Many students want to jump straight to large language models (LLMs) and generative AI, but without understanding regression, classification, and how simple neural networks work, a deep understanding of LLMs stays superficial.
My advice has always been: start with simpler models (linear and logistic regression) to really understand loss and gradient descent, then move to neural networks and finally more advanced models. That's exactly the order we follow in my "Introduction to AI" course.
For a full overview of this learning path's structure:
Full Structure of Google's ML Crash Course: https://developers.google.com/machine-learning/crash-course
Blog