Incentivizing the Mental Gym: Working Out Your Brain Before Supplementing with AI

Idea 31 for 2025

A few weeks ago, I listened to a podcast recorded in September on the future of higher education with the rapid ascent of generative AI. Published as part of The Globe & Mail's Machines Like Us podcast series, it featured a conversation moderated by Taylor Owen.

Created by Gemini by Google. October 31 2025.

It featured Conor Grennan, the chief AI architect at NYU’s Stern School of Business, where he’s helping students and educators embrace AI, and Niall Ferguson, a senior fellow at Stanford and Harvard, the co-founder of the University of Austin.

Ferguson lately, been making the opposite argument: that if universities are to survive, they largely need to ban AI from the classroom; however, in this conversation, the two were actually closer together, arguing that, just like, we incentivize working out our bodies in our culture, especially for younger people who populate our universities, we should also consider incentivizing the working out of our brains. As a consequence, we need to rethinking thinking and learning and conventional models of higher education (with Grennan and Ferguson's focus on the American (elite) university.

Here, with AI assistance, are some of their key arguments:

1. How Students are Using AI in Higher Education

The speed of AI adoption on campus has been astonishing. Only months after the release of ChatGPT, a survey found that 90% of college students were already using it. Today, that number is likely even higher.

Students are not just using AI to write essays; they are engaging it for every stage of the academic process, from generating ideas and conducting research to doing their assigned readings. The problem, as Ferguson sees it, is that the vast majority of this use is to "cut corners". Connor Grennan observed that students are highly incentivized to delegate because "the only thing that matters is grades," and a paper generated quickly by ChatGPT "helps their future prospects in life."

In essence, students are using the tools "to think for them," a trend that shifts the educational focus away from genuine inquiry toward maximizing efficiency and grades.

The most immediate danger of this cognitive offloading is the erosion of core intellectual skills. Ferguson argues that if students delegate the work of reading, thinking, and writing to AI, "you’re not learning to do those things." Writing, in particular, is viewed not merely as an output but as a vital form of intellectual "muscle flexing." By allowing AI to perform these demanding tasks, students risk incurring "cognitive offloading" and atrophy.

Ferguson uses the powerful analogy of the "obese brain." Just as we commit to physical exercise to stay fit, intellectual work is necessary to build "very, very fit brains" that can absorb complex data, think analytically, and communicate inferences. The consequence of constant reliance on AI, according to the speakers, is a generation that can produce work, "but not necessarily original thought," potentially plunging humanity into a "new dark age."

Furthermore, Grennan points out that students who skip the work do not develop the "muscle" to know "what quality looks like." This lack of foundational knowledge prevents them from being able to spot "slop," outdated information, or believable AI "hallucinations" (false outputs) when they enter the professional world.

3. The Mental Gym and the Cloister and Starship Analogy

To counteract this danger and re-incentivize the usage of human brains, Ferguson proposes the "Cloister and Starship" model. This vision is a new structure for education that intentionally separates mandatory intellectual training from AI-augmented learning:

• The Cloister (Foundation): This phase involves a mandatory period where students are denied access to AI, essentially forcing them to enter a "mental gym." This protected time is dedicated to building core competencies—reading, thinking, writing, and foundational mathematics—using just their brains. This phase is designed to instill the mental discipline required for independent thought.

• The Starship (Augmentation): Once equipped with this essential mental discipline, students can be "unleashed and let into the starship to use large language models." AI then becomes, as Grennan notes, potentially "the greatest teaching and learning tool ever created," capable of providing a bespoke, one-on-one tutorial experience that accelerates learning.

4. In Closing: Incentivizing the Mental Gym, or Working Out the Brain

The key to incentivizing this model lies in connecting mental fitness to real-world success. Firstly, only the "super fit" brain will be capable of using AI optimally—by asking the right questions, providing the right context, and, most importantly, evaluating the AI's output. Secondly, the job market is already providing external pressure: elite employers are moving past the hollow signal of a high GPA and are instead using new challenges to seek out people with "real intellectual brilliance."

The ultimate message for learning advisors is clear: the focus must shift from grading the output to testing the cognitive input. We must establish guardrails, like supervised exams and oral assessments, to make delegation impossible, thereby forcing students to engage the cognitive engine that AI is currently encouraging them to switch off.