Bing Image Creator

About

Bing Image Creator is the built-in AI art generator developed by Microsoft for use in Bing, as well as the Microsoft Edge browser. The image creator uses the artificial intelligence model DALL-E 3 and is available to all users, unlike other AI image generators that have waitlists, paid memberships and other requirements for use.

History

On March 21st, 2023, the official Microsoft blog uploaded an announcement that the Bing assistant would be able to create images based on prompts given to it.

On September 29th, 2023, it was then announced that OpenAI had added DALL-E 3 to Bing, without a waitlist, and that it was capable of putting words on the image, making it unique in this regard. That same day, X user @AImojopro^[1] uploaded a video to X in which they showcased how the process worked (shown below).

𝟮. 𝗨𝗽𝗱𝗮𝘁𝗲𝘀 𝘁𝗼 𝗕𝗶𝗻𝗴 𝗮𝗻𝗱 𝗘𝗱𝗴𝗲:

• OpenAI's DALL-E 3 for picture generation
• Search history-based personalised responses
• A shopping experience powered by AI
• Bing Chat Enterprise updates pic.twitter.com/9vFhskKfBA

— Aimojo (@aimojopro) September 29, 2023

Highlights

News of the capabilities of Bing's image creator update quickly spread, with people using it to create images that were unable to be generated using other generators at the time. For example, on September 30th, 2023, X user @itsPaulAI^[2] uploaded two example images, one of Spider-Man done in a comic book style and one of the Pokemon Charmander in the style of an anatomy chart (shown below).

As users began to experiment with the new tool, some began to see what its strengths and weaknesses were, as well as how the images were being generated.

For example, on September 30th, 2023, X user @Xploshi^[3] posted an image of Kermit the Frog eating Doritos as if part of an advertisement, postulating that the image creator is very skilled at logos and known things, hinting at it taking existing photos for its generation (shown below).

AI-Generated Pitbulls vs. Toddlers

On September 30th, 2023, an anonymous 4chan^[4] user started a Pitbulls vs. Toddlers thread on the /an/ board with an image of a demonic pitbull towering over a toddler generated with Bing Image Creator. In the thread, multiple other users shared more AI-generated pitbull memes, including giant pitbulls terrorizing toddlers and babies having a soulslike stand-off against the dogs (examples shown below).

In the following days, similar threads appeared on /v/^[5] and /pol/^[6] boards, with more images generated with Bing Image Creator being shared. On October 2nd, X / Twitter^[7] user @tulpapilled posted an AI-generated image of a pitbull using a baby vending machine, with the post gaining over 430 reposts and 5,400 likes in one day (shown below, left). Also that day, iFunny^[8] user MerilTheMoth posted another image that was featured and received over 12,800 smiles on the app (shown below, right).

Characters Doing 9/11

Bing AI Characters Doing 9/11 refers to a series of images created by users exploiting Bing's Image Creator to make it appear that characters from popular culture are flying planes into the Twin Towers, doing the September 11th, 2001 Terrorist Attacks. The images highlighted the challenge of creating adequate filters for AI image generation to prevent offensive content.

Various Examples

Search Interest

External References