I've always loved technology. I've just never been able to fully access it.
Not for lack of trying. I'd get excited about an idea, open a code editor, stare at it for twenty minutes, and close my laptop. The tools existed, but they demanded a fluency I didn't have and didn't have the patience to learn. So I'd hire someone, or I'd shelve the idea, or I'd build some duct-tape version in a no-code tool that did 40% of what I wanted.
That changed about six months ago. Not gradually. Overnight.

The Code Is My Voice Now
The thing that shifted is simple: the coding language is now my language. I describe what I want in plain English, and it gets built. Not perfectly. Not first time. But it gets built. And then I iterate on it the same way I'd iterate on a brief with a designer or a strategist. "Not quite. Try this instead. Closer. Ship it."
In the last six months I've shipped five products. A visual knowledge graph. A 42,000-line content engine. An energy audit dashboard. A booking bot. A proposal site that helped me negotiate a six-figure contract. I can't write a Python function. I don't really understand what a "hook" is. But these products exist, they work, and people use them.
I think this is genuinely what's next.

This Isn't Magic
I want to be honest about what this actually looks like day to day. It's not "say what you want and it appears." It's more like working with a very fast, very patient junior developer who never sleeps and costs £150 a month. You describe something, it builds 70% of it, and you spend the next hour saying "not like that" and "the spacing is weird" until it's right.
Things break. You debug by describing symptoms instead of reading stack traces. You make architectural decisions about systems you don't fully understand. You review code by reading it like a novel and asking "does this feel right?"
It works. Not in a "good enough" kind of way. These are real products that I use every day. The difference between me and a traditional developer isn't capability. It's vocabulary. I think in user experiences and business logic. The AI translates.

The Window
People who could build software 40, 50 years ago were going to change the world. And then they did. Microsoft, Apple, Google. The ability to make a computer do what you wanted was the most valuable skill on the planet for decades.
I think we're at a similar moment. There's a window right now where the people who learn to work with AI have an absurd advantage over everyone else. Not because the AI does everything for them, but because they're operating at a completely different speed. While someone else is hiring a developer, briefing them, and waiting for Sprint 1, you've already built it, tested it, pivoted, and moved on.
Eventually the AI will just do everything. No human in the loop. And this weird skill of "being good at telling AI what to do" will be worthless. But right now? It's the most valuable thing I've ever learned.
What It Takes
Clarity of thought. Tolerance for iteration. Taste. You need to know what "good" looks like even if you can't build it yourself. And shamelessness. You will ask your AI to explain things a first-year CS student already knows. Your users won't care, because the product works.
This blog post was written with my AI assistant. I described what I wanted, he drafted it, I edited it, and we shipped it together. That's the whole point. The thinking is mine. The execution is collaborative. Tools are tools.
The window is open. Walk through it or don't.
Rees Calder runs Levity, a lead generation agency, and serves as CPO of Thingiverse. He builds products with AI despite having the coding skills of an enthusiastic golden retriever.