From the course: AI Trends

OpenAI Sora: Text-to-video

From the course: AI Trends

Start my 1-month free trial

OpenAI Sora: Text-to-video

“

- Imagine you could describe a video and have it seemingly appear out of thin air. Well, OpenAI's Sora allows you to do just that. Sora is a text-to-video diffusion model. It receives a prompt, starts off with a noisy sequence, then iteratively removes noise until it has a clean and crisp video. Sora leans heavily on the transformer architecture, a model architecture that has yielded incredible technologies, such as ChatGPT and DALL-E. While this is extremely exciting, there is some room for caution as this technology could be used to generate misinformation. Now, at the time of watching this, you may not yet have access to Sora or a video generation tool. There are some ways, however, you can get ready for this technology. For starters, you can work on your prompt engineering skills using a text-to-image generation tool. These skills will likely carry over to something like Sora. Another thing you can do is get familiar with the vocabulary of cinematography. Find out the names of different camera angles. Figure out which cameras and lenses were used to shoot your favorite film. As someone who creates video and tells stories through video, I'm extremely excited for this technology and cannot wait to see what it will bring about.

Contents