The Alibaba team used a dataset created by the University of Minnesota, which includes videos from various TikTok creators. This practice of using datasets created for academic research for commercial purposes has been common, but it's becoming increasingly controversial. Legal battles have emerged where artists and content creators are suing AI companies for using their work without permission. This has significant implications for AI models like "Animate Anyone", and raises questions about the ethical considerations of using real people's content in large datasets.
Key takeaways:
- Researchers from Alibaba have developed a new AI model called "Animate Anyone" that can replicate dance sequences from TikTok videos, raising concerns about the potential misuse of the technology.
- The model has been trained on a dataset that includes videos from famous TikTok creators, which were used without their consent, leading to accusations of content theft.
- There are growing legal challenges against AI companies for using artists' works without permission, with a recent lawsuit against Epic Games for using a choreographer's dance moves being a notable example.
- The article suggests that academics should consider the potential future ramifications when creating large datasets of real people's content, as these can be used by large companies for commercial purposes.