How higher ed leaders can train students for an unpredictable AI future

AI is tearing up the old playbooks for how organizations train talent and gear up the next generation for a future that's more question mark than roadmap. For educational institutions, the mandate is particularly stark: How do you forge skilled graduates for careers still in the ether, using tools that could be museum pieces by the next academic year?

"It's terrific for people to think, 'I have a plan to be really good at AI.' But it's actually smarter to say, 'I'm just going to keep leveling myself up' and pay far less attention to the prognosticators," says Scott Marshall, President and CEO of Semester at Sea. He believes the key is to "listen to and understand AI, but don’t listen to predict its future—because frankly, no one truly can."

Embracing the unknown: "I'm not a techie," says Marshall, who oversees the unique shipboard global study abroad program. His approach to AI is practical, stemming from organizational realities: "We're a small organization; we have to be super resource efficient. AI can be one of those tools." He focuses on equipping his team and students to navigate the uncertainty, ensuring staff can "use the tool in a way that improves their work life and ensures that any privacy issues are dealt with."

Internal upskilling, careful adoption: This practical approach translates to careful AI adoption for his staff. Marshall explains that Semester at Sea, a Google platform company, has opened up tools like Gemini but with strict guidance: "We have to put guidelines in, like do not put any financial figures in there, do not put any marketing funnel data in there. That's not something we want Google eating up." Semester at Sea works to "keep the level across the organization moving up over time," and plans dedicated sessions so everyone understands the tools, policies, and their practical applications.

The student dilemma: "What does it mean for a college student, 20 years old, to live into this world in a way that they're proficient when no one, not a single person, can predict what it means in five or ten years from now?" Marshall asks, framing the central puzzle for educators. "I listen to a lot of podcasts, I talk to a lot of people, and I can assure you no one can predict it. There's a lot of experts, a lot of prognosticators, and they all think that they have the answer, but no one does."

Piloting the future: Instead of implementing broad, sweeping changes, Semester at Sea is experimenting. "We decided to partner with an Nvidia-certified instructor to offer a small cohort of our voyagers a Fundamentals of Deep Learning one-day course," Marshall highlights. This extracurricular program provides a certifiable skill set for about 25 self-selected students. "That is a pilot to see, is this the right way to address it?" he says. "We've really opted for recruiting this cohort who has self-declared interest and a belief that it will benefit them."

Curriculum agility: The rapid pace of AI makes integrating it into traditional, fixed curricula a genuine tightrope walk. "This is moving so fast in unknown directions," Marshall explains. "So to think way ahead of time and say, okay, we're going to hire faculty, there's a risk to that because you design a syllabus and within three months it could be out of date." Offering the deep learning course as an extracurricular, adjacent to the core curriculum, allows them to "follow the cadence of Nvidia updating these courses rather than the cadence of a general higher ed institution."

Building trust with human oversight: "Our IT manager has worked with our CFO to publish a set of guidelines," Marshall says, detailing how Semester at Sea builds trust around internal AI use. "Here's what we can use it for, and here's what we can't use it for." He also stresses the importance of maintaining an authentic organizational voice: "AI, whatever tool, cannot fully represent our unique, authentic voice. Everything that comes out has to have multiple reviews."

Experiential learning first: While technology is a tool, Marshall is quick to emphasize Semester at Sea's commitment to its core, hands-on mission. "Our model is experiential community, ship-based learning, going to countries. It's not counter-tech; it just leads with experiential learning," he clarifies. "We use an LMS, we have Starlink on the ship, we have dual authentication. We have all these kind of things that are really tech-based. But we're community and experiential learning first." This foundational context, including the reality of limited internet bandwidth on voyages, guides their tech integration.