Last week, four major research universities announced sweeping budget restrictions. Northwestern cut non-personnel budgets by 10%. MIT froze hiring. Washington State began planning for what their provost carefully termed “near-term futures where reductions are necessary.” These moves come as the Trump administration’s new policies threaten to slash federal research funding, with some institutions facing losses of over $150 million annually.
In times of constraint, emerging technology investments – particularly in artificial intelligence (AI) – often become prime targets for reduction. After all, AI can seem like tomorrow’s problem when today’s lights need keeping on. But before we reach for these seemingly logical cuts, we need to examine a fundamental shift happening in our workforce that no budget cut can stop.
The Generational Imperative
Consider this: In just three years, the class of 2028 will graduate college. These students will have never known higher education without AI tools. Their entire academic experience – from research to writing to problem-solving – will have been shaped by AI collaboration. Four years later, we’ll welcome professionals who have used AI throughout both high school and college.
This isn’t speculation about some distant future. These students are in our universities right now, developing personal workflows and approaches that integrate AI as naturally as previous generations integrated calculators or computers. They’re not just using AI – they’re reinventing how their work gets done.
What happens when these AI-native professionals enter organizations that have pulled back from AI investment? Imagine telling a 2032 graduate, “We don’t use AI here” – it would be like telling a graduate today, “We don’t use computers for spreadsheets.” The disconnect isn’t just technological; it’s fundamental to how these professionals approach problem-solving.
But the risk of falling behind goes beyond just attracting talent. History shows us that early technology decisions can shape entire industries for generations to come.
The Power of First-Mover Advantage
To understand the long-term implications of today’s technology decisions, consider the keyboard under your fingertips. In the 1870s, Christopher Sholes designed the QWERTY layout to solve a specific problem: mechanical typewriter keys jamming when typists worked too quickly. His solution? Deliberately make typing harder by separating commonly used letter combinations.
Fast forward 150 years. We type on keyboards in virtual reality where no physical keys exist to jam. Yet the QWERTY layout persists, even in these virtual worlds. The layout’s dominance isn’t about superiority – it’s about momentum (or path dependence in technical parlance). Early choices in technology create patterns that can persist for generations, even when the original constraints disappear.
This persistence of early technology choices should give us pause as we consider AI investment cuts. The organizations that maintain strategic AI investment during this downturn won’t just preserve capabilities – they’ll help shape how their entire industry uses AI. Those who wait may find themselves forced to adopt standards and practices set by others, potentially for decades to come.
Unexpected Innovation: The BreadBot Lesson
But maintaining AI investment doesn’t always mean massive expenditure. Sometimes, the most valuable innovations emerge from unexpected places and modest beginnings.
In 2007, Japanese software developer Hisashi Kambe took on what seemed like a modest project: helping bakeries identify pastries at checkout. The challenge proved so complex it nearly bankrupted his company. But by 2013, his team’s persistence paid off with BakeryScan, a successful AI system for identifying baked goods.
The story could have ended there – a nice case study in retail automation. But in 2017, a doctor at Kyoto’s Louis Pasteur Center for Medical Research noticed something remarkable: cancer cells under a microscope looked surprisingly similar to bread. This observation led to Cyto-AIscan, a tool that helped oncologists identify cancerous cells more quickly and accurately than traditional methods.
Think about that moment of discovery. A medical researcher, probably enjoying his morning coffee, happens to see a TV segment about bakery automation. Most viewers might have changed the channel. But because this doctor understood both the medical challenge and could recognize the AI potential, he saw a connection that transformed cancer detection.
This is the hidden power of maintaining AI investment across an organization. Innovation doesn’t just flow from the obvious places like the R&D department or the product team. It emerges from unexpected intersections, from moments when someone in marketing sees an AI solution that could revolutionize product development, or when a facilities manager recognizes how an AI system could transform sustainability efforts.
But these connections only become visible when people throughout your organization understand AI’s possibilities. If you cut AI investment and expertise now, you’re not just reducing current capabilities, you’re potentially blinding your organization to these transformative intersections. You risk creating an environment where the next BakeryScan insight goes unnoticed, where game-changing connections remain hidden in plain sight.
Yet maintaining this commitment requires more than just preserving budgets or spotting innovative connections. It requires a deep understanding of how technology expertise, once lost, becomes difficult to rebuild. This brings us to an unexpected source of wisdom on the matter: Sun Tzu’s Art of War.
