A little over a century ago, medical education in America was transformed. Before the reforms triggered by the 1910 Flexner Report, you could become a doctor largely by listening to lectures and reading textbooks. Afterward, aspiring physicians learned medicine by practicing it. They learned from more experienced doctors on rounds and in residencies, and eventually with patients and surgical simulators. Young doctors could make mistakes and be corrected before anybody died (with some engaging in what Matt Beane calls “shadow learning” to get even more practice). Today, no one would board a plane flown by a pilot who had only read about flying, or submit to surgery performed by someone who had merely discussed it in a seminar.
Yet that is, more or less, how we still train the people who run our companies. Business education remains organized around lectures, the case method and the odd group project. Cases, in particular, are problematic from the perspective of learning how things in business actually happened. We forget that for a company to agree to have a researcher write about them, the company has the right to edit the case. As the old line goes, history is written by the victors.
What a simulation teaches that a case cannot
In a simulation, in contrast to a case, the outcome is open-ended. Participants might discover that their clever pricing move has destroyed their working capital buffer, or that the marketing investment they championed starved the operations capacity they needed to fulfill the demand it created.
This is the kind of experience that improves judgment. Employers prize people who can make high-quality decisions under pressure, with incomplete information, and the understanding of how choices ripple across an entire system. It is also the capability that is essential in a world where competitive advantages are transient and leaders must continuously test assumptions, learn cheaply, and redirect. Simulations are discovery-driven planning in miniature: make assumptions explicit, test them against an unfolding reality, and adjust before real money is spent.
We do something like this in my Columbia Business School Executive Education program, Leading Strategic Growth and Change. Every participant prepares a personal case, which is a live strategic challenge from their own organization, and works on it throughout our week together. The frameworks taught by our faculty are immediately stress-tested against a real situation. It simulates their real-life dilemmas, run at accelerated speed in a room full of peers. Participants don’t leave with notes about what they learned. They leave with a plan for action.
Why this is the moment
AI makes simulations that once took years to build easy to generate, customize, and make adaptive at a fraction of the cost. Scenarios can respond dynamically to participants’ choices. Every student can face a different market, a different crisis, a different negotiation counterpart, and receive rich feedback without an army of teaching assistants. It’s classic disruption, in which something that was once difficult becomes easy and something that was once expensive becomes cheap (as Scott Anthony explains). The institutional momentum is picking up, even as there is an overall decline in interest in business education, as evidenced by the number of applicants taking the admissions test.
A January 2026 report tracking 48 business schools notes that “the transformation of business education through artificial intelligence has accelerated remarkably since the first iteration of this report in July 2025. What was then an inflection point has become an imperative . . . The pace of change has compressed what might have taken years into months, with institutions making decisive moves that reshape curriculum, pedagogy, faculty development, and institutional strategy.” Rewiring institutions to capitalize on the new pedagogy, however, is going to be uncomfortable.
A simulation has no department
A well-built simulation is inherently cross-functional. The departmental silos that represent home base for academics aren’t set up in a way that makes these interrelationships easy to identify. Departmental silos become obstacles if the goal is to see the organization operating as a systemic whole. It’s useful to remember that these structures were never created to inform action in the real world. They were created in response to charges that business schools didn’t have the right amount of rigor to be legitimate constituents of universities.
As the late Carter A. Daniel (formerly of Rutgers) noted, “In October 1959, both the Ford Foundation and the Carnegie Corporation for the Advancement of Education issued reports very critical of America’s business schools. Five criticisms dominated their attacks: (1) poor students, (2) untrained faculty, (3) unintellectual curriculum, (4) lack of theoretical research, and (5) unclear mission. Just about everybody jumped on the let’s-humiliate-the-business-schools bandwagon. The New York Times said the reports “assailed” the schools; Business Week said they “knocked the stuffing out of business schools”; and another observer said they “sent the schools reeling in disgrace.”
Daniel argues that the writers of these reports “betrayed an utter ignorance of business and business education. The authors, all traditional liberal-arts academics, wanted “A” students, PhD-holding faculties, big libraries, and lots of theoretical research, because those are the things that qualify as “good” in traditional academic departments like history and philosophy.” The reports also led to the creation of two “industries” in and of themselves: the PhD in business (rather than a practice-focused DBA or a degree from a core academic discipline) and the academic journals in which these PhDs publish their research.
Critiques
Critiques of this system are many. In a scathing article authored by George Washington University’s Michael Harmon, he observes that, “Competition for status among U.S. business schools has obliterated any evident connection between research productivity and the furtherance of any praiseworthy social, practical, or intellectual values.” Don Hambrick, when he was President of the Academy of Management in 1993, gave an equally critical address called “What if the Academy Actually Mattered?” Richard Bettis pointed out that the reward structures in most business schools favor statistical findings that are often flawed or do not mean what the researcher claims, in a similarly negative piece called “The search for asterisks: Compromised statistical tests and flawed theories.” Even my co-author and mentor, Ian MacMillan, had little positive to say about the relevance of academic research in the management field.
With the advent of AI in the classroom, we may be at the start of an inflection point. When an employer can see evidence of what a person can do, not just what degree they possess, the signal represented by a degree isn’t necessary. I’ve written about the potential for breaking up the degree stranglehold elsewhere.
A better approach may be staring us right in the face. Medical schools long ago reorganized their training around organ systems, clinical rotations, and integrated problem-based learning rather than on disciplines alone. Business problems are similar. Business schools serious about AI-enhanced learning will increasingly need to organize their teaching around decision domains and building models, rather than revisiting the frameworks of a 1970s textbook.
The faculty reward system promises to be a herculean obstacle to changing how business schools work. Who, in a top-tier business school, would invest in AI and simulations in the classroom? For a junior professor, that work would be career suicide. It would count as teaching innovations, rather than contributions to research. If schools do not create a track that rewards the design of learning environments the way we now reward journal articles, the work may well be outsourced to vendors, or become the province of entirely new institutions.
Taking our own advice
For decades, business schools have told companies to break down silos, organize around the customer, reward cross-functional collaboration, and disrupt themselves before someone else does. The AI revolution is a very good time for us to take our own advice. Some schools are starting the journey. Consider Drexel’s LeBow College of Business. The school emphasizes its co-op program, real problem-solving by students, and integration of technology into the curriculum. As one of the architects of their system, Dean Vibhas Madan, observes, a visiting Google executive said, “I can find 20 people who can code. But I’m going to hire the one who can persuade a team and think critically.”
Those schools that can reinvent themselves in light of demands for practical knowledge and technological savvy are likely to thrive. Those who can’t will find themselves on the wrong side of an inflection point.