One year ago, Elon Musk, tech magnate and founder of X, hinted at a unique feature for his AI chatbot, Grok, known as the ‘Unhinged Mode’. Now, new information has surfaced on xAI’s FAQ page, the Musk-operated creator of Grok, helping to demystify this impending feature.
‘Unhinged Mode’ is set to offer responses that are potentially contentious, inappropriate, and even offensive, resembling the risky humour of a novice comedian.
Despite multiple attempts, the function is yet to be discovered in Grok’s user interface on X, suggesting it’s not operational yet.
Grok was launched by Musk around two years ago as a frank and eyebrow-raising chatbot, poised to navigate the edgier landscapes of conversation often avoided by other AI systems. Grok has proven to be capable of borderline explicit language unlike its counterparts, such as ChatGPT.
However, Grok currently draws a line at fiery political topics, demonstrating a left-leaning bias on issues like transgender rights, diversity protocols, and inequality. Musk has pointed to the prevalence of ‘woke nonsense’ in the training data (public webpages) as the cause and has committed to steering Grok towards a more politically balanced stance in the future.
Accusations of AI chatbots suppressing conservative viewpoints have been rampant, with figures like President-elect Trump’s AI and Crypto expert, David Sacks, criticising OpenAI’s ChatGPT for being deliberately ‘woke’ and deceitful about politically sensitive matters.
Stay updated with more AI news – subscribe to TechCrunch’s dedicated AI newsletter today.
Original source: Read the full article on TechCrunch
