OpenAI has recenetly released a platform called Sora. This platform has expanded exponentially in a short amount of time. When looking at videos made by AI we need to be able to measure a few things. We need to be able to measure how realistic it is which would be the quality of the video, the temporal coherence which is if it flows together or not, does it relate to the promt that it was given and if it remembers its surroundings which is known as object permanence. Unlike DALL-E 3 that makes static photos, Sora can make real time videos of anything that you can imagine and promt it to do. In just a short amount of time it has already learned so much from itself and had so many improvements. Over time it will only get clearer images and better results over time which can lead to us not being able to tell the difference between a video created by AI and a video filmed by an actual camera. This could create many problems, especially right now with the election coming up, people will not know if videos they are seeing about candidates are true or not and could be dangerous to the future of society.
