I have built software for most of my working life, and I will say something that might sound strange coming from a software company: the traditional software developer, the person whose value is measured by the code they type, is about to become a much rarer job. Not in a generation. Within a few short years.

In its place, a new role is coming to the surface. Call it the software orchestrator, or the software designer. This person does not win because they know a framework better than anyone else. They win because they can imagine an outcome, describe it clearly, direct AI agents to build it, and judge whether the result actually works for the customer it was meant to serve.

The code is no longer the hard part

In early 2023, Andrej Karpathy, a founding member of OpenAI and former head of AI at Tesla, joked that "the hottest new programming language is English." In early 2025 he gave the idea a name, vibe coding: describing what you want, accepting what the AI produces and barely looking at the code at all. It went from an in-joke to Collins Dictionary’s word of the year.

Nvidia’s Jensen Huang went further at the World Government Summit in 2024, saying it is the industry’s job to create computing technology so that nobody has to program, and that the programming language is now human. You can debate the timing. You cannot really debate the direction.

I see it in our own work every day. The projects we deliver at etech increasingly start with a well-written brief, a clear design and a set of AI agents doing the heavy lifting, while our people spend their time on the decisions that matter: what to build, for whom, and how we will know it worked.

From developer to orchestrator

The data is already showing the shift. Stanford’s Digital Economy Lab found in its “Canaries in the Coal Mine” study that employment for software developers aged 22 to 25 fell by nearly 20 percent from its late-2022 peak, while more experienced workers in the same roles held steady. The researchers link the decline to tasks AI can automate, and the growth to roles where AI augments human judgement.

Read that carefully. The work that is disappearing is the codified, repeatable work: translating a specification into syntax. The work that is growing depends on judgement, context and taste. That is exactly the work of the orchestrator.

Tomorrow’s most valuable software people will spend their days on four things:

• Imagining the outcome a customer actually wants, not just the feature they asked for.
• Designing the experience and the rules that deliver it.
• Orchestrating AI agents to build, test and deploy, and knowing when to step in.
• Proving value: measuring whether the software changed anything for the customer.

It won’t be your technical skills that matter most in the future. It will be your imagination, and your ability to create software that produces the outcomes your customers want.

The white-collar labourer with a billion-dollar idea

Here is the part I find most exciting. If writing code is no longer the barrier, then the people who understand a problem best can build the solution themselves. The claims officer who knows exactly where an insurance process breaks. The bookkeeper who has watched small businesses struggle with the same spreadsheet for a decade. The nurse who knows which handover step loses information.

These people have always had the insight. What they lacked was a development team. That gap is closing fast. Sam Altman has said that he and other tech CEOs have a betting pool on the first year a one-person company reaches a billion-dollar valuation, something he called unimaginable without AI.

We are not there yet, and honest commentators such as Bloomberg’s Dave Lee point out that nobody has won that bet. But the trend is unmistakable. In 2025, Wix bought Base44, an AI app builder created by a single founder in a matter of months, for around US$80 million. Solo founders are now running businesses with the output of entire teams, as Fortune reported this year. The headcount needed to build something valuable is collapsing toward one.

So yes, I believe a white-collar worker, someone who has never written a line of code, can and will build a billion-dollar business. The tools are arriving. What they need is a clear idea of the outcome, the discipline to test it with real customers, and the imagination to see what nobody else has built.

AI doesn’t reduce our creativity. It releases it

The most common fear I hear is that AI will make us lazy and unoriginal. I think it does the opposite, provided we use it the right way.

A study published in Science Advances in 2024 by Anil Doshi and Oliver Hauser found that writers given AI-generated ideas produced stories rated as more creative, better written and more enjoyable, with the biggest gains for the people who started out least creative. AI lifted the floor.

The same study carries an important warning: the AI-assisted stories were more similar to each other. If everyone accepts the same first suggestion, everything starts to look the same. That is not an argument against AI. It is an argument for the orchestrator. The value moves to the person who refuses the average answer, combines ideas no model would put together, and knows their customer well enough to tell the difference.

For decades, most people with a great idea for software had to hand it to someone else to build, and watch it get diluted in translation. AI removes that translation layer. Your idea goes from your head to a working product with far fewer people in between. That is not less creativity. That is creativity with the handbrake off.

What this means for you

• If you are a developer: move up the stack. Learn product thinking, design, security and how to direct and review AI agents. Your judgement is the asset, not your typing speed.
• If you are a business leader: stop asking how many developers you need. Ask which outcomes you want, and who in your organisation understands them best.
• If you have an idea and no technical background: the barrier you thought you faced is falling. Start small, put something in front of real customers and learn quickly.

At etech, this is how we already work. We combine experienced engineering judgement with AI agents to take ideas from design to deployment faster, and with less risk, than a traditional team could. If you have an outcome in mind and want help orchestrating it, we would love to hear from you.