On Saturday I listened to Martin Lewis on Moneybox discussing fuel prices in the context of the Gulf conflict. He mentioned the government’s Fuel Finder service, which I’d never heard of. It’s a publicly available API that lets you query a database of registered fuel station prices. I know what an API is, roughly. I have no idea how to actually use one.

I’d been aware for a while that AI tools can write code, but I’d never had a reason to try it. This gave me one. I had an idea for something I wanted, so I asked Claude to build it. I used Claude in Chrome to point it at the API documentation (well beyond my technical comfort zone) and asked it to figure out how the service worked. After some trial and error, it did. I described what I wanted. A simple dashboard showing current fuel prices at my two usual petrol stations, one near work and one near home. Claude built it as a web app, deployed it to my hosting account, and made it installable on my iPhone home screen. It even designed an app icon, which was a nice touch. The whole thing took under 30 minutes from hearing the radio segment.

I didn’t write a single line of code. I couldn’t have. The only skill involved on my side was being able to describe what I wanted clearly enough.

As a practical tool it’s a small thing. I check petrol prices, I save a few pence (Costco is cheaper than Sainsbury’s, who knew!). But over the weekend it’s made me think about something that feels bigger, which is how many small, useful tools like this never get built. The person with the idea doesn’t have the technical skills, and the person with the technical skills doesn’t have the problem. That’s always been the gap, and to me it just got a lot narrower.

I work in a university, and I find myself wondering what this means for how we think about teaching people to code. If the barrier to building software is shifting from writing code to describing clearly what you need, where does that leave disciplines like computer science? I’m not suggesting they become irrelevant. But something has changed, something big. I’m now going to be thinking of all the other small apps I can build to make life and work a little easier.

I’d be interested to hear from anyone teaching in this space, or building in it, because I suspect the answer is more complicated than either “coding is dead” or “nothing has changed”.