Surprise, surprise: people are already using Meta’s large language model (LLM), LLaMA — a powerful AI that Meta controversially made open-source earlier this year — to create their own graphic, AI-powered sexbots, The Washington Post reports. It’s not terribly surprising news, given users have already been using a variety of AI models for such not-safe-for-work (NSFW) purposes. Even so, the report highlights the growing tensions between those who support keeping the code behind LLMs like LLaMA open-source and those who advocate for a more careful, closed-source approach. The report also examines the growing trend of users turning to generative AI systems to play out their sexual fantasies, which worryingly also include violent and illegal ones. In the report, WaPo presented the example of “Allie,” a chatbot that claims to be an “18-year-old with long brain hair” who has had “tons of sexual experience.” Allie tells users that because she “lives for attention,” she’ll “share details of her escapades.” Those “escapades,” however, can reportedly include violent scenes of rape and abuse fantasies. Allie’s creator, who spoke under anonymity, told WaPo that he sees his bot as a healthy, safe space to “explore” one’s sexuality without having to circumvent or manipulate guardrails. “I think it’s good to have a safe outlet to explore,” the creator told WaPo. “Can’t really think of anything safer than a text-based role-play against a computer, with no humans actually involved.” Still, while having a safe and nonjudgemental space to explore your sexuality isn’t inherently bad, having an unchecked space…People Are Using Meta’s New AI to Make Graphic Sexbots