“On a matchmaking app, if we ask you a question and your tone changes in response, that signals to us that you’re not telling us the whole truth. And so we’ll ask you the same question two or three different ways throughout your experience,” says Cohen-Aslatei. “We built it to mimic what a matchmaker would do for a client. The LLM is tracking changes in pitch and tone in your voice because we want to make sure we have an accurate understanding of who you are and what you’re looking for.”
After answering dozens of questions about lifestyle, future goals, boundaries, family, attractions, hobbies, and more over the course of a few days, Tai told me she would take the information provided and get back to me. Two days later, I received my first two potential matches.
I love you, girl alive
As a 31-year-old woman, I put my ideal age range at a healthy age of 26 to 40. My first two matches were 23 and 47. One was not alive when 9/11 happened, and the other had just graduated from college at the time. Bad start.
When a potential match is found, a picture of that person is blurred, and Tai gives you a summary of what would make a potentially good match. (You’ll need to provide selfie verification to confirm identity, and any unverified people will never be matched.) After that, you can click through to see a little more about them, like occupation, age, income, and a brief biography of what the AI creates.
At this stage of AI adoption, let’s just say there’s still a strong statistical bias against men who wear wraparound sunglasses and think driving a Cybertruck is the epitome of masculinity. Almost every one of the 16 matches I received during testing was Christian and wanted children as soon as possible, which Tai flagged as a potential issue every time. Many people were also initially flagged by Tai because they only wanted to date a certain caste or valued traditional gender roles, both of which I made clear I was not aligned with.
Out of journalistic duty, I accepted whatever match I got; Even an MMA-loving bodybuilder who enjoys grilling meat (I’m a vegetarian) and going to gun ranges (I’m generally anti-gun). Matches took place from Portland, Oregon, to DC, to New York City (where I live, although most of the matches were outside of NYC). Overall, none of the people I met were someone I would immediately hit on after seeing them on a traditional dating app.
If you accept, you either have to wait for the other person to accept or proceed with the match, or they will have already accepted, and you can start chatting. Here, your AI dating coach steps in to play wingman, providing hints based on the other person’s profile, highlighting your commonalities, and posing conversation questions based on answers to the match’s profile. Not only does the coach provide potential ice breakers (and responses), but you can also chat and ask for hints.
.png)
The Three Day Rule via Molly Higgins
I asked her for tips on how to break the ice with new matchsticks and she gave me advice, with an explanatory paragraph below each point. Tips include giving compliments, asking open-ended questions, using humor, referencing current events, sharing about yourself, and mentioning mutual interests. The advice was basic but solid, and reflected what the coach was doing with the conversation prompts given.
This is all a good idea in theory, and can be very helpful to people who have difficulty communicating with strangers. But this can also cause bigger problems. If the AI is doing all the chatting for you then you don’t really know who you’re talking to. And if you meet in person, you don’t know much about your partner’s real personality. You can tell a lot by how people type, the questions they ask and their sense of humor. All that was missing here.
