I’ve been thinking a lot about how AI changes not only software development — but software consumption itself.
For decades, software worked roughly the same way:
- Find a product.
- Register.
- Configure your account.
- Learn the UI.
- Adapt your workflow to the product.
That model made sense when software had to be handcrafted.
But AI changes something fundamental.
The Shift From Products to Intent
Today, a growing number of workflows can already be described as:
“Here’s my data. Here’s my goal. Figure it out.”
Instead of searching for a predefined SaaS product, users increasingly interact with systems through intent.
For example:
- “Take these Excel files and generate a financial report.”
- “Analyze these APIs and build a dashboard.”
- “Pull customer data from CRM and generate outreach summaries.”
- “Turn this dataset into charts, audio summaries, and video explainers.”
The interesting part is that users no longer necessarily care:
- which database is used,
- which framework is used,
- which UI component library is used,
- or even which application is being executed underneath.
They care about outcomes.
Software Becomes Fluid
This leads to a strange possibility:
software may stop being a fixed product.
Instead, software becomes:
- dynamically assembled,
- task-oriented,
- temporary,
- generated around a specific intent.
Not:
“Here is a CRM platform.”
But:
“Here is a dynamically generated operational workflow for your sales process.”
That is a very different world.
The UI Stops Being the Product
If AI systems can:
- generate interfaces,
- connect APIs,
- transform data,
- write glue code,
- orchestrate workflows,
- iterate automatically,
then the fixed SaaS interface becomes less central.
The real product becomes:
- orchestration,
- memory,
- trust,
- integrations,
- permissions,
- execution reliability,
- accumulated operational context.
In other words:
the software layer becomes fluid, while the operational layer becomes persistent.
Why This Changes Competition
One consequence of AI-assisted generation is that implementation itself becomes increasingly commoditized.
Many teams now have access to:
- the same foundational models,
- the same coding agents,
- the same infrastructure,
- the same frameworks.
As a result, software products may begin converging much faster than before.
The competitive advantage shifts upward:
- operational workflow quality,
- execution speed,
- ecosystem integration,
- user understanding,
- organizational memory,
- continuous iteration.
A Different Future for Software
Maybe the future is not:
“AI writes software.”
Maybe the future is:
“software becomes a continuously evolving AI-assisted execution environment around human intent.”
And if that happens, we may stop thinking about software as static products entirely.
We may start thinking about software as:
- generated workflows,
- operational environments,
- dynamic systems assembled in real time.
That possibility feels much closer today than it did even a year ago.
And I think we are only at the very beginning of that transition.