DALL-E now allows you to edit pictures via ChatGPT
DALL-E, an AI-based image generator from OpenAI, has received several new features that make it more user-friendly. Image editing tools are now available in both the web version and the ChatGPT mobile app. This allows users to edit images created by DALL-E directly in the chatbot interface. Another innovation is preset styles. DALL-E offers such options as wood engraving, gothic, hand-drawn, and others.
The third version of DALL-E, which was released last September, was integrated with ChatGPT, eliminating the need for users to write perfectly accurate text queries for images. Just tell ChatGPT what image you need, and it will generate a hint. The new option allows DALL-E users to correct and improve their results directly in the ChatGPT interface.
In addition, DALL-E 3 has added new style tips to serve as a starting point for the user. You can choose from styles such as woodcut, gothic, synthwave, and hand-drawn. When you hover the mouse over them, examples will appear. The choice of style will be added to the request, after which the user will be able to add the necessary details.
OpenAI is also working on the problem of trust in images created by artificial intelligence. Earlier this year, DALL-E 3 began adding visible watermarks and metadata to label images created by AI. This is a good step, but attackers can easily remove these labels. Solving this problem is not only a task for OpenAI, but also an important issue for the entire community.
