To Code or Not to Code? That Really Is the New Question

Artificial intelligence hasn’t ended the need for programmers. But it has fundamentally changed who gets to build. For the first time in my career, I’m as excited about creating software as I was more than forty years ago.

It Started with Punch Cards

People often ask whether AI means everyone should learn to code.

For me, it’s an interesting question because I’ve spent most of my career orbiting technology, sometimes as a programmer, sometimes as an architect, sometimes as a programme director, but always fascinated by how ideas become reality.

My coding journey began at college.

I was 17, studying COBOL, when programming meant sitting in front of a keypunch machine producing decks of punched cards. One misplaced hole could mean starting again.

I loved it.

Logic came naturally to me and mathematics had always been one of my strongest subjects. Programming felt less like learning a language and more like solving puzzles.

To put that into perspective, the personal computer was still seven years away from becoming a commercial reality. This isn’t nostalgia. It’s simply the world I grew up in as a young aspiring programmer.

The Day I Switched Off the Computer

My first role was in systems programming, working on large IBM mainframes.

It wasn’t glamorous. Much of the work involved parameter-driven configuration of critical operating systems, with enough authority to accidentally cause enormous problems.

Which, of course, I managed to do.

Still learning the ropes, I found myself in a system control panel I’d entered by mistake. Wanting to get back to where I had been, I saw a button labelled Reset.

That sounded reasonable.

It wasn’t.

I’d just initiated a full system shutdown.

The first indication something had gone badly wrong wasn’t the screen.

It was my manager sprinting into the computer room wearing an expression I’ll never forget.

Thankfully, I survived the experience.

Just.

Falling Down the Assembler Rabbit Hole

Later, at NCR during the early days of cash machines, I encountered what remains the most demanding programming challenge of my career.

Assembler.

One of NCR’s banking clients wanted ATM “Out of Money” printer messages transmitted across the network into their IBM mainframe systems. It sounded straightforward enough.

It wasn’t.

Very few people understood both the networking protocols and the mainframe internals required to make it happen, so I found myself writing the solution.

Nine months later, I was still writing it.

Assembler wasn’t like COBOL or the higher-level languages that followed. There were no abstractions. You lived inside the processor itself, mentally tracking registers, moving bytes between memory locations and constantly remembering where everything was before writing the next instruction.

It was intellectually fascinating.

It was also utterly exhausting.

I realised something important during that project.

Once you disappeared into assembler, it became very difficult to think about anything else.

Then Came the Internet

The next revolution arrived wearing a much friendlier face.

The web.

I was working at Intelligence Finance, one of the UK’s early online financial services pioneers, helping manage the critical path for the launch of IF.com at a time when putting banking services online still felt revolutionary.

What fascinated me wasn’t simply the technology.

It was the accessibility.

Microsoft had just introduced FrontPage, allowing people to create websites without writing everything from scratch. Compared with assembler, it felt almost magical.

I immediately started experimenting.

That experimentation became UK Theatre Network, originally a spare-time hobby driven by my love of theatre and live performance.

Remarkably, it’s still going today.

From Code to Creativity

As web platforms evolved, so did the way I built.

Like many people, I experimented with different website builders as they emerged, each promising to make publishing easier than the last. Some disappeared almost as quickly as they arrived, often taking any easy migration path with them, something that became painfully familiar after years of theatre reviews.

Eventually I settled on Squarespace.

It struck the balance I was looking for. Beautiful design, reliable hosting and the freedom to concentrate on content rather than constantly maintaining code.

When I entered the luxury yachting sector, that simplicity mattered. The photography was spectacular, the layouts elegant and I could focus on building a brand rather than wrestling with technology.

Sometimes the best tool is simply the one that gets out of your way.

Enter Codex

And then AI arrived.

Or more accurately, AI coding assistants arrived.

I’ve been using OpenAI’s Codex extensively since its relaunch earlier this year, and it has reignited something I hadn’t felt for a long time.

Excitement.

Not because it’s replacing developers.

Because it’s removing friction.

The corporate world is increasingly moving towards Claude Code and Cursor, and I’m on the waiting list to explore those too. But for now, Codex has become my coding partner.

Although “partner” might be stretching the definition slightly. I tell it what to build, it writes the code and, most of the time, it does a remarkably good job.

This weekend alone, I built three entirely new publications.

Lomond Insider, showcasing luxury property, travel and lifestyle.

Entertainment Insider, supporting the film and theatre communities I’ve spent years building.

And Lomond Logic, exploring technology, innovation and the future of business.

Most ideas now take less than an hour to move from concept to working prototype, including testing.

Some don’t work.

That’s fine.

Failure has become wonderfully inexpensive.

The important thing is that ideas no longer have to wait.

So… To Code or Not to Code?

After forty years, my answer is surprisingly simple.

Learning to code taught me discipline, logic and problem-solving. Those lessons still matter enormously.

But AI has changed the role of coding itself.

Today, the real competitive advantage isn’t always knowing how to write every line of code.

It’s knowing what to build.

It’s recognising opportunities, asking better questions and being curious enough to experiment.

The barriers between imagination and execution have never been lower.

For people with ideas, that’s incredibly empowering.

For someone like me, who wakes up with new ideas almost every morning, it feels like being handed an entirely new set of tools.

The question is no longer whether you should code.

The question is: what is it that you’ve always wanted to build, but never had the opportunity to try?

For the first time, that idea might be only a conversation away.