The Rise of Automatic Coding

It's pretty clear to anyone paying attention: LLMs can code. This fact is changing how we think about software development.

In the auto-code project, I'm exploring just how far we can push LLM coding abilities. I've noticed two main ways AI is being used for coding:

1. AI-Assisted Coding: This is where professional developers are heading. Tools like GitHub Copilot, and fancier ones like Cursor and Devin are becoming part of a coder's daily life. Even if AI ends up writing most of the code, the human is still in charge.

2. Automated Code Generation: This is for everybody, being able to describe a tool, or an app, or a game and have it built for you is immensely useful. It won't replace professionally developed software but it's undoubtedly going to be a huge trend in the future (no-code/low-code was already a fast growing sector before it had AI-boosters)

Through this project, I'm trying to figure out what AI can and can't do when it comes to coding. It's pretty exciting to think about where this might lead us in the world of software creation.