1. A few other libraries still depend on jQuery in turn, so I find that I still usually have it available in every app I work on anyway, so I might as well use it.
2. It "just works" and I already know it, and since it's already there, anyway (see (1) above) it just makes sense to use it. IOW, I don't feel enough pain from using jQuery to justify going off and learning what combination of new ecmascript (es6? es7? es9037253223252??? ???) features and/or libraries du jour that I would need to do everything jQuery does for me now.
That said, I mostly just use it for the selectors - which I understand can be easily replaced with native features now - and the ajax method. So in reality, yeah, I probably could ditch jQuery without too much pain, excepting for (1) above, but... why bother? I don't care too much about the size, since I don't typically link in tons of javascript libraries, ad networks, etc. to kill page load times, and page bloat seems to be the biggest argument I hear against using jQuery.
Plus, wouldn't actually using Vanilla JS be faster?
Right now it looks like people use JQuery for legacy code. But, they give no reason to use it for new projects, except if they need it for other libraries they're going to use. (based on legacy code anyways...)
JQuery looks redundant in the modern Javascript era on evergreen browsers.
So jQuery is still wonderful at what it does, but there isn't much reason to be directly manipulating the DOM now that there are alternatives.
It's a bit like the perl of web page JS for me. I wouldn't recommend writing a full complex application in it - although that is possible and by people who know what they're doing can be done cleanly and organised... but for super quick hackery, it's hard to beat.