Billionaire Elon Musk's company xAI recently updated a frequently asked questions page on its website, detailing a new mode for its AI chatbot Grok called "Unhinged Mode." This mode is designed to provide responses that are "intended to be offensive, inappropriate, and shocking," similar to an amateur stand-up comedian who is still learning the craft.
Musk first mentioned this mysterious mode in April of last year, but it has yet to be launched.
"Unhinged Mode" may be Musk's attempt to realize Grok's original vision. When Musk announced the launch of Grok two years ago, he described the chatbot as avant-garde, unfiltered, and anti-"woke," willing to address controversial questions that other AI systems would avoid. However, today's Grok takes a cautious approach to political topics, often leaning politically left on issues like transgender rights, diversity programs, and inequality.
Musk attributes this behavior to Grok's training data (public web pages) and has promised to "bring Grok closer to political neutrality." He stated that the internet (from which it was trained) is filled with various nonsense, and Grok will improve over time.
It's noteworthy that Musk and many of his allies have accused AI chatbots of censoring conservative viewpoints. They believe that such censorship leads AI systems to lie about politically sensitive topics.
With the exposure of "Unhinged Mode," expectations for the Grok chatbot are growing daily.
