Why Steve McCurry Doesn't Miss Film And What That Means for Coding in the AI Era

Last night in Melbourne, at Leica's 100 year celebration, I found myself standing in front of Steve McCurry, the man who captured one of the most iconic photographs of the last century, the "Afghan Girl." For me, someone who lugs around too many old film cameras and still wrestles with light meters, it was like stepping into a dream.

It was also my first time visiting Leica's new Melbourne store. The space itself had that mix of warmth and polish only Leica seems able to conjure. The staff were welcoming, I bumped into a few friends, and waiters floated around with trays of fancy drinks. Along the walls, McCurry's photographs were printed and hung with quiet reverence, the colours deep and rich, the frames perfectly lit. And there she was: the Afghan Girl, her piercing green eyes cutting straight through the room.

After some introductions, McCurry spoke briefly about his long career, the places he's travelled, the stories he's pursued, the people he's met. Then came the Q&A. Someone raised the question I was secretly hoping would be asked: Did he miss film?

He smiled and paused. "These days, a little," he admitted. "But not when it first happened." For him, the shift in the early 2000s had felt natural. Digital freed him. No more waiting for rolls to be developed. No more risk of running out mid assignment. No more wondering if the shot had landed. Efficiency replaced uncertainty, and he never looked back.

And that’s what caught me in the chest. As someone who still loads film like it's a ritual, hearing McCurry say he didn't miss it, not even for a day, was quietly heartbreaking. It felt like watching a door close on something sacred, and at the same time realising the world had already moved on.

The GenAI Moment

That conversation lingered with me. Because the shift from film to digital in photography feels a lot like what’s happening in software engineering today.

If you write code in 2025, you’re already in the middle of it. Tools like Claude Code, Gemini CLI, Cursor, Cline, and Roo are flooding into our field. Suddenly, every coder has an assistant, a co-pilot that can generate, refactor, and sometimes even deploy code.

There’s hype, of course. There’s always hype when the ground shifts. But beneath the noise, something irreversible is happening. We are crossing a line. Just as photographers once went from loading film cartridges to shooting endless frames on memory cards, we're moving from the analogue world of handcrafting every line to a digital world where code is faster, smarter, and less hands on.

You can almost feel the industry splitting in two. On one side, the voices warning: This is the end of real engineering. On the other, the optimists saying: No, this is just the next beginning.

Tools Change. Vision Stays.

McCurry reminded me of something essential: tools change, but the essence doesn’t.

Back in the 1950s, you could hand a Leica to a bystander, and the photos would be unremarkable. You could give a cheap box camera to Cartier-Bresson, and he'd still find the decisive moment. The lens was never the point. The roll of film wasn't the point. The point was the eye, the timing, the ability to find meaning in chaos.

The same is true now. Give an AI coding assistant to someone without clarity or judgment, and the results will be messy, confused code. Give a bare terminal to someone with insight and taste, and they’ll still build something remarkable.

In the end, whether you're holding a camera or staring at a code editor, what matters is not just how you click or type, but how you see.

Embracing Change Without Worshipping It

I’ll be honest: I’m embracing the AI tools. They’re intoxicating at times. I’ve shipped features in days that once would have taken weeks. I’ve used them to experiment wildly, to break things, to be surprised by what comes back.

But I’m also wary. Because there’s a seduction in believing the tool is the craft. It never is. Digital didn’t make McCurry a master. Film didn’t make Cartier-Bresson a poet of the street. And AI won’t make us better engineers.

It’s our vision. Our judgment. Our taste. That’s what shapes the work. And that’s what can’t be automated.

What Stays True

So maybe this is the real lesson. Technology always promises efficiency, and often delivers it. But efficiency is not the same as meaning.

Photography was never about the film. It was about seeing. Software engineering was never about typing every line by hand. It was about solving problems, building things that matter, shaping the invisible architecture of how people live and work.

AI will change how we work. That much is clear. But what will define us is not how fast we adopt the new tool, but how we hold on to the vision at the centre of the craft.

Whether you’re making a photograph or building an app, the edge is still the same: how you see the world, and how you bring that vision to life.