What you need to know
- Microsoft’s “special” Surface and AI event occurred today, where the company made many announcements.
- Among the announcements today, Microsoft indicated that Bing Image Creator would soon leverage OpenAI’s DALL-E 3 technology to generate detailed and engaging images.
- It’s unclear when this change will be taking place.
Microsoft has had quite the morning during its anticipated “special” event for Surface and AI, making a ton of announcements revolving around the release date for the Windows 11 2023 update, significant updates for Microsoft 365 and Bing & Edge, the introduction of two new Surface Devices, and more.
But what perhaps stood out the most (at least for me) was the improved and refreshed Bing Image Creator. As an avid user of the image-generating tool, it takes quite a bit of fine-tuning to generate the exact image you have in mind (but even then, not quite).
However, Microsoft promises to right these wrongs by integrating OpenAI’s DALL-E 3 into Bing Chat. While Microsoft didn’t provide a specific timeline for when users should expect the new model to ship to Bing, the YouTube video shared showcases how fast and detailed its responses are to prompts.
According to OpenAI, DALL-E 3 will make up for its predecessor, DALL-E 2, shortcomings and indicates that it will refine small details like hands, ultimately generating better and more engaging images. The company added that the images will be generated by default with “no hacks or prompt engineering required.”
Yusuf Mehdi, Microsoft’s Corporate Vice President & Consumer Chief Marketing Officer, posted on X (formerly Twitter) highlighting his excitement about the new addition that’s to come to Bing Image Creator. His sentiments were echoed by several other users on the platform, who also seemed eager to access the feature.
I’ve been having so much fun creating breathtaking images with DALL-E 3 from Open AI, now in Bing Image Creator. pic.twitter.com/LfxrnyX2OHSeptember 21, 2023
Bing Image Creator is about to get better
As highlighted above, I use Bing Image Creator a lot, both for work and just for fun. To be honest, the ability to create a unique image using words (“prompts”) still puzzles me.
But after getting used to it, small details like hands and eyes being blurry or odd-shaped started getting to me. Regardless of how detailed the prompt is, the tool couldn’t refine the image to the extent that I wanted. But Bing Image Creator stacks miles ahead compared to other AI-powered image generation tools like Midjourney.
With the DALL-E 3 coming to Bing IImage Creator, we will likely see a giant leap forward in image generation.
