I love the "There may exist some <greek symbol>...or there may not, it depends"
And I hate when they replace x with something and you wonder, "What's that squigly line". then the teacher responds, "That's basically like x"
Or when they use Theta sub Lambda in the same equation as Lambda sub theta and the instructor gets confused and mixes them up but doesnt realize it until half an hour later.
The point is, I hate math.
It even gets better when you are using L in formulas where in one part L means the inductance and in the other the length of the transmission line... (I think this occurs somewhere in the telegraphist's equation).
My motto about math and every other science is that nothing is rocket science, as long as you break the rocket science into small understandable parts... But I admit that breaking the rocket science into smaller parts can be rocket science itself at first.
I'm waiting for the half-life toilet-kill icon becomes a physics constant.
When dealing with AI, it seems use of greek letterage is a free for all. All 3 forms of sigma are used, alpha means something different for each type of neural network, and various other symbols have pseudo-constant meanings. A lot of this becomes a prolem when interfacing complex netowrks with physics functions, since you are almost guaranteed to have a variable collision.