buttsnguts asks:
Sorry-- have you thought about apply machine learning, or artifical intelligence, in your research-- for example, to classify body type or preferences for certain ones from photos?

What do I even know about AI? 

My experience with “machine learning” is very limited.  I’ve built a few systems using naive bayesian classification, decision trees, and clustering.  I have a preference for unsupervised learning systems, but usually work with data sets that are too small for that.

I do not consider myself knowledgable enough to provide a technical answer, but I will be happy to address the principle thesis of your question.

Let’s teach computers to classify us based on our body type. Surely, this will provide immense benefit to the gay community.

I actually built a “dumb” classifier for the gay cliques census.  I’m too lazy to look at the code, but there’s, like, 30-dimensions of data and almost 2-million rows of data.  Definitions are reinforced by people who take the survey, additional data is gathered from people who use the definition tool, and negative feedback is used when people disagree with their classification.

Super basic machine learning system.

All thats missing is a piece on computer vision to assess photos for these dimensions.

Its possible, but why would you do that?