Elon Musk and xAI introduced Monday that Grok customers will quickly have the ability to generate synthetic intelligence movies full with sound on its app.
The brand new function is known as Think about, and it permits customers to create each photographs and movies based mostly on textual content prompts and switch static photographs into video. At the moment, entry seems to be restricted to some workers and chosen influencers, and the @Grok X account wrote Monday night: “Video era is coming to Grok by way of our Think about function, powered by Aurora. Create prompt movies with sound from textual content prompts. Obtain the standalone Grok app, subscribe to SuperGrok, and be a part of the waitlist for early entry in October.”
SuperGrok is a paid tier that prices $30 monthly.
On Thursday, Musk teased “bringing again Vine, however in AI kind,” referring to the favored social media platform that Twitter shut down in 2017. An xAI worker stated in a reply that Grok Think about movies may be as much as 6 seconds, the size Vine movies had been famously recognized for.
Based mostly on movies shared on-line by a number of xAI workers and customers with early entry to Think about, the brand new function works by having customers kind in descriptions of what they wish to have of their movies, and the app generates a number of movies that correspond to the descriptions that customers can select from.
xAI workers and Grok lovers shared movies on-line showcasing the brand new function. An xAI worker posted an AI-generated video of a cat purring whereas strolling via house, and one other posted an astronaut strolling via an alien panorama.
An xAI worker stated on X that the brand new function guarantees some lewd prospects.
Mati Roy wrote in a submit sharing a video of a virtually nude lady, “Grok Think about movies have a spicy mode that may do nudity.” He stated in a reply to a different submit that the brand new video era function would have the ability to create real looking movies of people.
A couple of customers within the feedback expressed concern that the brand new options may permit customers to create specific deepfakes. An xAI spokesperson didn’t instantly reply to a request for remark.
The Trump administration has been pushing to crack down on the creation of specific deepfakes; President Donald Trump signed the Take It Down Act into regulation in Could. The act, initially written by Sen. Ted Cruz, R-Texas, makes it unlawful to publish nonconsensual sexually specific photographs and movies, whether or not actual or computer-generated.
New AI options on Grok had already drawn concern from some advocates due to the app’s vulgar and sexually specific “companions” — AI characters customers can speak to via textual content and speech. Musk had additionally introduced {that a} new companion he had been teasing named “Valentine” was now obtainable on Grok.
Grok stoked controversy this month when it revealed a slew of antisemitic posts. The official @Grok account later stated in a press release on X that it was “actively working to take away the inappropriate posts.” Most of the posts, together with ones during which the chatbot would consult with itself as “MechaHitler,” had been taken down.
Many AI platforms have already made AI video options obtainable, together with OpenAI, which launched Sora in February, and Google, which launched Veo 3 this month.
Musk responded to a person who in contrast the standard of Grok’s video era to Gemini’s, saying, “Grok Think about remains to be early beta and is optimized for optimum enjoyable, so needs to be evaluated as ‘quickest time to make a enjoyable, shareable video’, slightly than visible/auditory perfection. Our heavy responsibility video mannequin will practice on the 110k GB200s coming on-line subsequent month.” xAI is including 110,000 graphics processing items, specialised circuits designed to assist generate laptop graphics, to its supercomputer Collosus in Memphis, Tennessee, in line with Musk.
