Product management + AI (and the illusion of efficiency).
This week: We hit on some of the drawbacks of relying too heavily on AI in your product world.
AI tools are creeping into every corner of product management, making it easier to draft user stories, prioritize backlogs, and analyze data. Off-the-shelf AI tools are a lifesaver for product people like me who have spent most of their days in traditional, old-school, monolithic companies, especially those navigating compliance-heavy environments. They simplify the manual work and help us look polished to stakeholders.
But here’s the question I keep coming back to: Are we really working smarter, or are we just getting better at creating the perception of being awesome product people? And what happens when we lean too hard on AI without fully understanding the nuances of our products?
The (mostly obvious) upside of AI tools.
We have all seen a ton of posts about the AI wins we get as product people. I recently hit on a few of my favorite ones in a post titled “How are intelligent tools reshaping the role of product?"
Tools like JIRA and MIRO’s AI-powered features can make life much easier. They help product managers churn out user stories, draft epics, and prioritize tasks quickly. These tools shine in the early stages of planning - when you need to get concepts out of your head and onto paper, especially for stakeholders who don’t need all the nitty-gritty details yet.
In my experience, AI has been especially helpful for creating high-level frameworks. It allows me to articulate ideas faster and with more polish, which is invaluable when presenting new initiatives. But as great as these tools are, they have limitations.
The trouble with relying on AI for the details.
The challenge comes when it’s time to transition from those high-level drafts to actionable tasks. That’s where things can fall apart if you’re not careful.
Developers need more than a polished user story - they need clear, thoughtful context. If the AI-generated stories are vague or overly generic, they become painfully obvious once the team digs in. I’ve learned that while AI can help you look like you’ve got it all together, it won’t fool your dev team for long if you don’t truly understand what you’re building.
There’s also the risk of overloading your backlog. It’s easy to generate dozens of stories in minutes, but you're just creating more noise unless they’re prioritized and tied to meaningful outcomes. It’s not about how much you produce - it’s about whether what you make matters.
In my experience, this is where trust comes into play. You risk eroding credibility if your team feels like the work isn’t well-thought-out. Other product managers who work closely with me rely heavily on these tools. Whether they realize it or not - it’s super obvious based on all the missing, more personalized vocabulary, such as specific system names. Until AI can seamlessly combine legacy application history, expected behavior, and your intent, you must bridge that gap.
The perception of productivity vs. actual impact.
One of the most interesting dynamics I’ve noticed is how AI can create the illusion of high productivity. Rapidly generating user stories or epics might make it seem like your team is moving faster, but are you driving better outcomes?
It’s worth asking: does leveraging AI tools actually boost velocity, or are we just getting better at looking busy? This is the line I’ve been trying to walk - using AI to accelerate the early stages of planning while ensuring that the final outputs are refined and meaningful.
Take, for example, a tight deadline for a stakeholder presentation. AI can help you create a polished deck with clear user stories and epics. But once you take those same stories to the dev team, they must be tailored and prioritized. Without that extra layer of refinement, you risk spinning your wheels on low-value work.
Striking the balance.
Here’s where I’ve landed: AI is an amazing tool for jump-starting the work, but it’s not a replacement for thoughtful product management. It’s best used to complement your expertise, not replace it.
When I use AI, I try to think of it as my first draft, not the finished product. It gets me 70% of the way there, but the remaining stuff - the nuance, the context, the prioritization - still requires my input. This approach has helped me maintain trust with my team while taking advantage of AI’s efficiency.
It’s also a reminder to stay focused on the fundamentals. Developers need clarity, stakeholders need a strategy, and the product needs to deliver value. If AI can help with those things, great - but it’s not the whole story.
Final Thoughts
For me, the most exciting part of using AI is how it frees up time for more impactful work. But this only works if we’re deliberate about using that time.
Instead of focusing on churning out more stories or tasks, I see AI as a chance to prioritize bigger-picture questions:
Are we solving the correct problems?
How do these features tie into our long-term strategy?
Are there creative solutions we haven’t explored yet?
AI tools shift our role from doing the work to guiding the work. That’s a huge opportunity - but only if we embrace it. It means spending more time thinking critically, asking better questions, and ensuring our teams feel connected to the broader product vision.
AI is powerful but can’t replicate a product person’s ability to inspire, align, and make judgment calls. That’s where we come in. By leveraging AI’s efficiency while doubling down on these human elements, we can redefine what it means to be a great product manager - not just now but in the years ahead.
How are you balancing AI’s strengths with the need for precision and trust in your teams? Hit this fancy comment button below and share your thoughts with us.