Programming with Abstract Data Types (1974) [pdf]

> The programmer is initially concerned with satisfying himself (or proving) that his program correctly solves the problem. In this analysis he is concerned with the way his program makes use of the abstractions, but not with any details of how those abstractions may be realized. When he is satisfied with the correctness of his program, he turns his attention to the abstractions it uses.

I can imagine how to a modern reader the usage of the male pronoun here might sound excessive, but I think it's useful to note that this would be standard academic writing for the time. If I'm not mistaken, the switchover—away from using the male pronoun as a gender neutral pronoun—began (very slowly) in the late '70s. I remember reading an article Douglas Hofstadter wrote about how it bothered him once he became aware of the issue, but he'd mostly been oblivious until it was brought to his attention. The article also discussed various solutions to the problem, many of which are now common place (but if I'm not mistaken were fairly novel at the time). I think that article was from '79.

Edit: I was looking for the actual date Hofstadter's article was published, but I'm not sure since all the references I can find to it are from his '85 book Metamagical Themas whose contents was mostly (but not all) re-printing stuff he'd written starting in the late seventies. I noticed (just now) his article was included in a bunch of academic histories of gender in language, so it may be a decent reference point after all. So it looks like the idea to not use the male pronoun this way was pretty novel even in '85, probably.