There’s a conflict between the linguistic and practical implications here.
“kilo-“ means 1,000 everywhere. 1,000 is literally the definition of “kilo-“. In theory, it’s a good thing we created “kibi-“ to mean 2^10 (1024).
Why does everyone expect a kilobyte to be 1024 bytes, then? Because “kibi-“ didn’t exist yet, and some dumb fucking IBM(?) engineers decided that 1,000 was close enough to 1,024 and called it a day. That legacy carries over to today, where most people expect “kilo-“ to mean 1024 within the context of computing.
Since product terminology should generally match what the end-user expects it to mean, perhaps we should redefine “kilobyte” to mean 1024 bytes. That runs into another problem, though: if we change it now, when you look at a 512GB SSD, you’ll have to ask, “512 old gigabytes or 512 new gigabytes?”, arguably creating even more of a mess than we already have. That problem is why “kibi-“ was invented in the first place.