A reader was excited to share an article about a new cryptography method from MIT that would enable you to find out if someone is attracted to you, without revealing your own attraction to them. Could we implement a online dating service with this protocol?
I am generally skeptical about technological solutions to human problems, so I checked out the article’s academic credibility before actually reading it. There were only 11 references cited, including some cryptography papers (one of which the authors admitted to not fully understanding) and also Shakespeare, Fiddler On the Roof, a Hollywood movie, a (very relevant) XKCD cartoon, and a silly poem about gender expectations in dating. The article was an anonymized pre-print so I had to search for information about the authors. That led to a Hacker News discussion of the paper, which clarified that it was a satire presented at an academic-humour conference. (Note that the Hacker News discussion is not entirely sympathetic to socially-awkward people.)
Even if I had not done these credibility checks, the paper itself (in section 4.2) states why their proposed method would not actually work so well. If one person asks one other person to use the method, they have just revealed their attraction, and are now at risk of rejection. It works better to recruit many people to use the method, but what if the person you want to know about isn’t participating?
If it sounds too good to be true, it probably is.
As the Hacker News commenter DennisP says, there are already online dating services like Tinder that will reveal who “likes” you, if and only if you “like” them. No need for fancy cryptography, just a third-party database that lots of people use. Of course Tinder is not to everyone’s taste, and it might not include the person you are crushing on.
So I think we need a human approach to the challenge: if you are crushing on your friend, should you reveal your attraction? How to handle the risk of rejection? That’s a topic for another post.