Who would anyone want to pair a subpar language with a subpar ML framework?
That is the thing, what lessons were learnt from it, and how will Mojo tackle them.