Why Facebook thinks AI can help you land better dates (www.businessinsider.com)

Meta rolled out an AI assistant inside Facebook Dating that lets users skip endless swiping by asking a chatbot to find compatible matches or give dating advice. Built on Meta’s Llama models, the assistant searches the dating ecosystem for explicit, publicly listed profile signals (interests, locations, etc.) — for example, “find someone who goes to EDM concerts” — and can also suggest what to say to a match or where to go on a first date. Meta stresses the tool doesn’t infer hidden traits or adopt a human-like persona; it only uses information a user has made visible. The company is launching the feature in the US and Canada and is also testing “Meet Cute,” a weekly automatic match generated by Facebook’s matching algorithm. This matters because it signals a shift from swipe-driven discovery toward conversational, ML-mediated matchmaking that aims to reduce “swipe fatigue” and surface more relevant leads. Technically, using Llama lets Meta combine natural-language prompts with structured profile metadata to translate vague desires into search filters, which could improve match relevance and engagement. The move also joins broader industry trends—Bumble and Hinge are exploring AI features—raising questions about model bias, user trust, and how personalization at scale will be governed, even though Meta positions this rollout as limited to explicit user-provided data and non-monetized for now.