It looks like Google Bard will be adding its own AI image generator in the coming week. The news was leaked in advance by a tipster on X after they discovered some unreleased notes.
Google has been experimenting with AI image generation since the ‘early days’ of text-to-image prompts with their generator called Imagen. However, it hadn’t yet been released for public use due to biases that the researchers were finding in the system.
Dylan Roussel leaked the patch notes this week on X, saying, “Here’s what’s coming next in Bard tomorrow. Image generation with Bard will use Imagen, Google’s Text-to-Image ‘diffusion technology’.”
The tweet was posted on Wednesday, January 18. However, as of Thursday, there is still no image generator on Bard (at least not in Europe). Of course, Roussel included a caveat to say that the release date and other details could still be subject to change.
Interestingly, PetaPixel asked Google Bard if it would soon have an AI image generator of its own. “Google is indeed planning to add an AI image generator to Bard,” said Bard, “and it’s exciting news for everyone who enjoys using this powerful language model.”
The chat model followed up by saying that it didn’t know the release date yet, either. Although it then added that it would probably be “sooner rather than later.”
Stock site Shutterstock uses Imagen for its own AI generator and assures users that the data was trained ethically. However, Google has never disclosed where it sourced the training data from. Up until now, Adobe Firefly is the only truly ethical AI image generator, having trained its machines on images it owned the rights to.
Talking of Adobe, it wasn’t long ago that Adobe and Google announced a partnership between Bard and Firefly. I wonder what happened there…
[via petapixel]
