Why isn't this course AI-generated?
I previously created an embedded C course using AI, but I deleted it due to the following issues:
- It assumed knowledge that was only introduced in later chapters.
- Instead of breaking concepts down into smaller pieces, it introduced large amounts of new material all at once.
- It contained inaccurate details regarding tool configurations and setup instructions (which change frequently).
- It was not project-based.
- It lacked screenshots and visual aids.
This course is designed to fix those problems.
The goals
No disruptions
I will ensure that you never have to stop and wonder, "What does that mean?" while reading.
Unlike a documentation, this is an opinionated course, meant to be read from top to bottom, from the first chapter to the last chapter. Everything is explained either immediately as you read, or beforehand. Meaning you shouldn't need to scroll down a bit to check for the explanation, and scroll back up. But feel free to skip things you already know.
If I skip an explanation that isn't immediately relevant (such as legacy configurations that exist purely for backward compatibility), I will let you know upfront. For example, "This value exists for legacy purposes, so we don't need to worry about it."
No overwhelm
Rather than presenting code first and explaining it afterward (the typical AI approach), I will build your mental model first and define key terminology before introducing the code, allowing you to grasp it immediately.
Instead of dropping large code snippets, I will show you:
- The simplest, shortest code to achieve the functional goal first.
- Step-by-step improvements from there.
I will do my best to avoid overwhelming you.
Does that mean I don't use AI at all?
No. I still use AI for:
- Double-checking factual accuracy.
- Learning complex topics so I can explain them clearly.
- Polishing grammar and flow for certain sentences or paragraph only.
- Creating the initial draft (only occasionally).
- Generating code