OpenAI Debuts DALL-E 3, the Latest Text-to-Image Tool After GPT-4

OpenAI Debuts DALL-E 3, the Latest Text-to-Image Tool After GPT-4

OpenAI has introduced the third version of its AI art platform called DALL-E. This new version allows users to use ChatGPT to create prompts and offers enhanced safety options.

DALL-E is a tool that turns text prompts into images. In the past, even DALL-E 2 had its shortcomings, often not fully understanding specific instructions. However, the latest version, according to OpenAI researchers, has a much better grasp of context.

One notable feature of DALL-E 3 is its integration with ChatGPT. Instead of crafting a detailed prompt themselves, users can now simply ask ChatGPT to generate a prompt paragraph (DALL-E performs better with longer sentences), which DALL-E 3 will then use. Of course, users who have specific ideas can still provide their own prompts.

Currently, DALL-E 3 is available as a research preview, but ChatGPT Plus and Enterprise users will gain access to it in early October.

OpenAI has taken measures to make DALL-E 3 safer, preventing it from generating violent, adult, or hateful content. It also restricts requests for images of public figures by name and images that mimic the styles of living artists. Creators have the option to exclude their work from future text-to-image training tools.

OpenAI faces competition from other players like Alibaba’s Tongyi Wanxiang, Midjourney, and Stability AI, all working on improving their image generation models. There are ongoing concerns about AI-generated images, with a recent court ruling in Washington D.C. stating that AI-created art with no human input cannot be copyrighted under U.S. law.

ChatGPT plays a crucial role by taking short prompts and making them more descriptive, providing better guidance to the DALL-E 3 model.

DALL-E 3 brings improved image quality, especially with longer prompts, and better handling of challenging content like text and human hands.

This release of DALL-E 3 coincides with increasing competition in the generative AI field, particularly in image synthesis. Rivals such as Midjourney and Stability AI are refining their image generation models, putting pressure on OpenAI to stay at the forefront.

OpenAI’s plan is to introduce DALL-E 3 to premium ChatGPT users in October, followed by research labs and API customers. However, there’s no mention of when or if they will release a free web tool, similar to DALL-E 2 and the original DALL-E model.

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *