How the Best Product Managers Use AI (And What Everyone Else Gets Wrong)

Fireside Product Management

Content provided by Tom Leung. All podcast content including episodes, graphics, and podcast descriptions are uploaded and provided directly by Tom Leung or their podcast platform partner. If you believe someone is using your copyrighted work without your permission, you can follow the process outlined here https://podcastplayer.com/legal.

6M ago 57:40

MP3•Episode home

AI is transforming product management, but not everyone is using it effectively. Too many product managers treat AI as a glorified autocomplete—something that speeds up basic tasks but doesn’t fundamentally change their workflow. But after my conversation with Mustafa Kapatiya, a former Google executive and AI consultant, I walked away convinced that the best PMs aren’t just using AI—they’re redefining their entire approach around it.

Mustafa has been deep in the trenches, helping product teams harness AI to maximize efficiency and impact. He shared insights from working with organizations that are getting it right—and from those that are stuck at the surface level, frustrated that AI "just doesn’t get it." The difference? Knowing how to ask the right questions, structure inputs properly, and train AI to think like your best-performing team members.

Why Most PMs Are Stuck in the Shallow End

Most product managers start with AI the same way they started using Google Docs or Jira—it’s a tool, not a game-changer. They pop open ChatGPT or Claude, ask for a quick summary, maybe a user story template, and move on. Mustafa has seen this pattern again and again:

“Most PMs use AI in a very surface-level way. They play around with ChatGPT, get some decent results, but then get frustrated when AI ‘doesn’t get it.’ The reality is, they’re not using even 20% of what AI can actually do for them.”

The key distinction between elite PMs and the rest? Elite PMs don’t just ask AI for one-off answers. They integrate it deeply into their workflow. They train it to understand their company’s OKRs, their team’s strengths and weaknesses, and the nuances of how decisions get made.

“Top PMs think about AI in three key dimensions: speed, quality, and effort. They don’t just use it to go faster—they use it to produce better work and to minimize the time they personally spend on low-leverage tasks.”

AI as Your Second Brain

Imagine you’re a PM trying to make sense of customer feedback from hundreds of app reviews. The average PM might copy-paste a few into ChatGPT and ask for sentiment analysis. The great PM, on the other hand, does something entirely different. They:

* Train AI on their past decisions, OKRs, and company priorities.

* Give AI structured data and ask for a synthesized, weighted summary.

* Use AI to determine which insights actually move the needle, rather than just producing a generic report.

“If you ask AI the wrong question, you get a generic response. If you train it on your context, give it structured inputs, and refine its responses, it becomes a true second brain. That’s where the magic happens.”

Mustafa demonstrated this in real time during our chat. He uploaded raw Figma app reviews into Claude, structured a thoughtful prompt, and within minutes, AI produced a highly actionable summary: key pain points, frequently requested features, and a breakdown of sentiment trends. But he didn’t stop there. Instead of just handing that report off to stakeholders, he used AI to determine which insights mapped to team OKRs and who on his team needed to see them first.

This is where AI becomes more than a speed tool—it becomes a decision-making engine.

The New Playbook for AI-Powered PMs

So how do you move from surface-level AI use to best-in-class execution? Mustafa outlined three core shifts that separate the best from the rest:

1. Write Better Prompts (And Reuse Them)

Most PMs write one-off prompts each time they need AI to do something. The best PMs treat prompts like reusable assets, refining and improving them over time. Mustafa has even built a PM Playbook—a collection of prompts that cover everything from user research to roadmap prioritization.

“You should never be writing the same AI prompt from scratch every time. Write it once, refine it, and reuse it. A great prompt is like a great framework—it saves you hours every week.”

2. Train AI on Your Context

If AI doesn’t understand your company, your team, and your unique challenges, it’s going to spit out generic advice. The best PMs create “digital twins”—structured datasets that teach AI about their org structure, key stakeholders, and product priorities.

“I train my AI coach on five key dimensions: company context, my role, team structure, product specifics, and my strengths/weaknesses. This allows AI to give me insights that feel deeply relevant, rather than just surface-level observations.”

3. Use AI to Navigate Organizational Complexity

PMs don’t just build products—they navigate company politics, stakeholder expectations, and resource constraints. AI can be a powerful tool for this. Mustafa showed how he uses AI to analyze who in his org is most likely to support or resist a given initiative, helping him craft better pitches and build alignment faster.

“One of my favorite use cases? AI as a political coach. It helps me figure out how to frame conversations, who to bring in early, and how to avoid roadblocks before they happen.”

The Future: Leaner, Smarter Product Teams

AI isn’t just changing how PMs work—it’s changing who gets hired and how teams are structured. I shared a hypothesis with Mustafa: what if, in the near future, a small team of elite PMs, deeply trained in AI, could outperform a traditionally structured product org with layers of GPMs, APMs, and specialists?

Mustafa agreed—and took it a step further:

“I actually think we’re about to see a huge shift in how product teams are structured. Instead of large teams with lots of layers, we’ll see small groups of highly effective PMs, each managing multiple products with AI as their co-pilot.”

If this happens, the role of the PM will evolve. Instead of spending 80% of their time on documentation, synthesis, and reporting, PMs will focus on decision-making, strategy, and creativity. AI will handle the grunt work—PMs will steer the ship.

What This Means for You

If you’re a PM today, this shift is already happening. The question is: are you ahead of the curve, or are you still treating AI like a toy?

Here’s what you can do today:

* Start treating AI as a second brain, not a search engine. Train it, refine it, and make it work for you.

* Build your own AI playbook. Save and refine the prompts that work. Share them with your team.

* Use AI to navigate your organization, not just write documents. It can help you build alignment, anticipate roadblocks, and move faster.

Mustafa and I barely scratched the surface in our conversation, but one thing was clear: the PMs who master AI will have an unfair advantage. The ones who ignore it? They’ll be left behind.

If you’re serious about leveling up, I highly recommend checking out Mustafa’s AI playbook and following his work on Substack and LinkedIn. And if you want to get direct coaching on AI-driven product management, feel free to reach out.

The future belongs to PMs who know how to work with AI. Are you ready?Mustafa’s Free Assessment (limited seats): https://bit.ly/4kesaVj

Mustafa’s Linkedin: https://www.linkedin.com/in/kapadiamustafa/Tom’s Coaching Page

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit firesidepm.substack.com

106 episodes

#Tech #MBA #Business #Tom Leung #Productdevelopment #Productmanagement #Productivity