Michael Olenick: At Michigan Ross, Training Students To Leverage Increased Cognition

Latest in a series featuring short interviews with faculty at various B-schools focused on the use of AI in business and business education. See also Rebuilding The Role Play (Kellogg), Learning Requires Friction (Wharton), The Deck Is Dead (Stanford), and We Teach MBAs To Apologize About The Best Tool They Have (INSEAD).

I’ve been working on the application of AI to strategy for many years, first alone, then with INSEAD professors Chan Kim and Renée Mauborgne. The concept didn’t generate much enthusiasm for quite a long time.

As research progressed, we became a small group inside INSEAD focused on the idea, especially Peter Zemsky and Hyunjin Kim. One name that stood out and came up repeatedly was Felipe Csaszar of the Ross School of Business at the University of Michigan. Felipe has been a prolific scholar in the field. His work asks what strategy even is once machines can out-read, out-model, and out-iterate the people making it.

AI is likely to touch every field of business education, but the nature of strategy is different than many. Whereas finance and accounting usually have a strong right or wrong answer, strategy is much more in flux. There are often quite a few clearly wrong answers and nearly as many right ones. What works for a firm depends on elements not easily measured: what is the firm good at? What does the staff care about? Are the customers willing to try new and different things? Is the staff even willing to try different things?

The marketplace is no more sure than the business schools. Hiring at McKinsey and Bain is down but BCG is staffing up. Felipe is more enthusiastic than some of his peers. His argument: the internet cut communication costs, AI cuts cognition costs, and strategy is the discipline most exposed to that change.

---

Poets&Quants: We’ve been running a series on what different business schools are doing around AI, teaching materials, and how things are changing. Nobody has talked specifically about strategy yet. How is AI changing how strategy is taught?

Csaszar: There are many ways to approach AI and strategy in the classroom. One is to make almost no changes, just update the cases. Instead of Coke and Pepsi, you teach OpenAI versus Anthropic. You need some of that. Students need to know the main players and how the technology works. But I think the effect of AI on strategy will be much larger than swapping out cases.

One important piece is giving students a practical understanding of what AI can and can’t do. In my classes, I recommend that everyone install a coding agent. Not just for coding. You can feed it spreadsheets, documents, ask it to generate figures and dashboards. We’re past the ChatGPT type of interaction. It’s agents now and knowing how to work with them is important.

An activity I’ve used when teaching AI and strategy: I give students 45 minutes to create a new company in a given space, say food and beverage. They have to produce a business plan, product packaging, an advertising campaign, and if they have time, a website, a jingle, a video. The things people produce in 45 minutes are remarkable. And students are always surprised by what they were able to do. That creates energy, and it creates a new way of seeing the world: there’s a lot of room for experimentation.

P&Q: You mentioned strategy specifically. How does AI change the actual substance of strategy, not just how it’s taught?

Csaszar: This is what I’ve been working on, and there’s an HBR article (of mine) that covers it. The core idea is that strategy exists because we are boundedly rational. The problems that strategists need to solve are too complex for any person to handle fully. So strategy developed tools, simple frameworks, powerful oversimplifications, as Bruce Henderson of BCG called them, that capture the essence of a problem well enough to act on it.

AI changes those limits. An AI has read more than any person. You can set it to think about a problem longer, from more angles, with more detail. So AI doesn’t eliminate the need for strategy frameworks, but it does change the constraints that made them necessary in the first place.

The framework I’ve developed looks at the three core cognitive operations behind strategy. One is search: you look for options until you find something good enough to adopt. Another is representation: you take a point of view that lets you see things others can’t. The third is aggregation: you combine opinions from multiple experts, customers, stakeholders, with the hope that combining them cancels out errors and gives you a fuller picture, the wisdom of the crowd. AI can enter all three of those processes very naturally.

What I want to do next time I teach the strategy core is push students to use AI in each of those three modes. To search for more opportunities than they’d find on their own. To represent problems from perspectives they wouldn’t have considered. And to aggregate a broader range of views.

P&Q: We’ve been seeing something like this in research: personas that turn into agents, analyze a framework like a value chain, then discuss it with each other.

Csaszar: That’s a great example of both representation and aggregation. You give the problem a framework, a value chain, that’s a way of representing it. And then have multiple agents look at it from different angles. That’s aggregation. You get the breadth of perspectives without any single one dominating.

P&Q: Which raises the question of where humans fit in. You mentioned at some point that the AI won’t go bankrupt or go to prison. So at some point you need a person.

Csaszar: For the foreseeable future, yes, and for several reasons. One way to think about it: having AI assist with strategy isn’t very different from hiring a consulting firm. If you bring in McKinsey, you still need a CEO and a top management team. McKinsey generates alternatives. You have to guide them, provide the context, evaluate what comes back, and then decide. AI improves the quality of the alternatives and the support behind them.

But you still have to formulate, select, implement, and take the risk. Someone has to be accountable. That doesn’t go away.

I actually think there may be more demand for strategy people, not less. A lot of what junior people were doing was producing analysis. Now AI does that. Which means more people will be needed to evaluate, to judge, to make the calls. That’s closer to the actual job of a strategist.

P&Q: We keep hearing about job Armageddon, but the numbers don’t show it. Developer hiring, for example, has actually been going up.

Csaszar: There’s still real uncertainty. But if I had to bet based on history, we’re at the beginning of a big technological revolution. Thirty years from now, companies will look completely different, probably more different than companies today look from companies in 1995. The internet reduced communication costs. AI reduces cognition costs. And cognition is everywhere in an organization, which means the effect will be broader and deeper.

People will find new uses for cognition, new products, new services. Someone will need to lead that change. And my bet is that the people best positioned to lead it are the ones who understand both the business and the technology. I call it the hybrid professional. That profile will be the most in demand for at least the next decade.

P&Q: Stefano made an interesting point along similar lines: that international business used to be its own category, and then the internet made global communication so frictionless that international business just became business. He wondered whether AI will do the same to AI courses in ten or twenty years.

Csaszar: Yes, I think that’s right. I’m teaching a course called AI and Strategy right now. But I don’t imagine it will exist as a standalone course forever. It exists to accelerate the transition. At some point, this will just be part of strategy, the way the internet is now just part of everything. You don’t have courses on the internet. Cases reference it, every discipline uses it. AI will be the same. It’s a tool.

P&Q: Any closing thoughts on where the classroom goes from here?

Csaszar: The hands-on piece is the thing I’d emphasize most. Students need to get their hands dirty, not just read about AI or discuss it theoretically. The 45-minute company exercise I described, using coding agents in class, creating prototypes and dashboards and competing to show the results. That’s where something changes. You can see the potential and the limits at the same time, and that’s worth more than any amount of abstract framing.

---

Michael Olenick, JD, is a former INSEAD research fellow and founder of VSTRAT.ai. He has spent four decades building AI and expert systems; his research has been used at INSEAD and, before then, cited by Congress and media on how to resolve the 2008 era financial crisis. Michael currently lives with his partner, Anastassia, in Austin, Texas.

© Copyright 2026 Poets & Quants. All rights reserved. This article may not be republished, rewritten or otherwise distributed without written permission. To reprint or license this article or any content from Poets & Quants, please submit your request HERE.