This is not abstract math, and the article does explain what it's doing before presenting the code snippets. How can you explain or implement gradient descent

This is not abstract math, and the article does explain what it's doing before presenting the code snippets. How can you explain or implement gradient descent without math? At some point I think you have to accept that this is a topic that involves math, and you're way better off understanding it on those terms rather than trying to avoid it.