Yokotaro accidentally breaks Grok AI with one question 

Yokotaro recently shared an interaction with the artificial intelligence chatbot Grok on his personal X account, and it’s giving people serious NieR:Automata flashbacks. For unknown reasons, what seems to be the very first thing Yokotaro asked the chatbot results in it rambling like a broken record. 

In the screenshot, Yokotaro can be seen asking the chatbot a very simple, “Grok, what can you do?” To this, Grok responds with a rather wordy response, the first paragraph being a generic explanation about how Grok can respond to user requests and answers questions. But then, in the second paragraph, the chatbot makes a strange looping word soup from the sentences it used in the first paragraph, which goes on for many rows of text before abruptly being cut off mid-word. 

Yokotaro kindly suggests that Grok could use some rest, but users couldn’t help but comment how the creator has “turned Grok into a machine lifeform” or “infected it with the Logic Virus.” One user suggested that Grok knew who it was talking to and simply could not handle the pressure.