OpenAI has released a new version of Dall-e, a text-to-image creation tool. The new Dal-e 3 will produce a more perfect image than before, even with imperfect instructions.
International news agency Reuters reported in a report that OpenAI unveiled Dal-e 3 on Wednesday (September 20). This is the company’s latest version of text-to-image creation.
Users can create images based on text instructions using the artificial intelligence (AI) powered chatbot ChatGPT. Along with that, changes can be made in the picture by giving instructions to the chatbot.
In a statement, the company said, the Dal-e3 can produce detailed and accurate images with minimal instruction.
In addition, security has been strengthened in this new version, according to OpenAI. As a result, it limits the ability to create violent, adult-only and hateful content.
Even if instructed to create a picture in the style of a famous person or a living artist, the tool has the ability to reject that instruction.
Creators can withdraw all or part of the artwork used to train the text-to-image tool, the company said.