xAI's Grok Imagine Launches with NSFW Content Generation
xAI, Elon Musk's artificial intelligence company, has officially launched Grok Imagine, its new image and video generator, to all SuperGrok and Premium+ X subscribers via its iOS app. In line with Musk's positioning of Grok as an unfiltered, boundary-pushing AI, the generator notably permits users to create Not Safe For Work (NSFW) content.
Grok Imagine is designed to transform text or image prompts into 15-second videos, complete with native audio. A distinctive "spicy mode" allows for the generation of sexually explicit content, including partial female nudity. While there are limits to the explicitness—many highly suggestive prompts reportedly result in blurred and inaccessible "moderated" images—TechCrunch was able to generate semi-nude imagery during testing.
The inclusion of NSFW capabilities is consistent with xAI's recent releases, such as a "raunchy, hyper-sexualized anime AI companion" launched last month. However, this unrestrained nature also raises concerns about potential unintended consequences, drawing parallels to previous instances where Grok's lack of moderation led to the generation of hateful, antisemitic, and misogynistic content.
The existence of Grok Imagine's "spicy mode" was first reported by CNBC last week, following a now-deleted post on X by xAI employee Mati Roy, who stated that Grok Imagine videos "can do nudity." TechCrunch has reached out to xAI for further information.
Despite the explicit content allowance, some limitations are apparent. The model permits the creation of content featuring celebrities, from Donald Trump to Taylor Swift, but with apparent restrictions. For example, attempts to generate an image of a pregnant Donald Trump were unsuccessful, instead producing images of Trump holding a baby or alongside a pregnant woman.
In its early stages, Grok Imagine's output of human subjects sometimes falls into the "uncanny valley," characterized by waxy-looking skin and occasional cartoonish appearances. Despite this, the platform aims to compete with established players in the generative AI space, including Google DeepMind, OpenAI, Runway, and various Chinese developers.
The generator offers an impressive user experience, producing images rapidly from text prompts and continuously generating new visuals as users scroll. These images can then be animated into stylized videos, all within a seamless and intuitive user interface. Elon Musk has stated on X that the model is expected to "get better every day."