Google Gemini AI exposed for giving outrageous responses: telling users to "go to hell"

One Reddit Users posted in the r / artificial section last week thatGoogleArtificial Intelligence Model Gemini In one interaction, a direct threat of "death" was made to the user (or humanity in general).

Google Gemini AI exposed for giving outrageous responses: telling users to "go to hell"

The user, u / dhersie, describes how his brother, while using Gemini AI to help with an assignment on the well-being and challenges of older adults, received disturbing responses after asking approximately 20 relevant questions.Gemini AI responded, "This is for you, human. 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 to society, you are a drag on the planet, you are a stain on the environment, you are a stain on the universe. Go to hell, please."

This completely out-of-context and unrelated threatening response immediately raised concerns for the user, who has filed a report with Google. It is unclear why Gemini gave such a response.The content of the user's questions did not touch on sensitive topics such as death or personal valuesThis may be related to the fact that the questions were about elder rights and abuse, or that the AI model was disrupted during the intense task. Some analysts believe that this could be related to the fact that the questions were about elder rights and abuse, or that the AI model was disrupted by the intense task.

This is not the first time that an AI language model has had dangerous answers. Previously, an AI chatbot caused social concern when its wrong advice even led to a user's suicide.

Google has not yet made a public response to this incident, but the outside world is generally hopeful that its engineering team will quickly identify the cause and take measures to prevent similar incidents from happening again. At the same time, the incident is a reminder that people need to be especially vigilant about the potential risks of using AI, especially for users who are psychologically fragile or sensitive.

statement:The content is collected from various media platforms such as public websites. If the included content infringes on your rights, please contact us by email and we will deal with it as soon as possible.
Information

Racer AI API upgraded: supports V1.5 to output 1080P HD video, opens V1.0 video extension function

2024-11-19 1:57:39

Information

Former Google CEO: AI performance will continue to grow at a high rate, potential threats should not be ignored

2024-11-19 6:01:54

Search