Downcodes editor reports: OpenAI is making every effort to upgrade its high-profile video AI model Sora. Sora has attracted global attention since its debut in February this year, but its initial version also exposed some shortcomings. In order to cope with the fierce market competition, OpenAI is working hard to improve Sora's video generation speed, quality and stability, and actively collects massive high-definition video materials to improve the model.
OpenAI is currently working on upgrading its video AI Sora, a project first unveiled in February of this year.
OpenAI is developing an upgraded version of its video AI Sora, which is designed to produce longer, higher-quality video clips faster than the original demo, The Information reports, citing sources close to the company's management team.
It's worth mentioning that Sora wasn't completely ready at launch. The original video reportedly took more than ten minutes to generate, and feedback from some filmmakers was quite negative. For example, producer Patrick Cederberg generated hundreds of videos in order to get a usable clip, but the model did a poor job of keeping frames, objects, and characters consistent. At the same time, the system also suffers from some physics errors, which are not uncommon in other video models.
To improve Sora's performance, OpenAI is collecting millions of hours of high-resolution video footage covering a variety of styles and topics to reduce bias and improve model stability. The purpose of this is to make the generated videos more consistent in style and content, and to better meet the needs of users.
In the meantime, OpenAI also stated that Sora is still in the research phase as they are in discussions with policymakers. Mira Murati, chief technology officer of OpenAI, said in mid-March that Sora will definitely be released this year, but the cost of generating videos still needs to be reduced because the current system costs more than existing AI systems. Much higher.
Competition in the video AI market has become increasingly fierce since Sora was launched in February. In China alone, four new video AI systems have emerged, including Keling, Vidu, Jimeng, and MiniMax’s Conch Video. Run ML also launched two new models and reached a cooperation with Lionsgate.
This series of developments shows that the field of video AI is evolving rapidly, and OpenAI's Sora must speed up its pace to remain invincible in the fierce competition.
Highlight:
OpenAI is upgrading its video AI Sora with the goal of generating longer and higher quality video clips.
⏱️ The initial version took too long to generate and had stylistic and physics bugs.
The video AI market is growing rapidly, and Sora needs to step up to keep up with the competition.
All in all, Sora's upgrade path is full of challenges, but OpenAI's efforts also demonstrate its ambitions in the field of AI video generation. Sora's performance in the future is worth looking forward to, let's wait and see!