How do they accomplish this when the other/background noise is speech, like in a crowded area; where you only want the person near you amplified?

One method is using directional information. My low-end aids have two microphones each, and can share audio information from one aid to the other. I believe there are also some other methods, but I don't know how those work