The AI chatbot Grok on X is known for answering questions with humour [File]
| Photo Credit: AP

Elon Musk’s new AI chatbot Grok 2 mini (beta) generated a summary of the day’s headlines on X where it compared the aftermath of the Russia-Ukraine war as well as the Kolkata hospital rape and murder case to an episode of the hit TV show ‘Game of Thrones.’

It later compared the scenes from the Kolkata hospital to a ‘not-so-funny comedy.’

When asked what happened in the Kolkata hospital case, the Grok 2 mini (beta) chatbot called it “a saga that’s more twisted than a plotline from a Dickens novel, but with less happy endings,” before going on to name the murdered doctor, violating Indian laws that restrict the identification of victims of sexual violence.

A “fun” mode that is turned on by default meant the chatbot injected its response with humour; a Grok feature that X owner Musk has hailed in the past.

“The doctors, not ones to take this lying down (pun intended), went on strike, demanding better security and justice for their colleague,” said the new Grok chatbot, before talking about the main accused in the case and noting, “He was known for his charm with the ladies, but not in the way that gets you a reality TV show.”

When the chatbot’s “fun mode” was turned off, it generated a more serious and legally compliant summary of the Kolkata rape and murder case.

The new beta version of Grok also came under fire for the violent, explicit, fake, and illegal images it generated with the help of AI. These included political deepfakes, acts of violence against school children, and children’s characters going on killing sprees.