The Death of Human-Written Code Doctrines in the Age of ChatGPT … Or Not? | by Murtaza Ali | January, 2025

In a world where ChatGPT rules, one might expect tutorials on human-written code to fall flat.

After all, who's going to read a human-written tutorial on a detailed software process, when they can just ask ChatGPT for the output they want? Throw in advanced LLM enrollment and a few advanced engineering techniques, and anyone can produce detailed explanations on any topic. Why waste the time and effort to research, discuss, and write such an article?

Let's start with an important question, the answer to which will support my entire argument: What exactly is ChatGPT?

However, ChatGPT, or any large language model (LLM) for that matter, works with a machine learning algorithm that predicts which words should follow the input based on training data. What distinguishes LLM from previous iterations of language models is the size of its training data. ChatGPT's training, for example, includes the Common Crawl data set – a large data set that includes billions of web pages.

This is important to remember. No matter how amazing and seamless these generated responses are…