Want to get in on the fun of generating images with AI, but your language level isn't high enough yet?
No problem! In this tutorial, we'll show you how to use an existing image to get a text and study vocab and grammar, and then put that back into an image 🤩 Can't wait? Let's dive right in!

Exercises

1) Open ChatGPT.

2) Try this prompt: Here is an image. I am a [LEVEL] [LANGUAGE] learner. I would like you to help me extract vocabulary that I can use to describe this image in my target language. Please generate a list of important words and sentences that describe the image. Add a note to any nouns to help me learn the gender of the words.

3) If you would like to hear it read out loud, we recommend copying the sentences and pasting them into Natural Reader (learn more in our tutorial: Tool Tip: Natural Reader). 

Prompt: Give me all the descriptive sentences in a paragraph. Answer in [TARGET LANGUAGE].

4) Now translate this sentence into your target language: “Generate an image with the following description.” Then paste it into the chat, followed by the paragraph of all descriptive sentences. The result will be an image.

5) Finally, think of one small detail you would like to change and tell ChatGPT to change it in your target language.

For example: “Changez le manteau de bleu à rouge,” (change the coat from blue to red) or “Bringe Schnee ins Bild" (add snow to the image).

6) Try to make the image look even closer to the original you uploaded!

🎤 Extra challenge: Speak those prompts to ChatPGT in your target language!