Elon Musk’s artificial intelligence startup, xAI, unveiled the beta version of its latest AI assistant, Grok-2, on Wednesday, introducing an image generation tool that rivals OpenAI’s DALL-E and Google’s Gemini.
Unlike its competitors, Grok-2 appears to impose fewer restrictions on the types of images it can create, including those involving political figures and copyrighted brands.
Grok-2, along with its smaller counterpart Grok-2 mini, is available to X users with ‘Premium’ or ‘Premium+’ subscriptions.
The new model builds on its predecessor, Grok 1.5, with improved capabilities in generating both text and images.
Early users of Grok-2 have shared images of prominent political figures, such as former US President Donald Trump wielding handguns and Vice President Kamala Harris in military attire.
Additionally, Grok-2 has been seen generating images of well-known brands and copyrighted characters like the Simpsons and Tom and Jerry, sparking discussions about the model’s lack of restrictions
Musk has positioned Grok as a less regulated alternative to existing AI tools, criticising OpenAI’s ChatGPT and Google’s Gemini for their perceived biases.
The launch of Grok-2 further underscores xAI’s focus on providing a more flexible AI experience for users.