Microsoft’s Bing artificial intelligence (AI) chat function has already proven useful for answering queries, and now it has been enhanced to enable users to create imaginative illustrations. Today, Microsoft is introducing an AI-based image generator to its Bing search engine. The Bing Image Creator leverages an improved iteration of OpenAI’s DALL-E model, enabling users to fabricate visuals by entering text descriptions.
The company has launched a preview of the Bing Image Creator that integrates OpenAI’s DALL-E AI image creation technology into both Bing search and a sidebar within the Edge browser. This incorporation of cutting-edge AI allows users to explore their creativity, providing them with a simple and convenient means of generating images that align with their artistic vision.
To generate an image using the new Bing Image Creator, all you need to do is provide the chatbot with a description or follow-up to a previous inquiry. For instance, if you desire an image of an elderly man with a walking stick and a cheerful dog by his side, you can prompt Bing to generate design concepts based on your criteria. Check out the result:
Yusuf Mehdi, Microsoft’s Head of Consumer Marketing, has stated that those who have access to the Bing preview will be able to enjoy a fully integrated Bing Image Creator chat experience. The feature will initially be available in Creative mode. “By inputting a textual description of the desired image, offering supplementary details such as location or activity, and selecting an artistic style, the Image Creator will bring to life a visual representation from your own imagination,” Mehdi explains.
Furthermore, Microsoft is introducing the Bing Image Creator to its Edge browser. A new icon will be displayed in the Edge sidebar, allowing users to produce images via a text entry prompt. At the moment, Microsoft is only offering this image creator in Bing’s creative mode and is intending to optimize its functionality in multi-turn conversations. “Initially, we will restrict the Image Creator to the creative mode of Bing chat, with the goal of making it available in balanced and precise mode in due course,” Mehdi explains.
Bing has three distinct modes of responses, namely creative, balanced, and precise. Creative mode focuses on generating responses that are “innovative and imaginative,” while the precise mode prioritizes accuracy and relevance for more factual and succinct answers.
In an effort to avoid any potential controversies, Microsoft is taking extra care with the introduction of its image creation feature on Bing. After facing criticism for the inappropriate responses generated by the chatbot in certain instances, the company is implementing a more deliberate approach to its image creation feature. With the aim of ensuring that users receive high-quality and relevant images, Microsoft is initially launching the feature only in the creative mode of Bing chat. The company plans to make the feature available in other modes over time, once it has been thoroughly tested and refined.
Introducing image creation in the new Bing and Edge. Now you can use Bing Chat to create images with just your words. https://t.co/mCSrEADARl pic.twitter.com/SF1jmkF1aZ
— Yusuf Mehdi (@yusuf_i_mehdi) March 21, 2023
Microsoft has started to roll out the Bing Image Creator, which has been integrated into Bing chat, to Bing preview users on desktop and mobile devices from today. For those who are not a part of the Bing preview program, the Bing Image Creator can be accessed separately via bing.com/create for English text-to-image generation. Microsoft plans to add support for more languages in the future. Bing preview users can access the Image Creator via the standard chatbot prompt or through Microsoft Edge.
While Microsoft’s rapid development of AI-powered text-to-image features for Bing is impressive, experts have raised concerns about the potential risks associated with such technology. Some experts warn that text-to-image systems could produce images that are used for propaganda, harassment, misinformation, and other criminal purposes. As with any new technology, it’s important to carefully consider the potential risks and take steps to mitigate them. It remains to be seen how Microsoft will address these concerns and ensure that its image creation feature is used in a responsible and ethical manner.
Microsoft has stated that it is taking measures to prevent any potential misuse of its AI-powered text-to-image feature. Yusuf Mehdi has clarified that the company has incorporated OpenAI’s safeguards into the Image Creator and added extra protections to limit the generation of harmful or unsafe images. Additionally, Microsoft has implemented controls that can detect when a prompt may generate a potentially harmful image and block it, providing a warning to the user instead. It’s reassuring to see that Microsoft is taking a proactive approach to addressing concerns about the potential misuse of its technology.
Microsoft continues to incorporate more forms of AI into its Bing search engine. In addition to the text-to-image feature, the company is now rolling out AI-based Knowledge Cards that provide “dynamic” quick-glance information like charts and timelines. These new Knowledge Cards aim to provide users with more in-depth information and a better understanding of their search results.
Moreover, Bing is also offering “Stories” which provide images and short videos related to searched topics. The Stories feature is powered by AI and provides a visually engaging way for users to explore and learn about various subjects.