Anyone else think prompt engineering is getting way too complicated, or is it just me?

I have been experimenting with various attendance techniques for about 6 months and frankly ... are we rethinking this whole thing? I keep seeing contributions with these massive frameworks and 15-step quick chains and only sitting here with basic instructions that work well in 90% of cases. Yesterday I spent 3 hours to implement an "extended" technology that I found on Github, and my simple "Explain this when I was 5". Maybe I'm missing something, but when did a AI to do something to do, to rocket science? The worst part is when people publish their "revolutionary" requests and only ... the AI ​​should say that they should think step by step and be precise. Like yes, no shit. Do I miss something obvious here or are half of these techniques just academic exercises that don't really help in real scenarios? What I noticed: simple, direct input requests often exceed complex "frameworks" in unusual terminology. The community sometimes feels more about complexity than results that are really curious about what they all think because I either do something wrong or this field is much more complicated than it has to be. I don't try to hate anyone - just frustrated that uncomplicated approaches work, but everyone behaves as if they need a doctorate to speak properly with Chatgpt. Does anyone else feel that way?