Whilst it is not framed as such, there are some interesting high level takeaways in here for AI, particularly around e.g program synthesis/induction and ARC.

I think Wolfram could contribute a lot if he shifted his focus towards these domains.