OpenAI unveiled DALL-E 3 image creation artificial intelligence

OpenAI from the third version of the artificial intelligence tool for converting text to image DALL-E It unveiled itself that it now allows users to send their own text commands from ChatGPT use and also includes more security options.

DALL-E’s artificial intelligence can convert users’ text messages into images, but the second version’s performance was problematic, ignoring certain phrases. However, OpenAI researchers have now announced that DALL-E 3 understands user text much better.

One of the new features of DALL-E 3 integration with ChatGPT Is. With this feature, you no longer need to provide detailed descriptions to DALL-E, but instead ask ChatGPT to submit your request to it, and given that DALL-E performs better with longer sentences, naturally This will give you a better output. Of course, you can still use your own commands to create the image.

The new version of DALL-E will first be released to ChatGPT Plus and ChatGPT Enterprise subscribers in October, and then to OpenAI’s research labs and API service in the fall. It is not yet clear when this new tool will be available to the public for free.

Improved DALL-E 3 artificial intelligence security

In addition, OpenAI claims that it has worked on many security features in the DALL-E 3 model to prevent it from creating inappropriate images. DALL-E 3 will also not be able to reproduce images of famous faces, provided that the user’s request does not mention a specific name.

Sandini Agarwal, a policy researcher at OpenAI, said she has “a lot of confidence” in the security features of the new model, but explained that the model is constantly being improved and is not yet perfect. Also, representatives of OpenAI have announced that DALL-E 3 has been trained in such a way that it does not imitate the style of other artists.