AI, Storytelling, and the HBS Case Method
Inés (2013) by Spanish artist Jaume Plensa near Aldrich Hall; photo by Nicholas Ng
I have been personally and rudely starved of unlimited “free” access to academic articles since the day I earned my undergraduate degree approximately eighty-four years ago. As a result, I am going to tie together some disparate ideas on AI, evolution, and the HBS case method. I’ll cite my sources and, because 1) my essays are no longer graded by people with PhDs and 2) I chose the way of business instead of academia, no one can tell me I’m wrong! Yippee!!!
Unusually, Harvard Business School emphasizes and enforces a classroom free of technology: no phones, no computers, and no (gasp) AI. This feature—moreso in my opinion than course content or volume—is the single most consequential and important part of this experience.
Let’s talk about why.
Storytelling
Evolutionary psychologist Dr. Michelle Scalise Sugiyama described storytelling—narrative—as a “species-typical, reliably developing, complex cognitive process whose design is unlikely to have emerged by chance. Moreover [...] narrative content is consistent across widely divergent cultures.”
In other words, humans:
Be telling stories
Been telling stories
And, across disparate and distinctly different groups that have their own established rules, language, and norms, their stories share themes.
Sugiyama continues: “By substituting verbal representations for potentially costly first-hand experience, narrative enables an individual to safely and efficiently acquire information pertinent to the pursuit of fitness in local habitats” (Sugiyama, Oral Storytelling…).
Telling stories—as opposed to living them—demands fewer resources, or costs less—and serves as a mechanism that enables our species’ continued survival. Please note: “pursuit of fitness” in this context is NOT looksmaxxing—plz someone save us from ourselves.
Explains Sugiyama: “all societies practice some form of storytelling [...]. Although narrative skill varies from person to person, the ability to generate and process narrative is not limited to the exceptionally intelligent.”
So:
No, like, LITERALLY everyone tells stories
Some folks are better storytellers than others
But, to understand stories, you don’t actually have to be smart at all!
Bietti et al speculates that “the specific adaptive value of storytelling lies in making sense of non-routine, uncertain, or novel situations, thereby enabling the collaborative development of previously acquired skills and knowledge” (Bietti 2019) (I REALLY forgot how to cite articles). Bietti emphasizes that it is in the process itself of wrapping our minds around complex, chaotic situations that we draw upon, challenge, and develop what we already know. It is then in the act of sharing this experience with others that the survival value compounds, strengthening the “intra-group identity and clarifying intergroup relations.” Telling a story, in other words, doesn’t just serve to inform a chronological series of events that may or may not have happened—some call it fiction and non-fiction—it teaches the audience who they are in relation to the story, the person sharing it, and their fellow audience members. Storytelling begets social bonding; social bonding begets survival.
HBS
Stick 90-something people in a room from 30-something countries, assign them a letter to coalesce their identity around, and invite roughly six people to tell a bunch of stories for approximately four months. Rinse and repeat after a short break; at the end of THOSE four months, depart for a different country and endeavor to apply your newly learned—oops, wait, sorry, wrote that before FIELD got cut…moving on!
Boom: you’ve completed your first year of Harvard Business School.
Now, I am bastardizing the complexity of the structure of this institution. But isn’t it also true that, if The Venerable Harvard Case Method had a different, dumber brand manager, it could just as easily be called “Storytime”?
Generations of students bemoan the expectations of HBS’ approach to the case method: high volume, high technicality, and high expectations to not say something stupid in front of your peers. Professors recall Case X, and proceed to solemnly, generously warn you about the soon-to-follow Case Y. “The best resource to help you get through,” they nod slowly, “is each other. Don’t rely on a robot to teach you.” They must know: They Are Professor.
For them, the strength of the program is in students cyclically teaching themselves and each other.
I believe that this holds: a student willing to explain a new or complicated subject with clarity, concision and patience to their fellow classmates is worth her weight in gold.
However, with AI closing the gap for a student to move from zero understanding to at least a foundation, I believe that this deep interaction between students and faculty—what I would argue is effectively co-storytelling—is where the value of this program might hold over time. Because, at their core, that’s all cases really are: a clean definition of a complex situation, complete with a protagonist, conflict, and character arc. Rudely, they also include numbers, but begrudgingly I submit that these numbers are their own form of storytelling.
