YouTube Shorts is integrated with Veo 2, Google Deepmind’s latest video model, allowing creators to generate AI video clips for posts.
VEO 2 is Google’s response to Sora, Openai’s text-to-video generator. Google-owned YouTube already allowed creators to generate AI backgrounds for shorts using text prompts via a feature called Dream Screen. However, VEO 2 allows creators to generate standalone video clips and add them to one of their shorts.
According to a blog post by the YouTube Director of Product Dina Berrada, these AI tools run faster than before.
“VEO 2 gives a better understanding of real-world physics and human movements, making its output more detailed and realistic,” writes Berrada. “You can also specify style, lens and film effects to make your dream screen an easy and fun way.”
To show that these videos were created using AI, we use DeepMind’s SynthID tool, which lets AI watermark and identify the content generated. YouTube also says that these videos are clearly marked as being made with AI, but these labels do not affect the potential harm of sharing synthetic content that could mislead viewers. It doesn’t completely reduce.
To use VEO 2 with YouTube shorts, creators can open the shorts camera, select the green screen, then go to the dream screen and enter a text prompt to generate the video.
This feature will launch on Thursday for all creators in the US, Canada, Australia or New Zealand. YouTube will be expanding its access later.