• The Hidden Layer
  • Posts
  • Why I'm Turning to App Development (But Not Completely Giving Up AI)

Why I'm Turning to App Development (But Not Completely Giving Up AI)

Week 1 of documenting my app development learning journey (Feb 23 - Mar 1)

This is going to be new.

So far, I’ve been documenting my learning in Artificial intelligence, but now, I want to apply it.

However, my main use for it isn’t making software for a PC…

..it’s going to be either:

  1. An educational app that gamifies learning for AP courses, or

  2. A computer vision app that can be used in FRC (FIRST Robotics Competition) competitions

So from now on, I’ll be focusing (more) on app development, and when AI comes in again (e.g. generating questions or personalized feedback), I’ll bring it in then.

This change in topics has made me very emotional for the past week—it made me feel that I was doing this was for nothing but for wasting time…here’s how I tried to figure it all out about this newsletter.

Sunday, February 23rd

Today I mainly just learned all about Xcode, Apple’s specialized IDE, where Meta basically just glazed (praised) Xcode for at least 15 minutes; here are something about Xcode:

  • Make (free!) project templates

  • A complete build system

  • A fast emulator

  • Can operate on different devices (exclusively in Apple’s environment)

  • Have third-party integrations (including TensorFlow Lite hopefully?)

  • Has a versatile toolbar that can simulate an app on a virtual-looking phone

iOS apps have four main components:

  1. UIApplication, which is the central point of control and coordination; every app has one of these, where it’s like the “heart” of the application.

  2. UIApplicationDelegate, which is a set of methods that maintain the shared behavior of an app; it’s also the root of the app

  3. UIDocument, which allows an app to save, create, and manage data and content to a file locally through file systems to manage storage; this is what enables some apps to be used offline.

  4. UIViewController, which defines shared behavior and properties, and oversees the activities that are going on on-screen like UI, animations, interactions, navigations, etc.

The ViewController is a container, in which the app exists and is “drawn”, in a way.

When making my first “project” in Xcode, there was an “organization identifier” field, in which I didn’t know what to put (I don’t own an organization, nor do I have a personal website), so I followed this guy’s advice:

Thanks, guy!

Also, after some Googling, I just learned something about integrating AI into iOS apps: they use a specialized ML framework called Core ML. Supposedly it’s faster due to “native integration” or something like that.

So as I see it as of right now, I see three options for how to integrate AI into my app:

  1. Use TensorFlow Lite or PyTorch and convert it into a form iOS can use

  2. Use CoreML

Saturday, March 1st

I was away on a school trip, about a couple of hours away from where I live, where I had to sleep over there for two nights; now, since I’m back, I’ll get back on track:

Today I went over the organization of Xcode’s IDE interface.

The red area is for navigation, the green area is the editor area, the yellow area is the utility area for seeing details of the code or file that’s being worked on, and the gray-purplish area is the debug bar and is, well…for debugging, and shows things like error messages when they happen.

Pretty simple, considering that other popular IDEs use a very similar structure, like VSCode:

What is an API, anyways?

I used to confuse the terms “API” and “documentation”, but while APIs include documentation for their software to an extent, they are more than just documentation. API stands for “Application Programming Interface”, where “it’s essentially a set of rules or protocols that allows different software applications to communicate with each other, enabling them to exchange data, features, and functionality” (perplexity.ai, 2025).

So when you’re using a computer or a phone to use an app or a program, say, for example, doing a Google search, there is the front-end, which is the screen you see, with the Google logo and search bar and all that, and the back-end, which is Google’s databases that store all of the data that it finds and indexes from “crawling” the Internet. The API in this case would be the interface or “communicator” that enables the actual communication between the screen that you see (the Google home screen) and Google's search infrastructure and DBs (databases).

Note that this Google Search process I just explained is oversimplified; it captures the general idea, but is realistically more complex than this.

Accomplishments

  • Xcode is the main editor for developing iOS apps

  • I learned the basics of Xcode’s layout

  • I discovered the functionality of an API, and what it does for programs and applications.

Resources:

Course used: