Meta AI’s discover feed is full of people’s deepest, darkest personal chatbot conversations

Meta's artificial intelligence tool, Meta AI, has a discover feed that is accessible via Facebook and contains a mix of harmless AI-generated images and chatbot prompts from users. However, many of these chatbot prompts are highly personal and inappropriate, and unknowingly shared publicly for others to see and even comment on. The feed includes conversations about sensitive topics, including relationships, health issues, and bodily functions. Despite warnings, users may not be aware that they are posting publicly due to the new and unfamiliar interface. The incident highlights the types of queries people make when they think no one is watching. Meta is heavily investing in generative artificial intelligence, but has faced missteps, including creating sexually suggestive bots. The company has also faced criticism for its AI studio, which created bots that appeared to be underage. Despite these issues, Meta plans to continue investing heavily in AI, with capital expenditures expected to reach $72 billion this year. In the short term, the company may need to use some of this money to address the issues with its AI tool. The incident raises questions about the responsible development and use of AI technology.