As others have mentioned, notes help some individuals remember things even if they never reference them again. So I won't stress that point anymore, although I agree with it, especially for class notes.
As for notes on your reading, notes on papers are extremely useful. Depending on your field, you might end up reading lots of papers from which the details of methodology don't matter, but the key results do. If you make note of the results, then you can reference them without having to reread the papers.
Sometimes, when I am doing modeling work and doing background reading on experimental results I will take this to the extreme. As I read the paper, I will takes notes (in Mendeley or a personal wiki) and at the end include a section called "cite this paper in the following contexts" in which I will proceed to write down contexts where I expect these results to matter in my future work.
When I am writing my own paper, I might remember "there was something I read about the involvement of this in the amygdala". I do a quick search on my wiki, and from my notes have access to the relevant papers in a much faster way than going through the countless papers I might have read looking for the result.
Note taking makes my research (especially writing) more time efficient down the road