Last April, Company X's billionaire owner Elon Musk teased the mysterious “unhinged mode” of Company X's AI-powered chatbot Grok. Almost a year later, xAI, the Musk-owned company behind Grok, has updated its FAQ page on its website to reveal the new mode.
According to the FAQ page, Grok in Unhinged mode will provide answers that are “intended to be offensive, inappropriate, or offensive,” and is “much like an amateur stand-up comic who is still learning his craft.” ”.
This mode doesn't seem to be available yet. TechCrunch tried to find it in X's Grok web interface to no avail.
Unhinged Mode may be Musk's attempt to realize Grok's original vision.
When Musk announced Grok nearly two years ago, he pitched the chatbot as edgy, unfiltered, and anti-woke. They are generally willing to answer controversial questions that other AI systems cannot answer. He fulfilled part of that promise. For example, if you're called vulgar, Grok will gladly go along with it, spouting profanity and colorful profanity that you won't hear on ChatGPT.
However, Grok as it exists today avoids political themes and never crosses certain boundaries. In fact, one study found that Grok leans to the political left on topics such as transgender rights, diversity programs, and inequality.
Musk blamed the behavior on Grok's training data (a public webpage) and promised to “move Grok closer to political neutrality.”
“Unfortunately, the internet (which the internet is trained on) is full of woke nonsense,” he said in a December 2023 post. “Grok is going to get better. This is just a beta.”
Musk and many of his allies, including President-elect Donald Trump's crypto and AI “czar” David Sachs, have accused AI chatbots of censoring conservative views. There is. Sachs specifically named OpenAI's ChatGPT as being “programmed to be woke” and false about politically sensitive subjects.