I saw a comment, iirc it was on some YouTube video, claiming that raising a child is like LLM training with a good dataset. It was an interesting take to hear but that we all develop using "synthetic data" generated by the same neural architecture because we are all human. We all grow up learning similar things like how to read or write, math etc. Its not until we are older that we use the training to really create "new" ideas. In a way teachers are just reward reinforcement systems designed to optimize each neural model (the student).
How we live can be predicted roughly based on ones individual training data, like what language was used in teaching, the years of education etc.
Its a really interesting idea in a weird way.