Can We Use Demographic Data to Make Content Recommendations Better?

I definitely see the argument for removing some of the demographic targeting, yet there are likely, I'll see ads that are not going to apply to me often. What could be a proxy of gender for showing kind of products which people may not be interested is if someone is looking into that kind of product or doing that sort of engagement which can lead to people's interest in certain kinds of products. People may not be comfortable with their years of data being shared. Can we use only one week of people's behavioral data, like what kind of things they are purchasing and how much engagement they have on the platform? Can these behaviors be used to model content recommendations better or ad recommendations better?