Read in English / 阅读语言 
ไทย 
简体中文
AI Creates Realistic Videos: Sora
AI-powered video generation is here! OpenAI, the creators of ChatGPT, have unveiled a new AI model named “Sora,” which means “sky” in Japanese. Sora can generate highly realistic animated videos with a maximum length of one minute.
Users can input text descriptions (prompts), static images, or even existing videos to extend them further. Check out some examples of what this new AI can do in terms of video quality.
The prompt for this clip was: A stylish woman walks down a Tokyo street filled with warm glowing neon and animated city signage. She wears a black leather jacket, a long red dress, and black boots, and carries a black purse. She wears sunglasses and red lipstick. She walks confidently and casually. The street is damp and reflective, creating a mirror effect of the colorful lights. Many pedestrians walk about.
For this clip, the prompt was: Photorealistic closeup video of two pirate ships battling each other as they sail inside a cup of coffee.
For this one, the prompt was: Tour of an art gallery with many beautiful works of art in different styles.
Limitations of AI Sora
From these examples, it’s clear that Sora captures detail impressively, especially compared to other services currently available on the market. However, it still has several limitations. According to OpenAI, the system cannot yet simulate complex physical movements or fully understand cause and effect. For instance, in a video of someone biting a cookie, the cookie might not show bite marks. Additionally, Sora sometimes struggles with spatial orientation, such as determining left from right.
Using Sora
If you’re eager to try out AI-generated realistic videos, you’ll need to be patient. Currently, OpenAI is allowing a small group of artists, developers, and researchers to test the system and provide feedback before it’s released to the public.
Although text-to-video AI isn’t entirely new, OpenAI’s Sora stands out as an exciting generative video tool. In the near future, it may become difficult to distinguish between videos filmed by humans and those generated by AI. Platform providers, like Facebook and Instagram, are already working on solutions to label AI-generated content, as evidenced by their recent policy announcement.
How much of an impact AI like Sora will have on the art world remains to be seen, but it’s a space to watch closely.
