3 ms·
Optimizing for Mutual Physical Attraction (OMPA) Saint Louis | CEO | REMOTE OMPA is currently a PyTorch model that predicts if you’ll find a face attractive or
by wildermuthn 5y ago
Optimizing for Mutual Physical Attraction (OMPA)
Saint Louis | CEO | REMOTE
OMPA is currently a PyTorch model that predicts if you’ll find a face attractive or not. It requires about 10 minutes of work from a user to gather enough data to train an accurate model specific to that user. It involves ranking 7 faces from most attractive to least attractive.
I have a working model that is good enough to be used as an auto-swiper, but my goal is more ambitious: solving the “stable marriage problem” and fixing digital dating’s well-understood problem: the most attractive users capture almost all the matches, leaving normal people like myself with less interest than we’d receive offline.
Also, we need to Optimize for Mutual Emotional Attraction. But one hard problem at a time! :)
To do this I need an equal partner who is great at “Product” — i.e., using an innovative technology like OMPA to create a product that people will actually find value in.
OMPA isn’t a product yet. It’s just a tool. The question is whether that tool enables the creation of a product capable of fixing digital matchmaking. I believe it can, but know myself well enough to know that without a brilliant CEO guiding it, OMPA will remain just a cool toy I tinker on.
I’d love to schedule a demo of anyone who is interested: nate.wildermuth@gmail.com