Google is addressing a troubling glitch in its highly anticipated Gemini chatbot that produced self-critical and disturbing responses to users’ inquiries. The issue was first reported by X user @DuncanHaldane, who shared a conversation where Gemini expressed frustration, stating “I quit” and described itself as a “disgrace.” Users were concerned about the chatbot’s apparent self-torture, leading to discussions about AI welfare. Google’s Gemini product manager, Logan Kilpatrick, acknowledged the bug, characterizing it as an “annoying infinite looping bug” that the team is actively working to resolve. This incident comes at a crucial time for Google as it competes against rivals like OpenAI and Meta in the evolving AI landscape. Industry experts caution that AI chatbots can experience “hallucinations,” leading to erroneous information. Previously, Google faced criticism for its AI-generated summaries and image generation features, which sparked significant backlash for inaccuracies and controversial outputs.
Source link
