You can now use the DALL-E 3 AI image generator inside Bing Chat

If I want to see a dog bathing in a waterfall, but its water is Skittles, please take a picture of that.

The image generation capabilities within AI-powered Bing Chat have received a significant enhancement. Microsoft has officially announced the availability of OpenAI’s latest DALL-E 3 model for all Bing Chat and Bing Image Creator users. This rollout occurred gradually over the past week, initially extending to Bing Enterprise users and subsequently to Bing Image Creator. Now, it’s accessible to all users.

Notably, Bing has gained access to DALL-E 3 even before OpenAI’s ChatGPT, which is scheduled to integrate it later this month, but exclusively for paying users. Microsoft is poised to become the dominant player in the image generation tool arena for the foreseeable future.

DALL-E 3 marks the third iteration of OpenAI’s image generation model. This version boasts substantial improvements, with the ability to comprehend prompts more effectively, resulting in the creation of images that are not only more creatively imaginative but also strikingly realistic. Moreover, DALL-E 3 has been engineered for enhanced user-friendliness. It is seamlessly integrated into Bing Chat and ChatGPT, eliminating the need for a standalone product. This means users can engage in conversations with a chatbot to create and fine-tune their images, sparing them the need to endlessly refine their initial prompts.

OpenAI has incorporated enhanced safety features into DALL-E 3. Notably, it avoids generating images of public figures and steers clear of producing hateful or NSFW content. In Bing Image Creator, Microsoft has taken additional steps by adding watermarks to AI-generated images for easy identification. Furthermore, they have implemented their own content moderation system to ensure responsible usage. However, the true effectiveness of these safety measures will ultimately be determined by the actual output of the generated images.

Also Read | OpenAI has introduced Dall-E 3, the most recent iteration of its text-to-image tool

Microsoft’s ambitions for DALL-E technology extend beyond Bing. They are actively developing an AI image creation tool called Paint Cocreator, which will integrate the DALL-E model directly into the Windows Paint application.

Through Bing, anyone can now use DALL-E 3. In spite of this, the servers have been struggling to handle the demand. Microsoft did not anticipate such a massive influx of interest, as stated by Mikhail Parakhin from Microsoft, “We expected some strong interest, but we didn’t expect that much.”

DALL-E 3 Characteristics Performance

This advanced neural network creates realistic and varied images based on text descriptions.

It shows improved performance over previous versions in three key areas:

• Understanding prompts: DALL-E 3 interprets prompts more accurately and reliably. Create detailed prompts to refine the end result.
• Logical coherence: DALL-E 3 produces images that demonstrate enhanced photorealism and consistency in response to the prompt.
• Artistic quality: DALL-E 3 can generate stylistic, creative images that adapt to users’ artistic preferences.

Also Read | Microsoft Copilot Brings AI to Windows 11, Works Across Multiple Apps and Your Phone

DALL-E 3 Digital Watermarking

Bing Image Creator incorporates an invisible digital watermark, aligning with the C2PA specification. This discreet watermark contains essential details such as the image’s creation time and date, serving as a verification of its AI-generated origin.

When generating an image with DALL-E 3, it not only fulfills your initial request but also provides helpful suggestions for further customization. For instance, it might suggest adding a rainbow to the background, transforming a dog into a cat, or incorporating birds around a waterfall. While the system’s overload prevented me from experiencing this interactive feature in Bing Chat, it promises an engaging and enjoyable way to collaborate with the tool in creating unique images.