DALL-E 3: OpenAI announced the new version of its text-to-image tool

OpenAI recently revealed an improved version of DALL-E, its text-to-image tool, which combines ChatGPT, OpenAI's popular AI chatbot, to lessen the burden of prompting.

The majority of modern AI-powered picture production technologies accept cues, such as descriptions of photographs, and create artwork from them in a variety of styles, from the photorealistic to the fanciful. But creating the ideal prompt can be difficult, to the point where "prompt engineering" is emerging as a legitimate profession.

DALL-E 3, a new tool from OpenAI, leverages ChatGPT to assist in filling out prompts. Through ChatGPT, customers of OpenAI's top-tier ChatGPT plans, ChatGPT Plus and ChatGPT Enterprise, can submit an image request and refine it through discussions with the chatbot, obtaining the results right in the chat app.

A prompt as little as a few words can be improved by ChatGPT to make it more detailed and give the DALL-E 3 model additional direction.

DALL-E 3 introduces more than only ChatGPT integration. According to OpenAI, DALL-E 3 also produces images of higher quality and more accuracy when responding to longer instructions. Additionally, it handles material like text and human hands better than previous iterations of image-generating models.

Additionally, according to OpenAI, DALL-E 3 features new techniques to lessen algorithmic bias and increase safety. For instance, DALL-E 3 would decline requests for images that are meant to represent public persons or be created in the style of contemporary artists. Additionally, artists can now choose not to have any or all of their works used to train OpenAI's text-to-image models in the future. (OpenAI and some of its competitors are being sued for allegedly training their generative AI image models using artists' copyrighted work.)

The release of DALL-E 3 coincides with the intensifying competition in generative AI, particularly in the area of picture synthesis. OpenAI is under pressure to keep up with competitors like Midjourney and Stability AI that are constantly improving their image-generating models.

In October, OpenAI intends to make DALL-E 3 available to premium ChatGPT subscribers, then to research facilities and its API users. When and when it will issue a free web tool, like it did with DALL-E 2 and the first DALL-E model, are not yet known by the company.