1. OpenAI’s Sora AI model can create impressive videos from text prompts.
2. The company plans to make Sora available to the public this year.
3. Some of Sora’s latest videos include a leafy elephant, a kitten pirate on a vacuum, and a rainbow waterfall.
OpenAI’s artificial intelligence video model, Sora, continues to impress with its ability to create realistic videos from text prompts. While the public will have access to Sora at some point this year, for now, the only way to see its capabilities is through Sora Sunday posts on social media. However, there is growing pressure on OpenAI to disclose the source of the data used to train Sora, as the company has been hesitant to confirm if public videos from YouTube creators were included in the dataset.
Some recent remarkable creations from Sora include a video of a green, leaf-covered elephant for St. Patrick’s Day, a rainbow waterfall, and a kitten riding a robot vacuum. Sora’s ability to merge different videos and follow simple prompts allows for the creation of unique and realistic content. The company emphasizes that they are sharing their research progress early to receive feedback and provide the public with a glimpse of the future capabilities of AI.
Sora’s ability to bring consistency of movement to its creations, such as the independent movement of leaves on the elephant, showcases its impressive capabilities. Additionally, Sora excels at realism, as seen in the video of a pirate kitten on a robot vacuum. The model’s potential for merging videos opens up opportunities for creating fun and engaging social content. Other creations include a rainbow waterfall, a chameleon showing its color-changing abilities, and a bee in flight, all of which demonstrate Sora’s versatility and potential for future applications in various industries, including wildlife documentaries.