It is a secret to no one that Bing Chat With its Artificial Intelligence, it has managed to position itself in recent months as the most ascending fame seeker, also becoming the first to have managed to compete with Google.
And it is all due in large part to its association with OpenAI to integrate ChatGPT and DALL-E technology into its engine, achieving a much more robust experience than the traditional one.
Now things have escalated to a new level with the announcement that the alliance between both companies reaches a new stage with the integration of the new Artificial Intelligence of DALL-E 3 to Bing Image Creator.
If you have closely followed the evolution of the OpenAI image generative platform, you will know that this third version is much more powerful than the previous one, now achieving much more precise and realistic results.
The first tests of this new AI have clearly demonstrated that DALL-E 3 could now be considered on par with other systems that were considered much more robust in the past, such as Midjourney.
At the beginning of the year the distance between both platforms was more than obvious, but this update considerably shortened the gap and Microsoft was not going to miss the opportunity to take advantage of it.
Bing Image Creator is more powerful than ever thanks to DALL-E 3 Artificial Intelligence
Through a post on your official blog The Microsoft team has confirmed that DALL-E 3 technology has now been fully integrated into Bing Image Creator to improve image quality and detail.
The logical coherence of the model has been greatly polished in this third incarnation, while its ability to understand and translate the prompts of text in images.
Just enter the official site of Bing Image Creator to check. Now Artificial Intelligence can produce more realistic and creative images but it is also essential to be more detailed with the instructions:
“Bing Image Creator works best when written descriptively. So, be creative and add details: adjectives, locations, even art styles like ‘digital art’ and ‘photorealistic’.
Here’s an example: Instead of a ‘creature’ text request: try sending a ‘fuzzy creature with sunglasses, digital art’ request.”
This is what the site itself points out with the update that integrates DALLE-3. In other words, text instructions should now be as close to those in Midjourney as possible, omitting details about image sizes.