Navigating Strategy In The Age Of AI With Roger Martin - TWMJ #1021

**Mitch Joel** (0:00)
Well, hey there, and welcome to episode number 1021 of Thinking With Mitch Joel. It's Sunday, February the 1st, 2026 This is the show that used to be called Six Pixels of Separation. Let's get on with the conversation.
Strategy became a buzzword before PowerPoint frameworks and KPI dashboards. Strategy was about making hard choices when the future refused to cooperate. Today's conversation is with Roger Martin, one of the clearest thinkers on what strategy actually is, and if I'm being honest, one of my favorite business thinkers in the world, without a doubt. Roger is the former dean of the Rotman School of Management, a longtime advisor to CEOs around the world, and co-author alongside AG. Lafley of a new and revised version of the amazing book Playing To Win. It's a book that quietly rewired how a generation of leaders think about winning, choice, and trade-offs. Today we talk about why strategy isn't about data, why analysis can actually make decisions worse, why AI is far more dangerous as an answer engine than a thinking partner, and how we've confused scientific certainty with human judgment, often with unsettling consequences. This is a conversation about thinking better, about resisting precision, and about why the most important strategic decisions can't be proven, only argued. If you've ever felt that business has become over-optimized, over-modeled, and strangely allergic to any form of imagination, this one's for you. Enjoy the conversation. So who are you and what do you do?

**Roger Martin** (1:55)
My name is Roger Martin and I'm a retired professor, strategy advisor to executives, and a writer. In fact, I do a lot more of that last piece than I used to do.

**Mitch Joel** (2:10)
Yeah, and you've got a newly updated and expanded version of a book that I always love, Playing To Win, which is still fairly new. It's a newish book, right? Yeah.

**Roger Martin** (2:20)
I mean, it's interesting. It was 2013, so that's not too terribly old, and HBRP, the publisher, just reissued it with two of the HBR articles that AG and I collaborated on, which I thought was a good idea. I don't think it's old enough to do an updated version. They were interested in me doing an updated version, and I said, I think I've got other stuff on my mind, but then they came up with this other idea to add content that we had already done in another medium, and I thought that was a great idea, so kudos to them.

**Mitch Joel** (2:58)
Yeah, it's a great idea too, because then I get to corner you for an hour and I have an excuse, which is like always...

**Roger Martin** (3:05)
You don't have to have an excuse to corner me. I always enjoy doing your podcasts.

**Mitch Joel** (3:11)
I appreciate that, but the work you've been putting out, you're talking about this before hitting record, has just been both substantive and so impactful. There are people... When I talk about... I literally built a ChatGPT business advisor after your work. I don't know if you know that, but it's pretty amazing. Like it's really good. It's like a good little voice in my ear when I'm not sure which direction I should take or not.

**Roger Martin** (3:40)
Somebody sent me a Claude. They asked Claude to summarize my work and sent it to me, and it was pretty good. So AI is pretty amazing. On that topic, I am working on something on that front. Because I know you're totally all over AI.
In the domain that I think we operate, or at least I'll just speak for myself, I operate, which is getting to strategy solutions. That's what people are trying to use AI for. Here's my situation, tell me what I should do. I'm not a fan of that. I'm not interested in having an AI, even though there's a company I've worked with, like just a client who is interested in it, and they created a Roger Bot to use internally. It's a big player in the AI space. What I've started working on, I don't know if I'll finish it, is more of a dialogue partner, where rather than the AI coming up with the answer for you, what the AI does is help you as your thought partner. It's sort of Socratic. It says, okay, Mitch, what's the problem that you're trying to solve? Well, okay, that's not such a great definition. Can we approve that? And to go through my way of thinking about strategy as a dialogue partner who is never going to give the person the answer, the person is going to discover an answer by being questioned. I'm more interested in that form of strategy AI than I am what normally people kind of want to do with it.