The lessons get you up and running. These are the longer reads for when you want to understand why — the ideas behind these tools and behind "AI" itself, with the jargon still left at the door.
What people really mean by "AI" is the judgment layer — software that makes a call instead of just following rules. What it is, how it differs from an ordinary program, and when you actually need it.
An AI can't decide anything until you put data in front of it — and gathering that data, then acting on its answer, is a coding job. Why data is where AI really lives, and code is the foundation underneath it.
APIs sound technical because the word is technical. The idea is not: it is the approved way for one program to ask another program for your data, with permission and boundaries.
For beginners, they are close enough that the best first choice is usually the one you already pay for. Learn the loop in an IDE, and stay tool agnostic as your preferences change.
One helps you hand off the task. The other helps you build the tool that makes the task stop being yours. Same scattered feedback pile, two very different paths.