Elon Musk's company, xAI, has provided more insight into an upcoming feature for its AI-powered chatbot, Grok, known as "Unhinged Mode." The feature, teased last April, is expected to deliver responses that could be deemed objectionable or offensive, as outlined on an updated FAQ page on xAI's website.
The mode will be designed to generate content that is similar to an amateur stand-up comedian still finding its voice, according to reports.
However, the feature has not yet been activated, with attempts to find it on the Grok web interface proving unsuccessful.Also read: Microsoft rolls back Bing AI image creator update after user complaints about DALL-E 3 performanceThe introduction of Unhinged Mode may align with Musk's original vision for Grok.
When the chatbot was introduced two years ago, Musk described it as edgy, unfiltered, and willing to answer controversial questions that other AI systems avoid.
While Grok has met some of those expectations by producing vulgar content when prompted, it still maintains certain limits. On political subjects, Grok's responses have leaned left, particularly on issues like transgender rights and diversity, according to a report by Tech Crunch. Also read: WhatsApp to bring photo polls and chat event features: Here's what we knowMusk has attributed these tendencies to Grok's training data from public web pages, claiming the content is influenced by “woke” biases.