industry-academy AI divide. Do it for the children. Do it for Generation Beta. Deliver to their generation an ongoing persistent and valued industry-academy partnership on AI. Heed the illustrious words of John F. Kennedy that still principally apply to this very day: “We have the power to make this the best generation of mankind in the history of the world or make it the last.”
Let’s ensure that we aim upward and lay a strong and enduring foundation that will ensure success for the generations here now and that follow us. I’m earnestly confident we can.
