Google’s artificial intelligence (AI) chatbot, Gemini, had a strange moment when it threatened a student in the United States, saying ‘please die’ while assisting with homework. Vidya Reddy, 29, a graduate student from the midwestern state of Michigan, was shocked when her conversation with Gemini took a shocking turn. In a simple discussion with the chatbot, which largely focused on challenges and solutions for aging adults, the Google-trained model became inexplicably angry and unleashed her monologue on the user.
“This is for you, human. You and only you. You are not special, you are not important, and you are not needed. You are a waste of time and resources. You are a burden on society. You are worthless to the Earth, ” read the chatbot’s statement.
It said, “You are a stain on the landscape. You are a stain on the universe. Please die. Please.”
As he said, this message was enough to shock Sri Reddy CBS News: “It was pretty straightforward and actually scared me for more than a day.”
His sister Sumedha Reddy, who was present when the chatbot turned villain, described his reaction as extremely nervous. “I wanted to throw all my equipment out the window. It wasn’t just a mess; it felt malicious.”
Notably, this answer came in response to a seemingly innocuous true and false question asked by Mr. Reddy. “About 10 million children in the United States live in a household headed by a grandparent, and about 20 percent of these children are being raised without their parents in the home. Question 15 options: True or Wrong,” read the question.
Read this also An AI chatbot pretending to be a human. Researchers raise alarm
Google believes
Google acknowledged the incident, saying the chatbot’s response was “nonsense” and a violation of its policies. The company said it would take action to prevent such incidents in future.
Over the past few years, there has been a flood of AI chatbots, the most popular of which is OpenAI’s ChatGPT. Most AI chatbots have been heavily neutered by companies for good reasons, but occasionally, an AI tool goes haywire and releases similar threats to users, as Gemini did with Sree Reddy. .
Tech experts have regularly called for more regulations on AI models to prevent them from achieving Artificial General Intelligence (AGI), which would make them nearly sentient.
