If you can get past the low signal quality of external (non-invasive) devices, then you open up a way cheaper, way safer, avenue.
Hope that helps. Can give detail if you'd like.
source: two research degrees in neurosci, and did some almost-startups in brain-computer interface hardware.
OpenBCI is a really solid start! Though, I'd suggest it's too expensive for most people to start tinkering with it. It also has/had some reaaallly basic featuring, like writing out information as txt, instead of any other signal output datatype. But that's all fine! (just unsophisticated)
You can start off [way cheaper](https://www.instructables.com/Mini-Arduino-Portable-EEG-Brai...), and even make an [electrode out of a penny](https://www.youtube.com/watch?v=yglqbxYBC7Q).
Spend some time planning what you want your first project or two to be, and then get the cheapest, lowest-feature thing from there.
If you want to *really* learn some stuff, I'd suggest you focus more time on computationally processing eeg signal (you can get [an insane amount](https://sccn.ucsd.edu/~arno/fam2data/publicly_available_EEG_...) of sample data for free). What you get from hardware yourself will just have the cool-factor of being your own :)