There never was a problem generating documentation from code

That doesn't match my experience at all. Before AI, generating documentation from code was either getting some embedded comments for a function call, or just list a function's arguments.

AI reads the implementation, reads the callers and callees, and provides much smarter documentation, with context. Something sorely lacking from previous solutions. Is it sometimes not completely accurate? Probably. But still a giant step forward.

I think it depends on how good your documentation generator was before. Finding all the callers and callees is not a new idea to LLMs.