ByteDance and the University of Hong Kong have introduced Goku and Goku+, AI models designed to generate both images and videos. These models aim to produce high-quality visuals, with applications in content creation, advertising, and marketing. The release includes demos showing detailed animations, cinematic shots, and lifelike video scenes.
While details about their accessibility and broader use cases are still emerging, the technology hints at a growing shift in how AI can assist in producing realistic video content efficiently. They show some of their versions of content that Sora previously produced – to my eye, the Sora ones look considerably better. You can explore more about these models here.
