I'm Feeling the Shift in Product
I'm officially feeling the shift in product. And I don't mean that in the generic "AI is changing everything" way that everyone is saying right now.
I mean I can feel it in the actual day-to-day work.
The way software gets built is changing very quickly. Some teams are starting to move at a speed that feels almost impossible compared to the way product and engineering teams have worked for the last 10 or 15 years.
A small group of people can go from idea, to prototype, to working software in days. Sometimes hours.
Claude Code, Cursor, Codex, coding agents — whatever tool you prefer — they are making it easier than ever to take an idea and turn it into something real.
And when you really get it working, the feeling is incredible. It's probably the closest thing to surfing a wave for a nerd like me.
You have this idea in your head, you prompt it, refine it, correct it, push it, and suddenly you are interacting with something that used to require a full team, a full sprint, and a lot of back and forth.
That part is amazing.
Speed Creates New Questions
But I think we are still very early in understanding what this actually means for companies. Because when some teams are moving 100x faster than others, it creates a whole new set of questions.
Which speed is the right speed?
Where is the bottleneck now?
Who owns fixing the entire system across a 500-person company, when there are 394 steps between an idea and a measurable business result?
And this is where I think the conversation gets more interesting.
The real challenge for companies is not going to be whether they can build more software. The harder question is going to be whether they are building the right things.
And that sounds obvious, but I don't think most companies are set up that way today.
The Real Problem: Too Many Ideas, Not Too Few
Most companies I talk to are not short on ideas. They are drowning in them.
Sales has a list of things they need to close deals.
Customers are asking for new features.
Support knows where the product is breaking down.
Executives are pushing strategic priorities.
Engineering sees technical debt, platform issues, and infrastructure work that has to get cleaned up before the system falls apart.
Marketing wants better stories.
Customer success wants better onboarding.
Finance wants efficiency.
Legal and compliance want control.
And now AI makes it possible for even more people to build even more things, faster than ever.
That sounds exciting. And it is exciting. But it also creates a very real problem.
If building gets easier, the organization has to get much better at deciding.
That is the part I think a lot of people are missing right now.
The Bottleneck Is Moving Upstream
The bottleneck is moving upstream. It is moving into product management. It is moving into strategy. It is moving into prioritization. It is moving into the messy middle between what the business wants, what customers need, what engineering can build, and what will actually create value.
I actually don't think AI means there will be less need for product people. I think there will be more need for great product leadership than ever.
Because the hard part is becoming everything that happens before the code.
The Questions That Matter Before You Build
Do we understand the customer problem?
Do we know how this ties to the business strategy?
Do we agree on what matters most right now?
Do we have the right requirements?
Do we know what success looks like?
Can we explain why this work deserves capacity instead of something else?
Can we connect the thing we are building to the metric we believe it should move?
Can we tell leadership, with a straight face, why this is the best use of expensive product and engineering time?
That is the real work. And it is getting more important, not less.
Faster Development Without Better Judgment
Faster development without better judgment does not create better companies. It creates more output.
More noise.
More things to maintain.
More releases for marketing to explain.
More features for sales to position.
More complexity for support to handle.
And worst of all, more things for the customer to understand.
That is the part that gets lost. Every feature you build has a cost after it ships.
It has to be supported.
It has to be sold.
It has to be explained.
It has to be maintained.
It has to fit into the customer experience.
It has to fit into the business model.
It has to fit into the company strategy.
If AI makes it easier to build, that does not mean every idea should become software. It means the quality of the decision before building matters even more.
The Context Problem
This is where I think a lot of companies are going to struggle. They have strategy in one place. Roadmaps in another. Tickets in another. Customer feedback somewhere else. Metrics in dashboards. Executive priorities in decks.
And then everyone is trying to make decisions from different pieces of context. The company keeps moving, but the "why" gets weaker at every handoff.
Business strategy becomes product interpretation.
Product interpretation becomes requirements.
Requirements become engineering tickets.
Engineering tickets become releases.
Releases become status updates.
Status updates become executive summaries.
And by the time the story gets back to leadership, everyone is trying to remember what the original point was.
Now add AI acceleration to that. Suddenly teams can build faster, but the context problem gets even more dangerous.
If the context is weak, AI helps you build the wrong things faster.
The Better Operating System
That is the shift I'm feeling. The best companies are not going to be the ones that just adopt AI coding tools the fastest. They are going to be the ones that create a better operating system around the work.
A better way to connect strategy to opportunities.
Opportunities to requirements.
Requirements to execution.
Execution to resources.
Resources to outcomes.
Outcomes back to leadership decisions.
That is the loop. And that loop has to get much tighter.
Because product and engineering capacity is too expensive to manage through vibes, Slack threads, disconnected roadmaps, and hopeful status updates.
The Questions You Should Be Asking
The question should not just be, "Can we build this faster?" The question should be:
Should we build this at all?
Is this the right bet?
What outcome do we expect?
How confident are we?
What customer evidence do we have?
What business metric does this move?
What else are we not doing if we do this?
How will we know if it is working?
That is where the leverage is going.
The Bigger Unlock: AI With Full Company Context
AI will help teams move from messy input to execution-ready artifacts faster. That is a huge unlock.
But the bigger unlock is when AI has the full company context — when it understands the strategy, the customer, the market, the roadmap, the tickets, the metrics, the constraints, and the outcomes the business actually cares about.
That is when it stops being a tool that helps people produce more stuff. It becomes a way to make better decisions.
And that is what companies need. Because most companies can already produce plenty of stuff. They have roadmaps full of stuff. Backlogs full of stuff. Quarterly plans full of stuff.
The real question is whether the stuff is focused on what matters.
Are teams actually working on the right things?
Is execution matching the plan?
Where are resources going?
Are we building what will create value?
What is this work producing for the business?
Those are the questions product leaders are going to have to answer more clearly in this next era. Not six months later. Not after the release. Not once the board asks and everyone scrambles to build the deck.
Now. As the work is happening.
The Product Shift
AI is going to make building easier. But deciding what is worth building may become the most important skill of all.
The teams that win will not simply be the teams that ship the most code. They will be the teams that combine business context, customer understanding, product judgment, execution visibility, and AI acceleration into a better way of working.
That is the future I'm seeing. Not just faster software.
Better decisions.
Better focus.
Better use of resources.
Better connection between what teams are doing and what the business actually needs.
And honestly, I think that is the most exciting part of this whole AI shift.
Because if we get it right, we do not just build more. We build what matters.
