It's been a while since I was told this, so not sure how true it still is, but there a was a niche but lucrative market for people who could maintain stuff in Fortran, COBOL and the like.
Because there were some critical antediluvian pieces of software in banking, big businesses, etc that some companies were terrified of having to replace one day.
I'd expect that by now most would have migrated to more common languages, but I don't really know.
A lot of codebases in skuff like fluid mechanics, meterological models, quantum mechanics etc. are still in Fortran. Largely because there is very little to gain from rewriting the code base in some other language.
I would choose Fortran for a new project 0/10 times, but to be fair, it's a completely viable language for developing complex and computationally intensive models, and it's better to have the 1-2 new guys learn Fortran every year than to rewrite a 200k line code base in some other language that offers few or no real advantages outside of personal preference.
I used to laugh about COBOL, who can take it seriously, right?
But now that I'm working in a larger company I can assure you that COBOL is the backbone of large industry. Aviation, finance, railway operations, insurance, you name it. We're talking software running nonstop for decades.
To be honest, the first draft of this had the Shakespeare Programming Language in the last panel but the test audience (ie my co writer) had never heard about that one, so I changed it to something that wasn't necessarily bad but rather just old and no longer really in use.