Sun Tzu's Warning
This observation from Sun Tzu might seem obvious – of course your army needs water. But this passage begs a question: why would Sun Tzu bother including this obvious information? That’s where our lesson lies. Sun Tzu is telling us, between the lines, that the leaders of his time were dilettantes, often disconnected from operational reality. Many had never fought in battle and did not understand how quickly thirst can decimate an army.
Today, we risk creating similar disconnects through aggressive budget cuts. In 2022, The Washington Post published a brilliant op-ed exposing the legacy stacks of the IRS. At the time, the IRS had been starved of funding for decades. As a result, they were still using technology from the 1970s, built by a company that no longer exists, maintained by a single individual, where the newest component is Windows XP from 2001 – an operating system old enough to drink its own woes away.
This isn’t just about outdated technology. It’s about the gradual erosion of institutional knowledge. When organizations cut technology investments, they often lose the expertise needed to maintain and upgrade these systems effectively. The problems compound over time: documentation becomes outdated, workarounds multiply, and the pool of people who understand the system shrinks.
Now imagine this scenario playing out with your AI systems. Today’s cutting-edge solution becomes tomorrow’s legacy system. Without continued investment in both technology and expertise, organizations risk creating their own future crises. Who will maintain your AI systems in five years? Ten years? How will you integrate new capabilities into aging infrastructure?
The cost of rebuilding lost technical expertise often far exceeds the initial savings from budget cuts. Just as Sun Tzu understood that military leaders needed practical knowledge of battlefield conditions, today’s organizations need to maintain their technological competency – especially in emerging fields like AI. This challenge becomes even more acute when we consider how rapidly AI capabilities can evolve, potentially leaving organizations not just behind, but permanently disadvantaged in their markets.
Practical Steps Forward
The challenges we’ve explored – from the incoming AI-native workforce to the risks of technical debt – might seem daunting, especially during budget constraints. However, organizations can take these strategic steps to maintain their AI momentum without requiring massive investments:
First, focus on targeted applications that create immediate value while building future capabilities. Like the BakeryScan example, start with specific, well-defined problems where AI can demonstrate clear return on investment. These focused investments often reveal unexpected applications across other domains, but only if you maintain the expertise to recognize these opportunities.
Second, invest in knowledge preservation and transfer. Sun Tzu’s lesson about maintaining operational expertise applies directly here. Document your AI systems and processes thoroughly as you build them. Create opportunities for cross-functional teams to understand AI capabilities. Build internal communities of practice that can sustain technical knowledge even during lean times. Documentation and cross-training are wildly cheap relative to the long-term cost of knowledge erosion.
Third, prioritize flexibility in your AI infrastructure. The QWERTY keyboard reminds us that today’s solutions can become tomorrow’s constraints. Design your AI systems with modularity in mind, making it easier to adapt as technology evolves. This approach often costs less in the long run than trying to modernize rigid, monolithic systems.
Finally, maintain a strategic view of talent development. The class of 2028 is already in college, developing AI-enabled workflows. Organizations need to begin preparing now for this shift in workforce expectations and capabilities. This might mean preserving key AI training programs even during budget cuts, or finding creative ways to keep AI expertise in-house.
Closing Thoughts
The pressure to cut AI investments during budget constraints is understandable. But as we’ve seen, the long-term costs of these cuts often far exceed the immediate savings. From the persistence of early technology choices to the hidden opportunities for cross-functional innovation, from the risks of eroding expertise to the incoming wave of AI-native professionals – the strategic importance of maintaining AI capabilities has never been clearer.
Organizations that maintain thoughtful AI investment during this downturn won’t just preserve their current capabilities. They’ll help shape how their entire industry uses AI, spot unexpected opportunities for innovation, and be ready for the AI-native workforce that’s already on its way.
The future is coming, whether we’re ready or not. The question isn’t whether to invest in AI, but how to do it strategically and sustainably, even in challenging times.
About the Author:
James Villarrubia
CTO, Applied AI Expert, Presidential Innovation Fellow at NASA
James Villarrubia is an accomplished CTO and innovator dedicated to fostering positive social change through engineering and applied artificial intelligence. As the Presidential Innovation Fellow for NASA Aeronautics, James is helping develop sustainable, next-generation aviation and AI solutions. On behalf of NASA, he co-hosts the Ecosystemic Futures podcast which delves into technological advancements poised to shape the future of society. Prior to his role at NASA, James served as a tech executive and advisor for various startups, spanning healthcare, cryptography, workplace equity, and online education. James brings a wealth of experience in policy and portfolio management, having worked in Washington, DC at the White House, DOJ, and DOD.
