By Jenny Karuna, Chief Product Officer, Katch
March 2026 – Recently, product leaders have begun to diverge sharply in their views on how AI will reshape the product manager role and what the job will look like in the years ahead. While the conversation hasn’t reached the same volume as the debate around developers, it’s still generating plenty of thoughtful, and often heated, discussion across X, LinkedIn, and Substack. The arguments tend to fall into two main camps.
One that says that product managers can now write code, and that some PMs want to or even need to write code (English is the hottest new programing language!) Product managers that have been dipping their toes into coding assistants like Claude Code or Cursor know that they have come very far in the last six months. In companies that value superfast development cycles or where teams are thinly staffed, PMs will get pulled closer to building. We’ll see more hybrid roles like forward-deployed engineers, product engineers, or PMs who can take an idea from napkin sketch to something working in hours.
The other school argues the opposite: AI tools enable product managers to focus on their core purpose: deciding what to build. In this version of the future, PMs stop drowning in Teams pings and meeting marathons. The classical PM–engineer partnership remains intact, just with less burnout, and faster, increased throughput.
The first scenario broadens the product manager’s responsibilities, giving them more autonomy and reducing dependency on others to bring an idea to life. The second scenario narrows the scope of responsibilities, offloading secondary responsibilities to an AI layer, allowing them to stay in the problem space, think deeply about what to build and the value that thing brings to the business or the user.
Two different futures, both plausible, and both likely to coexist. I’ll admit I don’t have a magic 8 ball, but if I had to pick one, I’d bet that the future will not be black and white. It will likely be a messy middle.
So maybe the right question to ask isn’t which scenario of the two is more likely to happen, but what skills might make a PM successful no matter which reality we land in.
Below are the foundational skills that matter whether AI is your co-builder or your co-pilot.
1.Problem framing
We now have the ability to build almost anything at a fraction of the cost and at speeds that would’ve been unthinkable even six months ago. High‑quality code generation is becoming easily accessible and extends far beyond vibe coding UI mockups. Claude Code can now spin up fully functional prototypes, mock APIs, data models, user flows, backend scaffolding, you name it, all in hours if not minutes. As execution gets cheaper, the critical skill becomes choosing what to execute.
The hardest and most valuable part of product work is no longer writing the code. It’s deciding what’s worth coding in the first place.
Strong PMs will identify real problems, articulate why they matter, and define what good looks like before code (AI generated or not) gets written. This brings us to taste.
2. Taste
Taste has always separated good PMs from great PMs. For those not familiar with the idea of taste in in product management, it is an editorial judgment, pattern recognition, an intuition for what “good” feels like visually and in terms of functionality. (While intuition sounds like it is something a PM is born with, it is a skill that can be developed and needs to be developed. Tip: Ask your AI assistant about this.)
After a PM identifies a worthwhile problem to solve, there’s now a choice: ask an AI assistant for solution ideas or lean on product intuition to determine exactly how to solve the problem? The answer is both, in the right order. Use AI assistants for breadth and exploration, then use taste as the ultimate filter. AI doesn’t carry lived context from your internal stakeholders, users, customers, or the market.
Execution is becoming cheap but product intuition, that ability to discern and decide will likely remain scarce.
3. Systems thinking
As AI features become more pervasive in a business workflow (as they will), PMs become systems designers, designing end-to-end workflows and experiences.
This means zooming out to see the big picture AND zooming in to the micro interactions between humans and AI. Future workflows look a lot like this: human → AI → human handoffs, machine‑initiated → human approval / overrides, in any order. The job is to make those transitions sensible and safe.
AI, unlike other tools, isn’t just another feature. It is also a user, an actor in the workflow. Someone must make that participation make sense. That someone is the PM.
4. Speed
If your discovery → prototype → build → scale cycle still takes months, that may be a path to irrelevance. AI coding agents compress build cycles dramatically. Work that once required coordinated handoffs across engineering, design, and QA can now be automated or accelerated to the point where iteration becomes continuous rather than episodic. The future-ready PM needs to be unfazed by this new speed. They need to embrace it to learn faster, test more, get user feedback sooner and make decisions with confidence. In our new reality, the PM who clings to old cadences is a bottleneck and the PM who embraces speed is a force multiplier.
So what does a successful PM look like in the future?
It is not the PM who makes all decisions and builds things solo. Neither is it the PM who delegates everything to AI. It is most certainly not the PM that clings to product operating models of the past that no longer make sense.
The PM who will thrive is the one who:
- Has clarity about problems that are worthy of solving.
- Exercises taste and judgment in how they should be solved.
- Designs systems that blend AI and human workflows sensibly.
- Operates at a pace that matches what AI makes possible.
Other traits will matter too, of course. The craft of product management is too complex to reduce to a short list of skills, although the ones outlined here are foundational for anyone that is involved in building products and will matter regardless of how the future unfolds.
About the Author
Jenny Karuna leads Katch’s product strategy, guiding the development of AI-powered solutions that maximize recoveries and unlock new growth opportunities. She brings nearly two decades of experience building SaaS and AI-first platforms across health care revenue cycle and payment integrity, with prior leadership roles at R1 RCM and Cloudmed. Known for her strategic mindset and analytical strength, Jenny excels at connecting dots, spotting patterns, and translating complex challenges into actionable, scalable solutions. She is passionate about building impactful products and fostering high-performing teams.
