Start again

I've decided to start my development journey all over again, after learning some valuable things.

After an evening of excitement and many WhatsApp conversations with my friend Szymon, I got excited about my prospects to rapidly build apps. I downloaded Windsurf, spent a couple of hours tinkering with leaf php, played with code, and got into it in a great amount of detail.

But something was missing.

I'm not a programmer by trade, and I haven't been keeping up with my learning - and it showed within hours. I could read code, but stumbled over basic principles. The PHP framework I was trying to get to grips with seemed really clear, but the mental leaps I was trying to make were too much.

Eventually, I stopped, and reverted to a conversation with Claude AI.

"Be brave, be critical - here's my thoughts on learning to code. Can I just use AI to rapidly prototype apps? I like to learn from a working codebase, and take things apart. Surely that's a good way to code?".

Claude disagreed. I did, after all, tell it to be critical. It quite rightly pointed out that relying on AI to write code isn't always faultless, and might even generate security issues that I wouldn't be able to fix or even understand. Building apps using PHP, databases, TailwindUI, and everything else was like trying to drive before I even knew how to get in the car.

And that's my failing. I was getting carried away trying to dive in too high up the stack, thinking I could work it out - but the cognitive load was too high and stumbling around only gets you so far. (before you crash!)

I decided to be realistic with myself. Sure, I have projects I want to build for my day job - but right now, the basic methods I have used (spreadsheets, no-code apps) work. They don't need an app, and I don't need to rush. It's like trying to learn a language by diving into complex conversation with a native. I can't do it at my level, and because I'm not reliant on code - I'm a writer, after all - I have struggled to dedicate the mental effort I need to stay focussed on a difficult problem.

So, as suggested by Claude, I'm starting all over again from the basics.

I have courses, and I am starting them again. From scratch. Not skipping the easy stuff, but writing simple code over and over again until I know that I understand the basic principles before moving on. The basics need to be second nature so I can build on them.

I'm still using Windsurf so I can get AI guidance as I progress - AI is a good teacher and explainer, after all - and I'm using a course to set a framework for learning.

But I'm starting again. With code. alongside the day job. with renewed optimism that I'll get there. There is no rush. I'll have apps when I'm ready to build them, rather than relying on a series of crutches that all too quickly collapse under me.