Dan Luu is great. If he's up your alley, you may also like Steve Sinofsky ("Hardcore Software"), Raymond Chen ("The Old New Thing"), and if you're in to the business & finance side of things, Patrick McKenzie aka patio11 ("Bits About Money").
My approach is mostly to curate your news/social/stream feed to individual people that post the sort of things you like, and to slower-moving publications. Limit your follows on social media. Prefer blogs and RSS over Twitter-style social. And/or read journals/magazines/books and reviews of them instead.
E.g. the old "The Morning Paper" blog by Adrian Colyer (https://blog.acolyer.org/) seems maybe up your alley? It's over now; dunno if what the new online versions of that are. For substantive programming stuff, these days I mostly read magazines and books, like CODE, Logic, and 2600. (IEEE Computer Society and ACMD have journals, but they're prety theoretical and academic; guessing that's not what you want here. I haven't found them very useful in my own programming work.) Sad to say, the current scene for this doesn't seem great. Old school Dr. Dobb's sounds like maybe what you want, but AFAICT there isn't really much like that any more? LWN and OSNews seem like modern online versions of stuff in that area, for system programmers at least. MIT Technology Review is also decent, but less about programming per se.
Big +1 to this. Much of the programming stuff I've read that's made a big difference to me and my skills and career has been the older "classic" and fundamental stuff. Unless it's directly related to the specific tools you're working with on your current project, and even specific issues with those tools, maybe let it age before you spend time on it.
"1+ year" actually sounds pretty short to me. Maybe 5 or 10 years to see if something has real insight and staying power and you want to spend your time on reading it. For stuff that's like, books or serious papers & essays, and not small news or release-notes or security-update posts. Many of the things I've read the last ~10 years that had lasting value were from the '90s or even '70s. (Though there's often updated editions of them, and AFAICT the new editions are pretty much always better.)
There's a "Somebody's Law" for this that basically goes "for a thing that's currently in use, the older it is, the longer it will probably stay relevant, and vice versa". Can't remember its name or exact wording, but it's basically reflecting that old things still in use have proven their value and flexibility, and are less likely to be just trends, and also have inertia and network effects.