Our industry is rapidly changing

Years ago, when I was first learning Ruby on Rails, I decided that I would create a blog application to give myself something concrete to work on. The code was riddled with mistakes and poor practices—the natural consequences of a self-taught developer honing his craft. As I slowly cobbled together a working application, the next barrier was hosting and deployment. Luckily for me, Heroku was still around, and with a little reading, both obstacles were soon overcome.

I then put my new application to use, with some success. For a brief period of time my blog tutorial was on the front page of search results for Rails 3 tutorials—quite the accomplishment at the time and something I'm still proud of. The skills I learned helped me progress my career, and I started coding eight hours a day, five days a week. This of course quickly removed any desire to code in my personal time, and the blogging stopped shortly thereafter.

Ten years later, the repo had remained untouched and incredibly out of date. It was still on Rails 4 and had a multitude of half-finished projects and other tech debt. Now, with AI, I thought it might be a fun project to resurrect the application and redeploy it. Originally, I spent months getting the application working and deployed. Having used AI for about a year now, I expected updating it to be trivial, but I wasn't prepared for exactly HOW trivial.

It took about an hour, while I multitasked, to update it, deploy it on a completely new platform and write this article. And not just updated but refined—gone are the dead code, over-engineered authentication, Angular, Bootstrap, mistakes and insecure commit history. Hello to a slimmer codebase, a simpler app, and expanded test coverage.

Our industry is rapidly changing.