I will probably end up writing it up for BTTR more comprehensively, but the past few days I decided to try and use Gemini to add features and update my portfolio site: nick.broughall.com.au.

I built that site earlier this year for a job application. I liked the idea of creating something I didn’t have to pay for hosting for, lol.

But originally it was just customising a template. I can’t code, have no understanding of node.js or React or anything that actually went into the site’s build.

As such, it was pretty stark. It lacked any customisation at all. So I tried to use AI to give me that versatility.

And it worked. The site now has things like an RSS feed on BTTR, and a subscribe box for the BTTR newsletter built in.

But the process? It was like pulling teeth. Every single instruction AI gave me created errors. I had to go over and over the process to systematically fix those errors.

Did it allow me to do things I wouldn’t have been able to do on my own previously? Yes, absolutely.

Did it do it well? Not at all.

If I had budget to hire someone to do the work, it probably would have taken them an hour. I spent hours working through the process of updating and fixing errors. It was exhausting.

A screenshot of my portfolio site, nick.broughall.com.au