Don’t Be a Product Manager If All You Want to Know Is the Dates
Why product managers need to own product quality, understand AI’s impact on users, and manage context instead of only chasing delivery dates.
Nowadays, with AI making code changes much faster, the role of a product manager becomes more important than ever. Previously, a good engineer could define good UI/UX and make some product decisions. A PM today needs to know and own more than that:
- What do users want AI to do in your product?
- What do users NOT want AI to do in your product?
- How do you define quality?
- How do you know you can trust your LLM judges, if you use them?
The other thing a PM needs to adapt to, and do way better, is managing their own context.
Think of an LLM’s context window. You give it information, decisions and constraints, and expect it to keep up. Sometimes it loses track. You remind it, and it may even apologise. But if you have to keep reminding it of the same things, you start questioning how useful it actually is. I’d like to think of people working in tech today in a similar way. Everyone has more work and context to manage, thanks to AI, so it’s okay to forget something and be reminded. However, if you’re still working with the same-old limited context window size a few years ago, while expecting everyone else to move faster, you’re outdated and should upgrade yourself.
So, as the title goes, don’t be a PM if you care more about the dates and way less about the product. Don’t be a PM if you’re too busy chasing when a ticket will be delivered to use the actual product you’re in charge of. And please find a way to upgrade your context window. If you can’t do that, ask nicely and I’ll remind you. But I can’t promise I’ll be nice every single time.