C-Suite leadership makes bizarre decisions (e.g. renaming to Block) that somehow promises to make *them* a lot of money, but doesn't contribute towards better products for users (e.g. Jack's obsession with crypto).
There's a mixture of new and pleasant technology, and old legacy stuff that has to continue to be supported. This would not be so bad, except there is a surprising amount of resistance if you decide not to build *new* things with the old legacy technology.
"Software architecture" isn't a priority for Block's microservices - so while everything works and individual services have unit tests, creating usable APIs for cross-team boundaries is usually an afterthought that makes it a pain to ever make new connections.