Elon Musk's company xAI has updated the FAQ page for its AI-powered chatbot, Grok, revealing a new "Unhinged Mode" that aims to provide responses that are objectionable, inappropriate, and offensive, akin to an amateur stand-up comic. This mode is not yet live, and attempts to find it in the Grok web interface have been unsuccessful. Musk initially introduced Grok as an edgy, unfiltered chatbot willing to tackle controversial questions that other AI systems avoid. While Grok can be vulgar, it currently avoids certain political topics and leans left on issues like transgender rights and diversity, which Musk attributes to its training data from public webpages. He has promised to make Grok more politically neutral.
Musk and his allies, including David Sacks, President-elect Donald Trump’s crypto and AI advisor, have criticized AI chatbots for allegedly censoring conservative viewpoints. Sacks has specifically targeted OpenAI's ChatGPT, accusing it of being "programmed to be woke" and dishonest about politically sensitive topics. Musk has expressed his intention to improve Grok, which he describes as still in its beta phase, and address the perceived bias in its responses.
Key takeaways:
```html
Elon Musk's company xAI has teased an "Unhinged Mode" for its AI chatbot Grok, intended to provide objectionable and offensive responses.
Grok was initially pitched as an edgy and unfiltered chatbot, but currently hedges on political subjects and leans left on certain issues.
Musk attributes Grok's current behavior to its training data and has pledged to make it more politically neutral.
Critics, including David Sacks, accuse AI chatbots like Grok and ChatGPT of censoring conservative viewpoints and being biased.
