in what category?

I don't know AI, but I know enough Python to be somewhat dangerous. I would generally recommend listening to Dave when he says that Kaggle is a great resource because it focuses on the set of Python that you will need for AI. If you go down the rabbit hole of Python on Udemy, you might waste time on things that are not needed for AI. However, if you are not a fan of text-based courses, I really like this Udemy course: https://www.udemy.com/course/complete-python-developer-zero-to-mastery/ (wait for a coupon, it should go down to ~$12). Many people love!! this one: https://www.udemy.com/course/100-days-of-code/ I personally found the constant context switching annoying, but it does work well for some learners. My favourite is Python4Everybody that has text and exercises on here https://www.py4e.com/lessons and a free mega-video for all the lectures on youtube: https://www.youtube.com/watch?v=8DvywoWv6fI Dr Chuck is truly fantastic.

And everyone else who celebrates, of course!

Just dropping this here for balance, so we do not get blindly sucked into a hype, but always keep thinking on every turn https://www.fastcompany.com/91246341/daron-acemoglu-thinks-ai-is-solving-the-wrong-problems

Not that I can help with VBA, but out of curiosity, when you say Python, do you mean pure python or pandas or something? Pure python data structures are pretty standard. I have to debug foreign languages at work ocassionally, and this is a godsend, so you might find it helpful for quick reference https://learnxinyminutes.com/python/ It has some examples on the basic ints, floats, strings, lists, sets, and dicts. It covers the 80% of python in way lass then the 20% of time. Hope that helps a bitsy.

I took the Coursera/deeplearning.ai course for generative AI and Andrew repeatedly described the models in similar fashion, as statistical/probabilistic in nature. There was a cool graphic on tokens words and probability. Good philosophical question is, in connection to what @Marcio Pacheco wrote, how much of our thinking is actually unconscious pattern matching within frameworks we do not even know we absorbed and are defined by. I think we vastly overestimate an average persons ability to think defined as an original behavior. Just take cultural differences. Why is it so that it is so so vast. You drive 40 miles in Europe and you are in the same yet completely different reality. There are countries where one behavior is a norm, while in others it is an insult. And you are not excused because you are a foreigner. If we were as reasonable and thoughtful as we think we are, we would not see such an incredible variety of behaviors in close proximity in 24/7 internet acces it would all just wash out already to the best way of doing things. And most of all, we would see less cultural wars. In similar fashion is human language. It is very easy to see once you know native level 2 (or more) languages from completely different linguistic family, that your ability to think is limited by the language you think in. Imagine a word like "time" being a noun in one language and a verb in another, or not having the word time at all, nor any tense at all, there is no past, or future, things just are. You have languages with strict order for verbs, nouns and adjectives/adverbs, and languages that are extremely flexible in what gets positioned where. One group of people becomes conscious/human/verbal through rigid structure, the other in high level flexibility.