OpenAI has released a model named Sora that can generate short videos instantly based on text instructions. The research report from China Merchants Securities believes that OpenAI's innovation in the training and input ends has contributed to Sora's success. Currently, some visual artists, designers, and filmmakers (as well as OpenAI employees) have been granted access to Sora, and it is expected to be further pushed to all users as the technology matures. Although OpenAI acknowledges that Sora still has weaknesses, such as difficulty in accurately simulating physical phenomena in complex scenes and understanding specific causal relationships, based on the current demonstrated results, Sora is significantly ahead of other text-based video models, pushing AI video generation into a new era.

#BNB打怪升级之旅