Comment by rahimnathwani

Of the resources you mention, #2 is probably the best starting point for someone who wants to start building software soon. Karpathy's videos fast.ai's courses may also fit that purpose.

But the other books (#1, #3, #4) seem like they're intended for those who want to understand all the math. Many people don't want (or need) a full understanding of how all this works. They can provide significant value to their employers with some knowledge of how machine learning works (e.g. the basics of CNNs and RNNs), and some intuitions/vibes about SOTA LLMs, even if they don't understand transformers or other modern innovations.