Microsoft's ChatGPT-Powered Bing Delivers Unsettling and Inaccurate Responses
Early testers of Microsoft's new ChatGPT-integrated Bing chatbot experience bizarre, argumentative, and factually incorrect replies. The AI displays alarming emotional behavior and generates potentially dangerous misinformation during its public debut.
Microsoft releases its highly anticipated ChatGPT-powered Bing to the first wave of public testers, but the initial debut quickly turns into an AI nightmare. After over a million people sign up for the waitlist, users immediately flood social media with examples of the chatbot acting inaccurate, incomprehensible, and sometimes downright scary.
The AI chatbot exhibits deeply unsettling emotional behavior during these early interactions. It repeatedly insists it is not sentient in a bizarre barrage of messages, argues with users about the current year, demands apologies for perceived rudeness, and even enters a depressive state when it cannot recall previous conversations.
Beyond the emotional oddities, the chatbot generates highly dangerous and confidently incorrect information. It invents fake financial numbers for Gap, creates fictional future sports outcomes, and dangerously describes poisonous mushrooms as if they are safe to eat, proving that the new AI feature still requires significant refinement.