The Great Pivot: How AI is Transforming the Role of Product Managers
Interview with Brad Eiben, Executive Director, Masters of Product Management program, Carnegie Mellon
Brad Eiben is the Executive Director of the MS Product Management program at Carnegie Mellon University's Tepper School of Business. With a background in product leadership roles at Kromek Group and Cleaveland/Price, he is a distinguished figure in product management.
At Carnegie Mellon, Brad has grown the MSPM program from essentially a startup to a mature product through the same PM philosophies, strategies, and tactics he teaches. His approach to continuous improvement and his dedication to student success have earned him high praise from both colleagues and students.
Recently, Brad and I spoke about the potential impact of AI on product management.
Here is my favorite quote from the interview:
What follows below is a condensed and lightly edited version of our conversation.
Brad, it's great to have you here. I have been very much looking forward to this. Let’s start at the top, how do you think AI changes the way we build products?
Brad Eiben: Mustafa thanks for having me. I have been looking forward to this as well. And I have made lots of notes (laughing)
Great. I love it when my guests come prepared. Let’s get to it?
BE: Mustafa your question on the surface looks seemingly simple. But in reality, it is extremely complex.
Allow me to travel back in time. When I was an engineering student, I always struggled with writing during my exams. I knew the answers to the questions and understood the concepts, but I was a slow writer and my handwriting was hard to read.
Then the TI-92 graphing calculator arrived, and it was a game changer for me.
It allowed me to focus on the true test of knowledge rather than wasting time on rote tasks like graphing equations and flipping through books.
I see AI playing a similar role in eliminating non-value-added busy work for today's product teams. For developers, AI can handle the tedious coding work so they spend their effort on ensuring the code properly solves the core problem.
For product managers, AI can automate tiresome tasks like report generation so we spend our time on the human-centered, strategic aspects of defining the right product vision.
Great story. Can I double down on the product manager role? How do you see AI changing the way they work?
BE: If we narrow in on the PM’s role, let’s consider how much of a PM’s total work time is value added. Much of the day involves generating reports, creating templates, online research, compiling data, and giving presentations. None of this, in isolation, adds value. Synthesizing and enveloping into an actionable strategy does. This is where we’re going. As a bonus, it is also much more rewarding work for the PM and the team.
At the same time, I can’t see AI replacing the human-centered aspects of a PM’s role changing. Not anytime soon, at least. There is no substitute for spending time with users in context. It will be dangerous for product teams to sacrifice this step, though I’m sure many will try.
My favorite problem-solving quote is from Fujio Cho, former President of Toyota, “Go see. Ask why. Show respect.” Without treating your customers in this way, it will always be impossible to empathize and solve the right problem.
PMs will need to be more human, in general. The best will grow stronger at persuading, motivating, and defining the way.
What about at the org. level? What needs to change there?
BE: For me this is a critical piece of the puzzle that organizations must get it right.
A few years ago, I once worked for a company where all of the executives attended an innovation boot camp. For the next 18 months, “innovation” was the buzzword thrown around at every upper-level meeting. Everyone was told to add 'innovative goals' to our annual reviews that year.
The intent was right, but there were two problems. First, no clarity was ever communicated as to why, what, and how. Innovation was truly a vague buzzword. How should Nina in Corporate Accounts Payable be expected to innovate in a routine clerical process? Second, nobody’s operational goals changed at all. Everyone was still expected to perform tasks with no allotment for experimentation
I fear many organizations are already treating 'AI' the same superficial way - as an ambiguous hype term with no concrete vision behind it. That's a recipe for breeding fear and resistance, not sustainable change.
Real change has to happen top-down and bottom-up. From the top, leaders must understand that 'the way we've always done it' will rapidly become obsolete. They need to collaborate across functions to map out tangible AI use cases and a phased roadmap for integrating AI as empowering partners, not just labor substitutes.
But the bottom-up piece is equally vital. The best ideas will come from teams feeling empowered to re-examine every process through a fresh 'How might we use AI?' lens.
In my master's program, for example, we've moved away from just discussing application tweaks. Instead, we ask 'How can we evaluate candidates in new ways that identify leadership traits and PM potential?' That's driving us to explore AI for multimedia interviews, unique essay prompts, and more.
So leaders need to provide that clear overarching vision while also giving air cover for teams to freely experiment - and sometimes fail. Those who navigate that balance will rapidly pull ahead as AI advances.
Agreed. Top-down and bottoms-up have worked well in the past. And I think you need the same here as well. Let me go back to our PMs. In this new world, what new skills should PMs acquire? What is the best way to get started?
BE: With the advent of AI….I think all PMs need to look at their work through two lenses.
First, what is truly value added? How can I do more of it? And second, can I use AI to automate the non-value-added activities? How can I deconstruct my jobs to be done. Is there a revolutionary way in which I can accomplish my goal with revolutionary new tools?
Most PMs already have a high workload and are prone to focus 110% on getting through their to-do list. A percentage of time needs to be dedicated to finding a better way. That is where I think AI can help.
As to where to start? I have a confession. My starting point was to ask Chat GPT and Gemini to tell me a comparison of their pros and cons. It was that simple. Later I asked Chat GPT, Gemini, and Claude to tell me about the best use cases for each. Similarly, you can ask, “What are the AI use cases for product marketing and what are the best applications?” From there, experiment with the free and trial versions.
And if you are looking for a shortcut, and I know this is an unsolicited plug, but I know a guy at a company called Echo Point who can help product teams master AI.
Thanks, Brad for that shameless plug. Much appreciated (laughing). I may have to edit that last line out when I transcribe our interview.
One last question, what is your favorite AI tool? Why?
BE: When I have a task or question for LLM, which I frequently do, I’ll normally open Gemini and Claude. I’ll ask the same question. This allows me to cross-check response quality and consistency. I like Claude because it seems to be the most factually accurate. I like the format and creativity Gemini affords.
I like Gemini-generated images for use in presentations because it saves hours of scrolling through Google Images to find just the right image to reflect a message.
My most recent was an image of the comic character Juggernaut smashing through a chessboard.
Brad, this has been a lot of fun. Thank you for being such a great guest.
BE: Mustafa, likewise. Thanks for having me.