I just had a long conversation with Claude in which it poked a bunch of holes in our product strategy at @Buffer. This was by far the most (and possibly only?) specifically invaluable experience I’ve had with an AI tool this year.

First - fortunately our strategy is not screwed! A lot of good points came out of the discussion that we’re already thinking about and/or working on, and I have a few other things to consider 😅

Separately - I’m not here to say that AI is useless garbage.

But I’m really struggled to find the 10x use cases in it in PM outside of vibe coding (which has been written about a lot already) or spam/slop creation (which, if you follow me, you know I have…feelings about)

I use it as glorified search & to automate a few admin tasks. But to be honest, getting answers via AI isn’t that much faster than a great Google or Reddit search, and the types of tasks I’m wanting to automate involve a lot of handholding/review, so I’m not sure I’m actually saving much time.

Dozens of PM influencer types really honing in on “better faster PRDs” and AI prototyping as the things that will level up PMs. It will, for sure, in some cases. It will also level up all your teammates, especially designers - who can take an abstract problem we need to solve and get spec & working prototype ready, design system conforming and all, ready to hand off to an engineer to integrate (or even code themselves), in minutes.

But Good Product Managers™ uniquely add value by looking out beyond the immediate horizon, immersing itself in the current and potential users of its product to find opportunities, and bringing a team together around big ideas that matter & executing.

Some of that is “showing” the idea, in which case a quick prototyping route does help, yes. But you’re also now directly competing with your own designer(s) in their own workflow, and they have strengths to bring to the table that you may not. At best, this creates tension; at worst, you risk yourself being too top-down too aggressively.

As a not-terribly visual thinker, I try to embrace this and focus instead on those other things PMs can bring to the table: Focusing on honing in on what matters, so that my team can feel empowered to design & build the solution to what matters, and applying discretion, user centricity & taste to all of that.

This is what I’m trying AI for as a result:

• As a sparring partner when workshopping BIG ideas that are several steps beyond what the team is thinking about now. I’ve only recently gotten this working well after switching from ChatGPT (which was too much of a “yes man”) to Claude.
• As a brainstormer for questions & discussion prompts I can share with my user community to better understand their pain points and opportunities.
• As a quick processor of big data sets (this is not unique to PM).

What else are you using AI for as a product manager that isn’t writing PRDs, making prototypes, or automating admin?