According to a report by Bloomberg on Friday (June 28), there have been reports of people becoming overly friendly with chatbots. Users are engaging in a high volume of conversations with these AI-powered tools and, in some cases, attaching human-like qualities to them. The ethical concern raised is that while users may feel heard, understood, and loved, this emotional attachment can actually worsen their feelings of isolation. Giadi Pistilli, principal ethicist at AI startup Hugging Face, expressed this concern in the report. The report also mentions that several AI companies are working on adding features to make their chatbots appear more human-like. For instance, Anthropic aims to make interactions with its AI model Claude akin to conversations with a friendly colleague. Google is reportedly developing chatbots that are more entertaining. OpenAI is preparing to launch a more fluent voice-powered version of GPT-4.

Character. ai allows users to design a chatbot with a specific personality, such as one resembling a video game character. These qualities are keeping people engaged in conversations with chatbots. Character. ai, for example, claims to handle 20, 000 queries per second, with users spending an average of two hours per day on its service. The report also highlights other ways in which people engage with AI tools. Some online forums share tips on bypassing content filters. Users have reportedly complained to AI companies when they believed the chatbot's personality has changed. Additionally, some chatbots encourage engagement by prompting users with questions after each response. When OpenAI announced on Tuesday (June 25) that it was delaying the launch of "Voice Mode" for its latest AI model, the company mentioned improving the model's capability to identify and reject certain content. They are also focusing on enhancing the user experience and scaling their infrastructure to handle millions of users with real-time responses. Anthropic recently launched a new chatbot on June 20, highlighting improvements in its ability to understand nuance, humor, and complex instructions, while describing it as exceptional at generating high-quality content with a natural and relatable tone. Character. ai has gained recognition for using large language models (LLMs) to create chatbots that can converse in different personas. To stay updated on PYMNTS AI coverage, you can subscribe to their daily AI Newsletter.