Elon Musk's artificial intelligence (AI) company, xAI, has announced the beta release of Grok-2, its latest AI assistant. This new iteration boasts an image generation capability reminiscent of OpenAI's DALL-E and Google's Gemini, but with a notable twist: fewer apparent restrictions on image content, the media has reported. The xAI team describes Grok-2 as a "significant step forward" from its predecessor, Grok 1.5. Alongside this major update, the company introduced Grok-2 mini, dubbed "a small but capable sibling."
X, formerly Twitter, subscribers with Premium or Premium+ status can now access the Grok-2 beta, which responds to prompts with both text and visual content.
"Grok-2 is our frontier language model with state-of-the-art reasoning capabilities. This release includes two members of the Grok family: Grok-2 and Grok-2 mini. Both models are now being released to Grok users on the 𝕏 platform," the company wrote in a blog post.
Pushing Boundaries In Image Generation
Early user reports suggest that Grok-2's image generation capabilities extend beyond those of its competitors. The AI appears to handle a wider range of prompts, including those involving political figures -- a domain often restricted by other platforms. Unlike its rivals, DALL-E, Gemini and Midjourney, Grok-2 appears to have far fewer restrictions on the types of images it can generate -- including prompts involving political leaders, says a report by Forbes.
Controversial Images
Several users on X, shared examples of Grok-2's output; they shared images depicting political scenarios, such as former President Donald Trump wielding firearms and US Vice President Kamala Harris in a military context in Gaza. Even a fictional boxing match between presidential candidates has been generated, showcasing the AI's versatility.
Copyright And Brand Representation
Grok-2's apparent lack of restrictions has led to the creation of images featuring copyrighted characters and well-known brands. Users report generating visuals of popular cartoon characters and branded products, raising questions about potential intellectual property concerns. Moreover, in what appears to be a subtle critique of recent AI controversies, Musk shared an image of Grok-2's rendition of George Washington. This move seems to reference the recent debate surrounding Google's Gemini and its historically inaccurate depictions.