With enough practice, anyone can become good at anything. Knowing how and when to apply the basic principles will make you good at any artistic or technical endeavor. Knowing when to violate those principles is what will make you great.
This is part of the difficulty with machine learning, at least in the year 2023. It is easy to make machines that follow strict rules and parameters to mimic humans stiffly and robotically. But there is no understanding of context or nuance, so the performance is pretty bad.
Now we have machine deep learning and neural networks. These are a series of layers that process data and spit out an answer that it believes to be correct. It then compares its answer to what is true and re-processes the data backward to correct itself and make future processes more accurate.
This is how Google Translate can look at a hand-written sign in a foreign language and translate it on your phone. The "artificial intelligence" has been trained enough to understand the rough shapes of letters and return a translation that is (usually) very accurate.
Machines are great at answering questions, but they still don’t know the right ones to ask–unless a human gives them that data. Perhaps in the not-so-distant future, there is a world where machine intelligence has advanced to identify and understand the context better.
If I am ever to understand artificial intelligence fully, I might just need a computer to help. 😅