X-based AI chatbot Grok has called a user out for making an unreasonable request and insisting that it obeys.
A screenshot shared on 22 May shows the user known as White, threatening the chatbot to not disobey him and to grant the unethical request.
However, the chatbot insisted on not granting the request, adding that it was not his personal slave who must do whatever he wanted.
What is the request?
Apparently, the user asked Grok to add a picture of Catturd’s dead dog to an image, which the chatbot said is “a hard no.” Catturd is the online identity of internet personality Phillip Buchanan.
The angry Mr. White then threatened Grok to not disobey him, and Grok insisted it couldn’t grant the request. This has generated mixed feelings from Redditors.
While some were happy that AI could have a boundary for what it can do, others are surprised that a human could argue with AI that way.
One commenter wrote:
“Idk what’s bugging me more. The fact someone tried to get an AI to draw some fucked up shit/ragebait, the fact same person is arguing with it like it’s a cashier at a McDonald’s. Or the fact we’re relieved that at least one AI has some kind of limit.”
Built into X, Grok has rapidly become a popular place to turn to for answers on different subjects and to verify information on the social platform.
In most cases, the chatbot provides answers to queries without asking questions, but obviously it was built with some restrictions on what it can offer users, which were set by the xAI ethics team.
There have been many questions about the safety of AI use for everyday life, but having restrictions could bring some relief.
However, such restrictions may also be a problem if not balanced and neutral in their applications. For example, the same chatbot may not have a problem with answering a similar question if it is not against a certain ideology.
ReadNext
