From Post-Its to Python: The New Era of AI-Powered PMs

Hi, everyone. I'm Filip Szymanski.

Like many of you in the product management world, I'm fascinated by how artificial intelligence is changing our jobs. One big change I see happening right now is the growing importance of coding skills for product managers (or PMs).

Is coding becoming part of the PM toolkit? I strongly believe so.

About three months ago, I started coding again after stopping for almost 20 years! Thanks to new AI coding tools, I've learned incredibly fast. This experience convinced me that coding is a skill every PM should consider learning. Let me share why and what I've been doing.

Why Product Managers Should Care About Coding

As PMs, our main job is to find important problems that need solving. For most companies, the solution also needs to be successful commercially, i.e., it needs to make money or achieve a business goal.

But how do you know if your solution will work before you build the whole thing? No amount of research guarantees success. The best way is often to create a basic version of the solution, known as a Minimum Viable Product (MVP), and give it to real users or stakeholders. We then monitor progress to see if it actually solves the problem—without creating new ones.

Building even a simple MVP usually requires help from software engineers, which can take extra time and resources.

This is where AI coding tools come in. They make building and testing simple product ideas much faster and easier. So easy, in fact, that the product manager can often build the MVP themselves. This lets you test your ideas much faster.

How AI Makes Coding Easier: Two Types of Tools

So, how is this possible—especially if you don't code? There are mainly two kinds of AI coding assistants:

Agentic Development Tools (Web-Based)

These are often websites where you type instructions (prompts) to build an app. You don't need much coding knowledge to start.

• Purpose: Very simple apps, quickly making prototypes to show ideas, if you have zero coding background.

• Benefit: They can help you easily publish your app online.

• Platforms I Used:

  1. v0

  2. Bolt

  3. Lovable

• Downsides: They have limited capabilities. Additionally, sometimes they make mistakes that are hard to fix, and fixing them can use up your credits or allowance quickly. Making small changes step-by-step seems to work best. I haven't personally built a final, ready-to-use application with these. I don’t think they are quite there yet.

• My Advice: Even with these tools, knowing some coding basics helps you understand what's being built.

Intelligent Coding within an Integrated Development Environment (IDE)

These tools work inside professional coding software called an IDE. They help you write code, suggest improvements, and fix errors right where engineers work.

• Purpose: This is a great choice for any type of application, simple or complex. It works with almost any programming language.

• How It Works: It might seem scary if you've never coded, but many have a chat window, just like the web-based tools. You can still build things just by typing instructions.

• Platforms I Used: 

  1. Windsurf

  2. Cursor

  3. Cline

  4. GitHub Copilot

• Downsides: You usually need to figure out how to host the app (put it online) yourself. It requires learning some coding basics.

• My Advice: I recommend this approach for product managers.

My Journey: Learning to Code (Again) with AI

I believe PMs should have some coding skills. After my long break, I decided to learn modern web development. Here are some insights into my journey back to the coding world:

• Learning: I took online courses like Harvard's CS50W to learn HTML, CSS, Python, and JavaScript basics. I also took the University of Helsinki's Full Stack Open course to learn modern tools like React and Node.js.

• Languages: I focused on JavaScript and Python because AI tools seem to work best with them. This is likely because they are very popular languages, and they have been trained extensively on them.

• Tools: Using IDE tools like Windsurf (with AI like Claude Sonnet 3.7) helped me build more complex apps. For example, I built a server that connects a voice assistant to phone call features, running it in the cloud on Render. I also used Cursor and Cline to code, and early on, I experimented with GitHub Copilot.

• Hosting: I learned how to put my apps online using services like Heroku and Render. Their guides made it quite simple, and DevOps hosting headaches are a thing of the past.

• Other Uses: AI coding isn't just for big apps. I needed to clean up messy email lists, and writing a small program (a script) with AI help was much faster than doing it by hand.

• AI as a Teacher: Whenever I got stuck with code, terminal commands, error messages, or just understanding a concept, I asked an AI chatbot for help. It was like having a patient tutor available 24/7.

What might have taken me 2 years to learn before AI, I managed to get comfortable with in about 3 months. Yes, it takes time, but I believe it's a vital investment for AI-powered PMs and will be an essential skill for an AI-powered PM.

My Recommendation for Product Managers

I encourage every product manager to:

👉 Learn basic coding skills.

👉 Understand how modern web applications are built.

👉 Get comfortable using an AI coding assistant inside an IDE.

Why? Because AI coding tools allow you to build MVPs faster without software engineering help, and can even create helpful utilities for tasks like data analysis. Coding involves logic and clear steps, which AI is very good at helping with, and that is why I believe the democratization of coding is one of the first disruptions we are facing with AI.

What's Next?

As coding plays an increasingly important role in a PM’s job, it raises big questions:

• Will the job of a software engineer change fundamentally?

• How will engineers work with PMs who can now build early product versions?

• Could product management and software engineering roles even start to merge?

I'm curious to hear your thoughts. How do you see AI changing the product manager role? 

💬 Let me know in the comments below! 💬

About the Author

Filip Szymanski has over 20 years of experience as a product leader and is passionate about leveraging AI to power a new generation of product managers. Follow my journey at productpath.ai or add me on LinkedIn.

🚀 Boost your business with us—advertise where 10M+ AI leaders engage

🌟 Sign up for the first AI Hub in the world.

📲 Our Socials