The recent public release of the Hunyuan Video generative AI model has intensified ongoing discussions about the potential of large multimodal vision-language models to one day create entire movies.