"When I first started using apps like Tinder or Bumble, I could feel myself slowly wanting to be a solipsist," Joshua recounted, the dejection still palpable at 31. "People ghosting one after another or simply not responding...it felt like you were commodified and you saw exactly what your worth was."Growing up in the internet age, Joshua watched as meeting people shifted from the organic corridors of real life to the curated digital world. Even in high school it wasn’t unusual for friends on different niche interest forums, online games communities, or even Facebook groups to become more than friends. It was after college, as his in-person social circle shrank, that Joshua jumped himself to the ubiquitous dating app scene. "Dating apps make people think that their choice is unlimited in some cases, or that they are alone in the universe in others," 33-year-old Bostonian Anton told The Miilk. "The first kind get overwhelmed by options and cannot make rational choices, the second get desperate and blame themselves for what they don't control."Apps and algorithms have dramatically redefined how we pursue love and relationships in the modern age. The pool of romantic potentials has expanded exponentially, allowing people to meticulously filter for an alarming number of preferences at any time. But as the options multiply, so does the ease of casual rejection through normalized behaviors like "ghosting".The impact of technology on human connection has become a contentious societal debate. While some celebrate newfound ability to meet partners outside one's typical social circle, many others lament the soul-numbing fatigue and dehumanization of contemporary dating's supply-and-demand economics.What's clear is that apps and AI-driven algorithms have irrevocably shifted how people first encounter romantic prospects - from the off-line world to the online realm. This transition is still underway, as evidenced by dating companies now pitching generative AI as a solution to revive stagnating user growth. However, the impacts of deploying such AI remain unclear given the problematic biases ingrained in the datasets used to train these models.