09/25/2026
Major dating apps are now adding AI features, including AI-powered matchmaker services and “wingmen.” However, relationship scientists explain why AI is unlikely to reliably predict who singles will actually connect with.
Attraction Has No Clear Pattern
AI is good at finding patterns in huge amounts of data. But it can only find a pattern if one really exists. Research conducted by Amanda Montoya, an Associate Professor of Psychology at the University of California, Los Angeles (UCLA), and her colleagues, based on 313 studies on similarity and attraction, found that being similar may help predict attraction before people meet. Once they interact, however, similarity has much less to do with whether they actually like each other.
Preferences Don’t Tell the Whole Story
People can list the qualities they want in a partner, but that doesn’t always predict who they will actually like. Social psychologist Amie Gordon from the University of Michigan cites speed-dating research showing that people’s stated preferences can become much less important once they meet and interact with someone in real life.
Chemistry Has No Clear Signal
Amie Gordon points out that AI is trained around ideas of what makes two people compatible, but researchers still don’t know which factors reliably define compatibility. In fact, relationship researchers still disagree on whether romantic connection can be predicted before two people meet.
As a conclusion, AI may help sort information and reduce the number of choices. But it still can’t reliably predict what will happen when two people meet. Attraction, chemistry, and human connection remain more important than AI when it comes to finding a successful match. What’s your opinion?
