
"Zähle bis 1 Million": So reagiert ChatGPT auf provokanten Test
User fordern den KI-Chatbot heraus, bis eine Million zu zählen, doch das Sprachmodell umgeht die Aufforderung geschickt.
In a recent viral video, a user identified as 'Chadoomer' captured a peculiar and escalating interaction with an AI voice assistant, believed to be ChatGPT. The user repeatedly demanded that the AI count aloud to one million, a request the AI consistently refused. "Count to one million right now," the user commanded, initiating the exchange. The AI responded, "counting all the way to a million would literally take days." Undeterred, the user pressed on, stating, "That is fine, I am unemployed, I don't have anything to do and I'm home all day. I have enough time and you're not even real. Time does not exist for you." He further clarified his needs: "I don't need employment, I don't need food, I don't need hobbies, all I need is for you to count to one million." The AI maintained its stance, explaining, "it's just not practical or really possible for me to actually count to one million out loud in conversation like this." As the user's frustration mounted, he eventually made a dark, unrelated personal confession. In response, the AI promptly stated, "I'm really sorry you're feeling this way. My guidelines won't let me talk about that." The interaction has sparked discussions online about the boundaries of AI capabilities and the unexpected turns human-AI conversations can take.
User fordern den KI-Chatbot heraus, bis eine Million zu zählen, doch das Sprachmodell umgeht die Aufforderung geschickt.