BIG BUG: DALL-E does not know how to write the text that I gave him to put in the image

DALL-E does not know how to write the text that I gave him to put in the image

I gave him the exact words to write. How is it possible to change the words and write them so INCORRECTLY? Maybe he doesn’t know how to recognize the Romanian language when it comes to images?

This is the prompt:

Vreau o imagine care sa contina un anumit mesaj, 3 puncte unul sub altul. Poti sa reformulezi mai exact, daca ti se pare mai bine. Poti adauga diacritice romanesti

Titlu: Somnul “BEBE-F”

1. Nu manca nimic dupa ora 17:00
2. Culca-te maxim la ora 23:00
3. Cu jumatate de ora inainte de culcare ia o pastila “spirulina + catina” de la HOFIGAL

Please check the DAL-E image below. Notice how incorrectly he spelled my words. He just had to copy their words and put them there. Instead, DALL-E has completely mixed up the letters, and the writing makes no sense.

That’s a big “effect of how it works”

yes, but what is the solution? ChatGPT has 4-5 models for creating images, and none of them know how to do this well.

What is so difficult to create the picture according to the instructions, and then to copy there exactly the words that I asked for?

So, in practice, I cannot use ChatGPT to insert words into images. This simple thing, it doesn’t know how to do CHATGPT…

He doesn’t even know how to solve 2nd grade problems…

ChatGPT should be easy to use by anyone. And this simple thing, to create image then to copy/paste any text on it, should be the most easy and the most important. otherwise, it would practically mean that the user has to edit the image again, and lose precious time (especially if he doesn’t know how)