> But in STEM fields, directory structure remains crucially important. Astronomers, for example, may work with hundreds of thousands of files in the same format — which can be unwieldy to scale to a searchable system, Plavchan says.

Anecdotally, I've found searchable systems to be lukewarm. Search engines like Google are good for out-of-the-blue questions, but when looking within local files or data inside programs, search systems struggle to get what I need quickly. If I remember a specific file name, Spotlight search on macOS usually tells me what I need, while the default search on Windows does more or less the same. But these search systems never really help me find things in-depth, perhaps some mathematical formula embedded somewhere in a PDF file compiled from LaTeX. Or these search systems are not as forgiving when I'm trying to look for something that isn't text-based, like pictures or 3D assets.

Search is complex, so what can I do to depend less on these built-in systems? How can I make things more searchable in general? What methodological approach can be adopted to make searching for something, regardless of context, better?

Mindful folder structures, tagging files, consistent naming, and associating content to regular text as much as I can is a start, but I wonder if there's more that I'm failing to do.

[1] https://www.theverge.com/22684730/students-file-folder-directory-structure-education-gen-z