Really? How familiar are you with how inodes work on a spinning platter vs an eMMC? Can you speak intelligently about how virtual machines segregate their ram boundaries whilst still maintaining decent performance? What is your familiarity with FIPS?
And you were born with an understanding of how a keyboard and mouse worked or how windows seperated different GUI contexts and what all those symbols around them meant? I'm not saying you have to understand the nuts and bolts because abstractions exist but those abstractions still present you with a model and you need to understand that model to use them. People often present git's model mixed in with its nuts and bolts, which I think is unhelpful because the model is actually really simple and the nuts and bolts are not so important.
Because they've already learned a model of how devices with screens work, albeit an imperfect one. Children are pretty good at learning. Software and systems should be easy to learn, but it's not a case of no learning.
I agree that git's UI is not particularly easy to learn, and it could be better in that regard (which is frustrating because its underlying model is about as simple as you can come up with, and I essentially can always think of what I want in that model easily but then have to google how to tell git to do it). What I object to is the idea that needing to understand something to use it is a problem.
You’re very incorrect.