“TikTok may soon offer a tool for generating AI-created videos. ByteDance is launching with a tool that allows the creation of videos and images from prompts.”
Following in OpenAI’s Footsteps
In February 2024, OpenAI, the company behind ChatGPT and the image generation tool Dall-E, introduced its video generation model Sora. This tool has sparked interest, prompting other companies, like ByteDance, to position themselves in the field.
TikTok’s parent company aims to meet the growing demand for AI-generated video content in China with its tool, Jimeng AI. The application was launched on Android on July 31 and is now also available on the App Store, but only in China.
For now, it’s known that users will be able to create videos from text, although specifics regarding video length and resolution are still unknown. However, it’s likely that these videos will be optimized for social media platforms like TikTok.
Image and Video
Jimeng AI has a significant advantage in the content generation field, as unlike OpenAI and Sora, Jimeng is capable of generating both images and videos.
The application allows users to create up to 2,050 AI-generated images or 168 videos per month, depending on the subscription plan chosen. The plans range from €8 for a monthly subscription to €84 for an annual subscription (these prices are based on a conversion from yuan to euros without taking into account potential taxes).
For now, there is no information available regarding the database used to generate these videos. The close connection with TikTok raises questions, and one might wonder whether the videos produced on the social network could be used to feed this AI.
“A practice we have already seen implemented by other groups like Meta. Last May, the company wanted to use our posts on Facebook and Instagram to enhance the learning of its artificial intelligence.”