For decades, building software required a special type of person.

You had to enjoy computers deeply. It involved spending years learning programming languages, debugging obscure errors, and understanding systems most people never even saw. The barrier to entry was high. If you had an idea for a tool or a workflow, you needed to find a software engineer. The same applied to a piece of automation. Typically, you also had to pay them to build it.

That barrier has just collapsed.

In the past few weeks alone, the capabilities of AI coding tools have crossed a significant threshold. This change feels fundamentally different from everything that came before. What once required years of training can now start with a simple prompt. Describe the tool you want. Refine it with feedback. It writes the code, debugs it, and even helps deploy it.

This shift will lead many people to conclude that the world will need fewer software engineers.

I believe the opposite is true.

AI coding tools will dramatically increase the number of people building software. The number of software-related tasks in the world will grow exponentially.

The Hidden Constraint of the Software Era

For most of the digital age, the limiting factor in software creation was not ideas.

It was builders.

There have always been more problems worth solving than engineers available to solve them. Every company has internal workflows that are automated. Every industry has inefficiencies that can be reduced with better tools. Every individual has repetitive tasks that can be turned into scripts, dashboards, or automations.

But building software required a rare combination of skills:

  • Understanding programming languages
  • Managing complex environments and dependencies
  • Debugging code
  • Deploying and maintaining systems

For most people, the effort needed to learn these skills far outweighed the value of solving a single problem.

The software that got built tended to be the largest ideas. They were the most economically possible. These products justified hiring teams of engineers.

Millions of smaller ideas simply never happened.

The Barrier Has Shattered

AI coding tools change the equation.

Today, a person can describe what they want in plain language and get:

  • working code
  • debugging help
  • architecture suggestions
  • automated tests
  • deployment instructions

Instead of learning syntax and frameworks first, users can start with the problem they want to solve.

This dramatically lowers the cost of experimentation.

Someone who earlier would have said “I wish this existed” can now say “Let’s try building it.”

The difference seems subtle, but the implications are enormous.

Lowering barriers does not shrink participation — it expands it.

Technology History Repeats Itself

We have seen this pattern many times before.

When spreadsheets appeared, millions of people suddenly became financial analysts. People who would never write statistical software began modeling businesses in Excel.

When blogging platforms arrived, millions became publishers.

When e-commerce platforms like Shopify emerged, millions became retailers.

In each case, technology did not remove professionals in the field. Instead, it dramatically expanded participation.

AI coding tools are doing the same thing for software development.

They are transforming software creation from a specialized discipline into a broadly accessible skill.

The Coming Explosion of Software Tasks

There is a major misconception about AI and coding. People assume the number of software jobs and the number of software tasks are the same thing.

They are not.

Even if the number of traditional software engineering roles eventually stabilizes or declines, the number of software-related tasks will explode.

Consider what becomes possible when building software is easy.

Individuals will start building:

  • personal automation tools
  • custom dashboards for their work
  • data analysis pipelines
  • niche applications for hobbies or communities
  • small internal tools inside companies
  • specialized SaaS products serving narrow markets

Most of this software will never become venture-backed startups.

But it will still be built.

In fact, the majority of software created in the next decade will never be publicly released products at all. Instead, it will be personal software — tools built to solve specific problems for individuals or small teams.

Spreadsheets enabled millions of personal financial models. These never became commercial products. Similarly, AI coding tools will allow millions of pieces of personal software.

More Builders, Not Fewer

Today there are tens of millions of software developers worldwide, and the number has been growing steadily for years.

But that number reflects the era when coding required specialized training.

As AI reduces the technical barrier, the definition of who counts as a “builder” expands dramatically.

Product managers will prototype features themselves.

Researchers will automate their data pipelines.

Small business owners will create tools tailored to their operations.

Students will build applications as easily as they create presentations.

The number of people capable of creating software will increase by an order of size.

Instead of tens of millions of developers, we will soon have hundreds of millions of software creators.

Why This Is a Golden Era for Software Engineers

Paradoxically, the rise of AI coding tools make it an even better time to be a software engineer.

AI is exceptionally good at generating code, but software development has never been just about writing code.

The real value lies in:

  • system design
  • architecture decisions
  • understanding trade-offs
  • ensuring reliability and scalability
  • translating ambiguous problems into clear solutions

These are precisely the areas where experienced engineers excel.

AI will handle more of the repetitive and mechanical aspects of programming. It will write boilerplate, scaffold projects, and generate tests. This allows engineers to focus on higher-level thinking.

The role shifts from code writer to system architect and AI orchestrator.

A single engineer working with AI tools can now build what once required an entire team.

Leverage increases dramatically.

The Age of Software Abundance

For most of computing history, software was scarce because the people capable of building it were scarce.

AI coding tools mark the beginning of a new phase: software abundance.

When the cost of building software approaches zero, the number of solutions increases dramatically. Problems that were formerly too small to justify development suddenly become worth solving.

The world does not run out of problems.

It runs out of builders.

By expanding the number of builders, AI will unleash a wave of experimentation. This wave will dwarf the startup boom of the past two decades.

The next generation of software will not just come from tech companies and venture-backed startups.

It will come from everywhere.

And from everyone.

Discover more from Startup Iceland

Subscribe to get the latest posts sent to your email.