Grok could soon have an 'Unleashed Mode'
Grok, the artificial intelligence powering various features on X, could soon implement an 'Unhinged Mode' that would allow it to use offensive language in its responses.
In April of last year, Elon Musk, the billionaire owner of X, mentioned an enigmatic "Unleashed Mode" for Grok, the AI-powered chatbot. Almost a year later, xAI, the company Musk founded responsible for Grok, has updated the FAQ page on its website, shedding light on this new mode. According to the available information, Grok in Unleashed Mode will provide responses “meant to be objectionable, inappropriate, and offensive,” similar to a novice comedian still honing their craft. This mode does not appear to be active yet, as TechCrunch was unable to find it in the web interface of Grok on X.
This Unleashed Mode could be an attempt by Musk to fulfill the original vision for Grok. When he introduced Grok about two years ago, he described it as a provocative, filter-free chatbot that opposes “politically correct” culture and is willing to tackle controversial questions that other AI systems avoid. To some extent, he has lived up to this promise; when prompted to be vulgar, Grok has no hesitation in using coarse language and colorful expressions that would not be found in ChatGPT. However, the current version of Grok is cautious on political topics and avoids crossing certain lines. In fact, a study revealed that Grok tends to lean left on issues such as trans rights, diversity programs, and inequality.
Musk has attributed this behavior to the data used to train Grok, which comes from public web pages, and has promised to “bring Grok closer to a politically neutral position.” In a post from December 2023, he mentioned, “Unfortunately, the Internet (with which it is trained) is rife with politically correct nonsense. Grok will improve. This is just the beta version.” Musk and several of his allies, including David Sacks, the "czar" of crypto and AI for elected President Donald Trump, have accused AI chatbots of censoring conservative opinions. Sacks has specifically pointed to OpenAI's ChatGPT as “programmed to be politically correct” and misleading on politically sensitive topics.
