From the course: Introduction to Multimodal Prompting for Generative AI
Unlock the full course today
Join today to access over 23,200 courses taught by industry experts.
Text to image in GPT-4
From the course: Introduction to Multimodal Prompting for Generative AI
Text to image in GPT-4
- [Narrator] A fun way to start using ChatGPT's multimodal capabilities is using the ChatGPT app. You can take photos directly into your phone. Here, I take a photo of this hourglass, and I ask ChatGPT to generate an icon for an AI powered time management app. The app is geared towards tech consultants. Generally speaking, the more details you give, the better, and this can be challenging. We tend to rely on the visual input, but it's very important to be specific when it comes to the text prompt. So here we have this icon that was generated for us, and I'm going to go ahead and click on it, and I'll go ahead and click Select. I'll erase some of the image. You can also modify the size of this selector, here on the top right. And I'll add some text, so "An hourglass with a flat base on the bottom and on the top in an app icon." There it is. So initially we had an image and a text input that created the prompt that generated this image. Then we went ahead and used an image as well as a…