My Android app experience is limited to reading a Kotlin book and trying "hello world" once.
I've been using Gemini AI Pro for a year now, thanks to an introduction from a friend. I tried making an Android app during the recent holiday break. It was a simple app without a backend or database, and it worked well. I didn't have to get involved in the actual code at all.
I focused on clearly communicating the requirements in an md file (Gemini handles the md file organization, haha).
As much as possible, I started by organizing the requirements and then created an md file. I gave instructions like "Make step 1" - "Make step 2". Since I was using Android Studio with a virtual Android emulator, it could even debug errors on its own. When I asked for help debugging ambiguous behavior, it suggested checking the current state of the virtual environment app in Chrome browser on Windows.
I was surprised that even though I was developing in a Windows environment, Gemini AI Pro could monitor the virtual machine through Chrome.
I was worried that there might be limitations on AI usage with Gemini AI Pro and Android Studio, but it worked fine all day yesterday when I gave it commands.
It's great to live in a world where you can easily create and use utilities for personal or internal company use, whether it's Windows or Android. Of course, this means that the time frame for engineers to earn a living may be shortened.