AI, storytelling and the in-person premium
An angel loses its wings every time another LinkedIn post claims “and THIS is what no one is talking about.” AI slop surrounds us from all sides, impacting not just individual experience but overall algorithmic culture. Perhaps in response, we’re seeing a rise of the “authenticity economy,” where increasingly the premium becomes in-person third-spaces and plain, unedited human messiness: bars and concerts are designed explicitly to be a no-phone environment, Partiful invitations to “eat an apple together” go viral in DC, and wedding hosts ask their guests to please-we-literally-hired-a-professional-photographer, put your phones away.
What AI does well in an academic environment is to help students with near-zero context on financial analyses (cough) move to some foundational understanding. A hypothetical interaction might look hypothetically something like “why…do…we…even…need…accounting…in…the…first…place…who…came…up…with…this…and…why…did…they…hate…themselves…and...also…me...send.” Hypothetical long night.
With defined parameters, AI adeptly explains dense, complex, and long-recorded concepts in 17 different ways until the student can at least delude themselves into believing they kind of sort of get at least part of it. But, inevitably, where the concept solidifies is in-person: they’ve read the case, which buttresses the core concepts looking to be taught; they’ve engaged with fundamental questions without eliciting frustration or projecting their confusion onto someone else; they’ve started exactly where they already are and moved their own understanding as far as they can within the time constraint afforded to them.
Now, in class, they hear it again, and this time it’s guided by Professor Storyteller and co-shared with Student Audience Members. It is a proactive-meets-active experience, where raising a hand is only part of the game: you must continue to engage, every day, in every class. For at least a few hours every day, there is no logging off, no muting—yourself or others’—and no unfollowing. There is no responding to notifications, asking the internet for that fact you can’t quite recall, or watching that viral video in lieu of daydreaming.
“Data privacy” is today’s great oxymoron, and many of us have cynically, tiredly accepted that the CIA or Amazon Alexa or some foreign government already knows everything about us.
For this Harbus tech editor, the HBS classroom might be the first tech-free environment I’ve sat through since the pink Razr flip phone first infiltrated my elementary school classroom walls. To not be giving a Mag 7 some form of my data for a few hours every day feels like an odd form of withdrawal; it also feels like coming up for air. The experience of the classroom remains largely protected as just…ours. Without the relative aids and handicaps of technology, we’re challenged, out loud and in front of ninety-ish strangers-to-colleagues-to-friends, to confront and acknowledge our beliefs, failures, and blindspots—to move not just around but through the initial impression, fear, and discomfort and to keep engaging, failing, and showing up. Technology is frictionless by design; here, we are challenged to seek friction.
The stories that AI can tell will always be a relative average of all stories it has calculated: there is nothing in the pace, structure, or delivery that was informed by a genuinely experienced set of circumstances, environments, or contexts. Nor do its “stories” serve any evolutionary purpose: there is no storyteller, there are no audience members, and there is no distinct identity that the shared experience of that story informs or enforces. In speaking to everyone, it speaks to no one, the value of its words approaching zero.
Even as AI giants gobble up old copies of books in an apparent attempt to rectify or improve this, the algorithm will never be able to say something genuinely surprising.
And this, dear reader, is where educational environments hold value over time. Learning technical concepts is no longer the largest barrier—today, it’s whether we’re willing and able to put ourselves and each other in an environment that places our mind, the most adaptive muscle, alongside other beings able to feel, laugh, and change.
And the best news? HBS does not hold a monopoly on this: our mission, should we choose to accept it, is to seek novelty such that it allows us to create, and to tell each other our stories. This is where we built our humanity, and this is how our humanity evolves.
Kacey Sorenson (MBA ‘28) is from the San Francisco Bay Area. She’s a washed up pre-med student, having studied English and Environmental Science with an intersectional focus on the cyclical influence of humanity on the physical environment and vice versa. Over her career, she’s added the language of technology, finance, and commercial insurance to her vocabulary, scaling Nirvana Insurance—an AI-native fintech startup— from Series A through D and a $1.5B valuation in four years. She misses using an em dash, and humbly aspires to use them here without reproach.
Comments