OpenAI, a leading artificial intelligence (AI) company, has just introduced Sora, a text-to-video application that can transform text into videos with realistic or imaginative scenes.
Sora is a Generative AI application, a technology that can generate new content based on existing data. Sora is an advanced development from Dall-E, a text-to-image application also created by OpenAI.
Examples of Sora’s Text-to-Video Results
With Sora, users can create short videos just by using text. For example, users can type “A snowy Tokyo city with several people enjoying the beautiful weather, and there are sakura flowers fluttering.” Then, Sora will produce a video that matches that description.
The videos produced by Sora are of very high quality, with very detailed details, complex camera movements, and characters with lively emotions. Sora can also create videos with a duration of up to 60 seconds, longer than other text-to-video applications like Lumiere from Google and Make-A-Video from Meta.
Sora Can Impact Politics, Economy, and Entertainment
However, Sora also raises controversy about the ethics of Generative AI. On one hand, Sora makes it easy for users to create creative videos without needing to have special skills. On the other hand, Sora also has the potential to disrupt the creative industry, human jobs, and truth of information.
One industry threatened by Sora is the film footage industry, which is short videos usually used for film production, advertising, or documentary purposes. With the existence of Sora, the need for film footage can be replaced by AI-generated videos.
In addition, Sora can also be used to create hoax videos, which are fake videos made with the intention of deceiving or damaging someone’s reputation. Hoax videos can be very dangerous, especially amid heated social and political conditions around the world.
To prevent misuse of Sora, OpenAI has taken several steps.
- First, Sora has not been opened to the public, only to selected designers, filmmakers, and visual artists to get input.
- Second, Sora is still being tested by OpenAI’s Red Team, a team tasked with investigating the risks and potential misuse of Sora before it is released to the public.
- Third, OpenAI adds a watermark to the metadata and images indicating that the video is the result of AI. OpenAI uses the watermark standard from the Coalition for Content Provenance and Authenticity (C2PA), which is also adopted by Meta to label AI-generated content on their social media platforms like Facebook and Instagram.
With this watermark, it is hoped that users can more easily distinguish between original videos and AI-generated videos. This is important to reduce the occurrence of misinformation, which is wrong or misleading information that can harm many parties.
Sora has also attracted attention from other technology companies, such as Apple. According to circulating news, Apple will reveal AI features in iOS 18 related to Sora at the WWDC event to be held in June. Apple is also rumored to be enhancing the Neural Engine, a special AI processing engine, in the Apple A18 chip that will be used by the latest generation iPhone.
Sora is one example of the rapid advancement of AI technology, but it is also full of challenges. What is your opinion about Sora? Are you interested in trying it? Or are you worried about its impact?”