Veo was trained on a large amount of footage, some of which may have been sourced from Google's own YouTube. Despite concerns about the use of user data for training AI models and the lack of a mechanism for creators to remove their work from training data sets, Google insists it is setting the "gold standard" in ethics. The company has already made Veo available to select creators and plans to gradually release it to a wider audience. However, due to limitations in the current generative AI, Veo will remain on a waitlist on Google Labs for the foreseeable future.
Key takeaways:
- Google has unveiled Veo, an AI model that can create 1080p video clips around a minute long given a text prompt, at its I/O 2024 developer conference.
- Veo can capture different visual and cinematic styles, and make edits and adjustments to already-generated footage. It is seen as competitive with today's leading video generation models.
- The AI model was trained on a large amount of footage, some of which may have been sourced from Google's own YouTube platform.
- Despite its capabilities, Veo has limitations such as objects in its videos disappearing and reappearing without explanation, and incorrect physics. Therefore, it will remain behind a waitlist on Google Labs for the foreseeable future.