Create Videos from text. SORA the new AI
2 Minutes of reading
Can you imagine being able to create realistic and imaginative videos just by writing a few lines of text?
This is the promise of SORA, the new generative artificial intelligence model developed by OpenAI, the organization dedicated to AI research and development.
SORA is capable of generating 60-second video scenes while maintaining visual quality and consistency with the input text. In addition, it can incorporate elements such as camera movements, characters with emotions, light and sound effects, and different artistic styles.
How does SORA work?
SORA is based on a deep neural network architecture that combines several image and video generation techniques, such as GPT-3 or DALL-E.
These techniques allow SORA to understand the meaning of text, search for relevant images on the web, and synthesize videos from them.
Video created with SORA from text on the official Open AI website
The SORA generation process is divided into three main concepts:
Text analysis: SORA analyzes the input text and divides it into segments, each with a description of a scene or event. It also extracts keywords and named entities from the text, and uses them to search for related images on the web.
Image generation: SORA uses DALL-E, a model that can create images from text, to generate an image for each text segment. These images are combined with images found on the web, and the most appropriate ones are selected for each scene.
Video generation: SORA uses DALL-E-Vision, a model that can create videos from images, to generate a video for each selected image. These videos are concatenated to form the final video, to which transition effects, sound and styling are added.
What can be done with SORA?
SORA has great potential for creative and educational applications, such as the creation of movies, animations, video games, documentaries, tutorials, or presentations. SORA can also be a useful tool for journalists, researchers, or students, who want to illustrate their texts with videos.
Some examples of videos generated by SORA are the following:
A trailer for a sci-fi movie about a spaceman wearing a red wool motorcycle helmet, set in a salt desert with a blue sky, cinematic style, shot in 35mm, vivid colors
.
An animated scene showing a close-up of a short, hairy monster kneeling next to a melted red candle. The art style is 3D and realistic, with a focus on lighting and texture. The mood of the painting is one of wonder and curiosity, as the monster gazes at the flame with wide eyes and open mouth. His pose and expression convey a sense of innocence and playfulness, as if he is exploring the world around him for the first time. The use of warm colors and dramatic lighting further enhances the cozy atmosphere of the image.
A drone view of the waves crashing against the sheer cliffs along Garay Point beach in Big Sur. The crashing blue waters create white-tipped waves, while the golden light of the setting sun illuminates the rocky coastline. A small island with a lighthouse stands in the distance, and green bush covers the cliff edge. The steep drop from the road to the beach is a dramatic feat, with the cliff edges jutting out over the sea.
What are the challenges posed by SORA?
SORA is an impressive tool, but it also poses some ethical and technical challenges. On the one hand, SORA could be used for malicious purposes, such as creating fake or misleading videos, which could affect the credibility of information or people's reputation. On the other hand, SORA still has limitations in terms of the quality, variety, and originality of the videos it generates, which are highly dependent on the quality and quantity of the input data and web images.
For these reasons, OpenAI has decided not to publish the code and model of SORA, but only an interactive demonstration on its website. In addition, OpenAI has established a set of rules and principles for the responsible use of SORA, including respect for copyright, privacy, security, and diversity.
Conclusion
SORA is one of OpenAI's latest innovations, demonstrating the power and versatility of generative artificial intelligence. SORA is capable of creating realistic and imaginative videos from text, opening up new possibilities for expression and learning. However, SORA also raises some ethical and technical and technical challenges, which require appropriate regulation and oversight.
SORA is, without a doubt, a fascinating tool that invites us to explore the limits of creativity and reality.
If you want to know more about this tool, you can visit the OpenAI website, where you will find more information and examples.
