OpenAI’s artificial intelligence video model Sora continues to display impressive ability to capture the real world from nothing more than a text prompt.
The company says it will be available to the public at some point this year, but until then the only way to experience its capabilities is through the Sora Sunday posts on social media.
The latest release comes amid growing pressure on OpenAI to disclose the source of the data used to train Sora. This was sparked by CTO Mira Murati’s refusal to confirm during a Wall Street Journal interview whether public videos from YouTube creators were in the dataset.
In the meantime we still get to enjoy its creations, including an uncanny ability to merge different videos, follow simple prompts and have multiple scenes from a single text prompt.
Amazing new Sora clips
This week we’ve seen a video of a green, leaf covered elephant to mark St Patrick’s Day, a rainbow waterfall and the waltz of the bumble bee.
OpenAI adds to every video it shares: “We’re sharing our research progress early to learn from feedback and give the public a sense of what AI capabilities are on the horizon.”
A leafy elephant
One of the most impressive features of AI video tools like Sora is the ability to create something that couldn't exist — and from a text prompt.
Unlike other models Sora brings consistency of movement. For example in this elephant video shared to Instagram the leaves on this elephant move independently.
Prompt: "An elephant made of leaves running in the jungle."
Kitten on a vacuum
Sora also does realism very well. This is a video of a pirate kitten riding on a robot vacuum and I had to do a double take to realise it wasn't just a video filmed on a smartphone.
One of the fun features of Sora is its ability to merge a pair of videos together and a future tool built on the model could see someone merge a video of their own pet with another video like that of a robot vacuum to create a new generation of fun social clips to share.
Prompt: "An adorable kitten pirate riding a robot vacuum around the house." Shared to Instagram by OpenAI.
Rainbow waterfall
More paint bucket than rainbow, but this shows a bizarre version of Niagra Falls looking like it has been taken over for a paint company commercial.
Shared to Instagram and made using the prompt "niagara falls with colorful paint instead of water."
Chameleon moving
This image was shared to the OpenAI Instagram showing a chameleon moving and showing its color changing capabilities. It is impressive how realistic it looks and could hint at the future of wildlife documentaries.
It was shared with the description: “this close-up shot of a chameleon showcases its striking color changing capabilities. the background is blurred, drawing attention to the animal’s striking appearance.”
Bee in flight
This is one of my favorites and makes me root for the little bee as it flies around hunting for pollen across a range of plants and backgrounds.
OpenAI shared it to Instagram with the prompt: “POV video of a bee as it dives through a beautiful field of flowers”.
