Many people predicts on social media – that it is end of Product Management.
First it was no code.
Then it was empowered engineering teams.
Now it is AI.
The latest discussion suggests that engineers can act as mini Product Managers for projects that take two weeks or less. In many situations, I agree.
If the customer problem is already understood, success metrics are clear, and the roadmap has already been prioritized, then AI and engineers can absolutely own much more of the execution. They can clarify requirements, make feature level decisions, and deliver faster than ever.
That is progress.
But I think this conversation misses one important question.
Who is responsible for maximizing the return on every engineering hour?
That is where Product Management becomes even more valuable.
Building software is no longer the expensive part
AI has dramatically reduced the cost of writing code.
A developer who previously spent three days implementing a feature might now complete it in a single day.
That sounds like incredible productivity.
But something interesting happens when building becomes cheaper.
The bottleneck shifts.
The question is no longer:
“Can we build this?”
The question becomes:
“Should we build this at all?”
That is not an engineering problem.
It is an investment problem.
An operational issue that looks obvious
Imagine an ecommerce company.
Every morning, warehouse operators complain that they cannot easily identify delayed customer orders.
An experienced engineer proposes a solution.
“We can build a dashboard showing all delayed orders with live filtering, color coded statuses, export to Excel, and configurable alerts.”
With AI assistance, the implementation only takes one week.
Everyone is impressed.
The dashboard works beautifully.
Six months later, usage analytics tell a different story.
Only five warehouse supervisors open it regularly.
Operators still rely on their existing workflow.
Customer delivery performance has barely changed.
The company spent valuable engineering capacity solving a problem that produced very little business impact.
The solution that was never considered
A Product Manager approaches the same problem differently.
Instead of asking:
“What should we build?”
They ask:
“Why are orders delayed?”
After speaking with warehouse teams, they discover something unexpected.
Eighty percent of delays happen because priority orders are mixed with standard orders during picking.
The warehouse management system already supports priority labels.
The labels simply are not being displayed on the picking device.
Instead of building an entirely new dashboard, the team spends two days exposing an existing priority flag.
The result?
Warehouse workers naturally pick urgent orders first.
Delivery performance improves immediately.
No new dashboard.
No additional operational process.
Minimal engineering effort.
Higher business impact.
Engineers optimize solutions. Product Managers optimize investments.
This is not about technical capability.
Many senior engineers could have discovered the same insight.
The difference is perspective.
Engineers are trained to solve problems efficiently.
Product Managers are responsible for deciding whether a problem is worth solving, whether another problem deserves attention first, and whether the expected outcome justifies the investment.
Every feature competes against another feature.
Every sprint has an opportunity cost.
Every engineering hour is an investment.
The PM’s job is not simply to maximize output.
It is to maximize return on investment.
AI makes this responsibility even more important
As AI accelerates development, companies will naturally build more.
That sounds exciting.
But building more is not always creating more value.
If engineering capacity doubles, companies do not suddenly have twice as many valuable ideas.
Without strong prioritization, they simply build low value features faster.
AI has reduced the cost of building.
It has not reduced the cost of building the wrong thing.
The future Product Manager
I do not believe the future PM is someone who writes detailed requirement documents or manages every small feature.
Those activities should increasingly belong to engineers supported by AI.
The future PM looks different.
They understand customers deeply.
They identify the highest value problems.
They connect product decisions to business outcomes.
They understand opportunity cost.
Most importantly, they treat engineering capacity as one of the company’s most valuable investments.
Final thought
The debate should not be whether engineers become mini Product Managers.
I hope they do.
The more ownership engineers have, the faster great ideas become reality.
But autonomy does not replace prioritization.
Someone still has to answer the hardest question in product development.
Not “Can we build this?”
But “Is this the best possible use of our limited engineering investment?”
As AI makes software development faster and cheaper, the ability to build is becoming less of a constraint. The bigger challenge may be deciding where that engineering capacity should be invested.
When teams can build more, the cost of building the wrong thing can become even greater. This makes prioritisation, opportunity cost and ROI increasingly important.
Ultimately, Product Management is not about deciding every detail of what engineers build. It is about ensuring that limited engineering capacity is directed towards problems that create meaningful customer and business value.
In an AI-driven world, knowing what to build, what not to build, and why may become one of the most valuable responsibilities of a Product Manager.