Microsoft
Microsoft Bing Chat is introducing DALL-E 3, which is an improved version of OpenAI’s tool that converts text descriptions into detailed artwork. Pixabay

Microsoft Bing Chat is rolling out DALL-E 3, which alludes to an upgraded version of OpenAI's text-to-image tool. A source from the company has confirmed that DALL-E 3 will be available to a "small group of users".

According to the folks at Windows Latest, the ChatGPT-powered DALL-E 3 works as advertised. In other words, you can convert descriptions of images into artwork with more text and details.

Unlike its predecessor, DALL-E 2, the new DALL-E 3 creates high-quality artwork. This improvement can be attributed to how DALL-E 3 handles text within images.

DALL-E 3: What does it offer?

The current generation AI can draw any picture and create unique artwork. However, the previous generation model wasn't good in terms of generating words, labels and signs within images.

The recently unveiled model is good at text generation, allowing OpenAI to produce more accurate labels or signs. For example, you can now instruct Bing Chat AI to create an illustration of an avocado sitting in a therapist's chair, saying, "'I just feel so empty inside with a pit-sized hole in its centre".

Meanwhile, the therapist (a spoon) scribbles notes. Bing Chat AI will be able to accurately produce the results. Also, the text within the image, "I just feel so empty inside", appears correctly.

The image confirms that DALL-E 3 handles text within images significantly better than DALL-E 2, which produces artworks with random alphabets. The new AI model in Bing Chat, on the other hand, showed the exact text as described.

It is unclear how many users on Bing Chat have access to DALL-E 3.

However, OpenAI recently confirmed in a blog post that paid customers of its AI platform (ChatGPT Plus and Enterprise) can access DALL-E 3 in October. In the meantime, the American AI company introduced fine-tuning for the GPT-3.5 Turbo model.

So, ChatGPT Plus subscribers can directly type prompts in ChatGPT and create images. Alternatively, you can use Bing Chat's DALL-E 3 image creator, which offers the same functionality at no cost.

Those who have access to the early preview of DALL-E 3 should be able to try the new image creator in Bing Chat in Microsoft Edge Canary. Aside from this, Microsoft recently updated the Edge browser with AI writing on the web.

DALL-E is better than its predecessors

The new text-to-image AI model boasts a slew of quality improvements over the previous models. To those unaware, the previous models could not handle high-quality image requests, labels and signs.

DALL-E 3 in ChatGPT and Bing Chat is reportedly based on the data trained on millions of images of artists, photographers, internet creators, stock images and more.

"Modern text-to-image systems have a tendency to ignore words or descriptions, forcing users to learn prompt engineering. DALL·E 3 represents a leap forward in our ability to generate images that exactly adhere to the text you provide", OpenAI said in its latest blog post.

Microsoft Paint to use DALL-E AI image creator

A separate report by Windows Latest suggests Windows 11's Paint app is set to receive a big AI upgrade comprising a DALL-E and Bing-powered "Cocreator" feature.

These features will enable the Paint app to create realistic images from text descriptions. So, it is safe to say that the imminent exit of Windows and Surface Chief Panos Panay isn't stopping the tech giant from rolling out significant updates to Windows 11.

The word on the street is that Microsoft is testing a DALL-E and Bing-powered image creator dubbed Cocreator in the Paint app. Notably, Bing-powered image creator uses OpenAI's AI system to create realistic images from a text description.

Despite the lack of an official confirmation, one of the references in the updated app suggests Microsoft will be using the most recent DALL-E models. The cocreator feature in Paint can come in handy for creating high-resolution DALL-E images.