Like building an operating system, or a hypervisor, or a vm for some made up language, or an autograd engine, or something that sounds absurd. What about an optimizing compiler?
Maybe venturing into project Euler could help? What about learning about approximation algorithms? Or randomized ones? What about numerical optimisation?
Maybe the magic is not in the not knowing, but in knowing how things are done, seeing the connections between things and building beautiful abstractions that just work.
I think it is a sign of the times that a lot of people think there isn't fun or magic in programming anymore.
With the internet there's always know-it-alls going "why should I use your programming language instead of Python?" when you show people an experiment. Or maybe you build an internal app that solves a problem at work and there's 20 people saying "why didn't you just use Notion".
But even before that... I remember building a primitive text-to-voice synthesizer in the 90s and showing some friends, and they just said it was crap because they saw something on TV that was better. I remember even as a kid wondering why they were comparing the accomplishments of a 14 year old with IBM... On the other hand, I made some internet forums back then and on-line people people loved it.
So maybe ask yourself who's doing the yawning. Is it other people vicariously basking in someone else's glory? Or is it yourself not enjoying the journey.
EDIT: That actually reminds of magic. Some people say that magic tricks shouldn't be revealed. Well, I will admit that I yawn at magic and magicians, but I enjoy knowing how the tricks are performed.
When you perform magic for other people it should be on par with if not better than what they've seen before.
When you perform for yourself... Anything goes. You set your own bar.
The second group really won't care about how it's made or what can be learned, and won't care about sharing the experience... Which is fine, but makes their opinion absolutely useless, unless your product is targeting them.
I guess it's harder to find time to learn the basics -- vs the old days where all you needed was a few `for` loops and some math.
I struggled with that for years before finally accepting that the magic wasn’t coming back. I’ve since moved on to other non-tech interests that are giving me that sense of magic again.
I think what a lot of us found to be so magical and entrancing was the initial process of discovery and learning. At least for me, it was downright addicting to figure out how to bend the computer to my will, but the thing about that phase is that it’s kind of a one time deal. Once you’ve learned how to write software to a professional level, the only way to keep the magic alive is to keep going deeper. And frankly I got to a point where I didn’t want to spend any more time in front of screen.
Also started "labor" hobbies (car maintenance, welding, etc) but I think a professional mechanic or welder would say the same thing about their jobs. Why? "Professional" (aka money) ruins everything after a while. You get to a point where all you do is the same thing over and over, maybe with slight variations in between. Unless one has full, money no object, choice of what jobs to accept, this will happen.
Starting out, writing some Borland C++ code to render a pixelated 16-color 3D sphere on a VGA monitor was magic! Now, even a beautifully ray-traced photoreal landscape would be an unimpressive yawn.
How do you find that novelty? That magic? Is this burnout, or something different?