Alibaba, the Chinese e-commerce giant, has developed an AI video generator called EMO that is capable of transforming still images of faces into actors and singers. In comparison to OpenAI's Sora, Alibaba's EMO stands out for its ability to make AI-created characters speak and even sing. Demonstrating this capability, Alibaba showcased videos on GitHub, featuring the Sora lady singing "Don't Start Now" by Dua Lipa and Audrey Hepburn speaking audio from a viral clip. EMO's facial animations appear to be more realistic than previous AI face-swapping technologies such as deepfakes.

While the demo is impressive, further testing is needed to assess its capability in expressing extreme emotions solely through audio guidance. Additionally, EMO showcases multilingual capabilities and effectively captures subtle emotional cues in facial expressions. The model behind EMO utilizes a large dataset of audio and video to accurately depict emotions without relying on 3D models. Although the demos are promising, the implications for the acting industry are potentially concerning.