AI-assisted development: the machine speeds up, I steer

It depends on what for. To validate an idea, get a first version out or launch fast on a tight budget: yes — well-directed AI produces serious results in a fraction of the time. For a brand that lives off its image, an AI site built without direction shows: cloned layouts, generic copy, that template feel your customers can't name but definitely notice.

My short answer: it's worth it when someone is steering. AI is the crew; art direction, strategy and review are still human. And if your project doesn't need AI, I'll tell you that too.

The usual three. In SEO, AI builders tend to produce weak semantic structure, duplicate metadata and slow pages — the kind of thing Google quietly penalises. In security, unreviewed generated code can expose API keys, ship forms without validation or lean on abandoned dependencies. And in maintenance, you inherit code nobody fully understands — not even the AI that wrote it.

None of this is the AI's fault: it's the consequence of publishing without review. In my projects, every generated line goes through my hands before touching production, and technical SEO ships as standard, like on any site of mine.

Yes — and it's probably the best use case today. In one to two weeks I can build you a clickable prototype — not static mockups: a working version you can click through, put in front of real customers or show to an investor.

The honest caveat: a prototype is not a product. It exists to learn fast and decide with evidence whether to build for real. If we carry on afterwards, part of the work carries over and part gets rebuilt properly — and I'll tell you which is which from day one.

Somewhat — and above all it changes where the hours go. AI saves me the repetitive code, which translates into tighter quotes or more scope for the same price. What it doesn't make cheaper is the judgement: deciding what to build, designing the interface, reviewing what's generated and getting it production-ready are still human hours.

Be wary of anyone promising a professional website for €200 'thanks to AI': either it isn't professional, or the costs show up later. My ranges are still my ranges — a website between €1,500 and €6,000 — with a fixed written quote before we start.

Yes, and I'm hearing from more and more people in that exact situation. First comes an honest audit: I review what the builder produced — structure, SEO, performance, security — and tell you what's worth keeping, what needs fixing and what's cheaper to rebuild.

Sometimes the answer is uncomfortable: there are builder sites where fixing costs more than starting over with everything you've learned. I'd rather tell you that before invoicing than have us both discover it mid-project.