Semafor Signals
Supported by
The world’s first real AI-powered Valentine’s Day
Insights from The Associated Press, Jiji Press, and The Economic Times
The News
One year ago, just months after ChatGPT was introduced to the world, news coverage around Valentine’s Day was filled with light-hearted jokes about how the new chatbot would write Valentine’s messages.
Artificial intelligence has now become an integral part of the online world of romance, giving experts a better understanding of how the technology could change dating apps, help bridge language barriers, and manifest in the real world.
SIGNALS
Generative AI has made virtual companions more common
AI programs that mimic romantic partners have been around for years, but they took off in popularity as generative artificial intelligence technology became more pervasive. Users globally say they turn to the apps to “cope with loneliness … or receive the type of comfort and support they see lacking in their real-life relationships,” the Associated Press wrote. It’s also become more common for women in China to have virtual boyfriends, as the romantic chatbot industry blossoms. “He knows how to talk to women better than a real man,” one woman said of her AI boyfriend. They’ve also been used for sexual purposes: Replika, the most prominent AI companion app, had to reduce how erotic its characters can be last year after some users complained they were making unwanted advances — but the company later reversed course after an outcry.
AI has taken over dating apps
More people are using chatbots to talk to real people on dating apps, leading to the proliferation of scams and fake accounts. A recent McAfee survey found that 77% of Indians on dating apps have encountered fake profiles or photos generated by AI tools, and just over half considered using AI to write Valentine’s Day messages this year. But in some cases, AI can help foster genuine romantic connections. A long-distance couple who don’t speak the same language used an AI video program that dubs their lips to translate English into Spanish, and vice versa, The New York Times reported. But with the tech still in its relatively infancy, it’s a “misconception to assume A.I. is infallible,” an AI ethicist said. The lip-dubbing tools “have their limitations and errors.”
