China launches text-to-video model before OpenAI

MiniMax – known as one of China’s AI “tigers” along with Zhipu AI, Baichuan and Moonshot AI – has rolled out the -01 video model to the public via its website, after revealing the new tool at a developer conference in Shanghai last Saturday.

Photo: GI

Video-01 allows users to enter text descriptions to create videos up to 6 seconds long. The process from text prompt to video creation takes about 2 minutes.

MiniMax founder and CEO Yan Junjie said that video-01 is the first version of the company's video creation tool. He pointed out that future updates will allow users to create videos from images and edit those videos.

Video-01 from MiniMax reflects how Chinese tech companies have been making aggressive moves into this emerging AI market segment, after OpenAI first released video samples generated by Sora in February. However, Sora has yet to be widely released to the public.

X

A video created by Sora with a caption that reads "The camera follows a white vintage SUV with a black roof rack as it speeds down a steep dirt road...".

According to OpenAI’s announcement, they will take a number of important safety steps before implementing Sora into OpenAI products. OpenAI is working with experts in areas like misinformation, hate speech, and bias to avoid the risks this model could create.

OpenAI will also give access to a number of visual artists, designers, and filmmakers to get feedback on how to improve the model to be most useful to creative professionals.

“Sora can generate complex scenes with multiple characters, specific types of motion, and precise details about objects and context. The model not only understands what the user asks for in the prompt, but also how those things exist in the physical world ,” OpenAI said in a statement.

Bui Huy (according to SCMP, OpenAI)

Source: https://www.congluan.vn/cong-ty-trung-quoc-vuot-mat-openai-khi-cho-ra-mat-mo-hinh-bien-van-ban-thanh-video-post310490.html