How can PMs navigate the AI conundrum?
AI AI every where,
And all the teams did shrink;
AI AI every where
But not a single KR moved
In The Rime of the Ancient Mariner by Samuel Taylor Coleridge, he writes the story of a voyager who found himself drifting south because of a storm, reaching the Antarctic.
I first read the phrase “Water, water, everywhere, but not a drop to drink” in my CBSE book many years ago, but the phrase has struck with me since then.
The current AI landscape seems similar.
Everyone around you is asking what should you do with AI. Your company leadership asks you to use AI tools in everything you do. Then there’s your over-excited colleague who brings you a vibe-coded “fully functional” app and asks to ship to production. Your UX colleague says they have Figma prototype that is ready to deploy, and your engg colleague says they have created the designs for your product.
Everyone is writing PRDs. Everyone is making designs. Everyone is writing code.
Your YouTube history is flooded with solopreneurs who are making millions building apps online. Lovable, Figma, v0, Vercel, Replit, n8n, and god-knows how many others want you to use their platform to build. Lenny shows you hour-long videos of people building with AI, and watching this content eats up your time, while stressing you up at the same time.
What the heck?
What does a modern-day PM do?Through this confused world, what is your role as a PM?
My take is that the Product Manager role hasn’t fundamentally changed: You still need to make decisions to help your users get the most value out of your product.
Irrespective of whoever does what on the team, the responsibility of success rests with you.
You own the decisions, not the PRD, or the mockup, or the code.
The craft of making good decisions for your users comes from:
Understanding your users really well (through data, user research, competitors)Understanding the “builders” (your UX, Engg and any other counterparts)Understanding the constraints involved (tech, budget, infra, etc.)In this modern world, don’t focus on owning the PRD, or any specific artifact. Focus on better evaluation of everything your team does. The one question you need to be asking and answering (as any great PM would anyways be doing) – is this going to help my user?
If the answer is Yes – encourage your teams to build it, but challenge and refine the initial concept / POC to reduce noise and focus.
If the answer is No – refine the idea if a variation of it could help. Or cut it off.
If the answer is Maybe – find a way to learn more about the user.
It is worth noting that not only product builders, but also its users are confused about how to use AI. Don’t be in a FOMO to ship that AI agent if you are not confident how it’s going to serve your users. Find out which of their problems is it going to solve.
At the end, all I will say is that as PMs, our craft is not to ship fast, but to ship long-term value for our users and our companies. Fast is not necessarily durable.
Stay focussed, cut the noise.
I published this post originally as a part of Magnifier. I plan to write more about my experience in Product Management. Make sure to subscribe to get the latest posts.
The post How can PMs navigate the AI conundrum? appeared first on Hemant R Joshi.
