For the past two years we've been asking the wrong question about AI and software development.
The question everyone keeps asking is:
Will AI replace developers?
But after spending time building software with AI coding tools, I’ve started to suspect something very different might happen.
AI might not eliminate developers.
It might create demand for more of them — just in places that never hired developers before.
In particular, it may create a new role that many companies have never had:
the internal developer.
The Company With No IT Department
Imagine a typical mid-sized business.
Maybe 30–60 employees.
They are good at what they do, but their internal systems look familiar to anyone who has worked outside the tech industry.
Spreadsheets everywhere.
Quotes built in Word.
Customer notes stored in email threads.
Information scattered across accounting software, shared drives, and manual processes.
None of it is completely broken. But none of it is particularly efficient either.
Historically these businesses solved software problems in three ways:
• buy SaaS products
• hire consultants
• live with inefficient processes
Custom software was rarely an option.
Not because the problems weren’t worth solving.
But because building software felt expensive, risky, and slow.
How Software Has Traditionally Been Built
For most companies, the model looked something like this:
This approach works, but it creates friction.
Projects take time. Requirements must be carefully written. The people building the system are often external to the business itself.
As a result, many operational problems simply never get solved with software.
They remain spreadsheets.
AI Changes the Starting Point
AI-assisted development tools change one critical thing:
They dramatically lower the cost of creating the first version of a system.
Now imagine someone inside that business starts experimenting with AI coding tools.
Within a few days they produce a rough internal tool:
• a quoting calculator
• a small workflow form
• a dashboard showing job status
It isn’t perfect.
But it works.
And suddenly the company realises something important.
Custom software might no longer be unreachable.
A Different Model Starts to Appear
Instead of software always being built outside the business, the workflow might start to look like this:
The company no longer needs to start with a consultancy engagement.
They can start with experimentation.
That’s a very different starting point.
The First Internal Tool
Imagine the first real success.
Someone builds a small quoting system.
The impact is immediate:
• quotes are generated faster
• pricing becomes consistent
• information is stored properly
• reporting becomes possible
The tool might only save a few hours each week.
But the business now sees something it has rarely experienced before:
A problem solved exactly the way their workflow needs it solved.
That’s when the mindset changes.
Instead of asking:
“Which software should we buy?”
The company begins asking:
“Could we build something for this?”
The Hidden Reality
Once that first internal system exists, a new reality appears.
Software isn’t just something you build.
It’s something you own.
The system needs improvements.
Someone reports a bug.
Another team wants a similar tool.
AI may make building software easier, but it does not remove the need for someone who understands how the system works.
That responsibility has to live somewhere.
The Internal Developer
At this point a new role quietly emerges.
Not a full engineering team.
Not necessarily even a traditional software engineer.
But someone inside the business who becomes responsible for the company’s internal software capability.
Someone who can:
• understand business workflows
• design simple systems
• use AI tools to build and evolve them
• maintain those systems over time
This person becomes the bridge between how the business operates and the software that supports it.
Historically many companies never needed that role.
AI might change that.
A New Kind of Demand
For decades, most companies outside the tech industry simply didn’t hire developers.
They relied on vendors, consultants, and packaged software.
But if AI makes building internal tools dramatically cheaper, the economics change.
Instead of outsourcing every project, many companies may decide it makes sense to have one internal builder who understands their systems and can create solutions when needed.
Across thousands of companies, that could represent a significant shift.
Not fewer developers.
Just developers working in different places.
The Real Opportunity
The interesting implication is not that AI replaces developers.
It’s that AI might make software creation accessible to businesses that previously couldn’t justify it.
That means:
• more internal tools
• more automation
• more operational software solving real problems
In other words:
more software built in more places.
Places that previously had none.
A Question I'm Curious About
If AI continues lowering the barrier to building software, do you think we'll start seeing more companies hire one internal developer instead of relying entirely on agencies?
Or will consultancies simply adopt AI themselves and keep the existing model?
I'm curious what other developers are seeing in the real world.
Top comments (0)