Google Bard Is Getting Its Own AI Image Generator Thanks To Adobe
Google is partnering with Adobe to give some serious content generation superpowers to Bard, its in-house rival to OpenAI's ChatGPT and Microsoft's Bing Chat system. At the I/O 2023 developers conference, Google announced that Bard is integrating Adobe's Firefly tool, allowing users to create images using text-based prompts. To recall, Firefly is Adobe's own generative AI tool that pulls the same image creation tricks as OpenAI's Dall-E engine. Firefly went into the beta phase earlier this year, and so far, Adobe says the tool has been used to create over 70 million images.
With Firefly now becoming an extension of Bard, users can create images directly from the Bard interface without any third-party app or browser extension shenanigans. In doing so, Bard is also catching up with Bing Chat, which already offers an AI image creation tool based on the same OpenAI tech that powers the Dall-E image creator. However, Adobe is touting an ethical upper hand for Firefly in Bard over Microsoft's Bing Chat.
Adobe
Firefly in Bard relies on content without any copyright snags and "hundreds of millions of professional-grade, licensed images" in Adobe's own stock library. That's a crucial differentiator, as the likes of Stability AI have been sued by Getty and face backlash from the artist community for using their work to train generative AI models without giving them due credit or compensation. Adobe says its partnership with Google involves Content Authenticity Initiative's (CAI) open-source technology, offering a higher degree of transparency over the AI-generated material.
How Firefly in Bard eclipses Bing Chat
Bing Chat's native image creator follows the same formula as OpenAI's Dall-E image creator, outputting images at 1024 x 1024 pixels resolution in response to text prompts. But once a set of images is created, the only options you get are saving it to an online collection, downloading it locally, or sharing the whole set. There is no option for tweaking or editing the final result, save for entering a fresh text prompt. That's where Firefly in Bard takes the lead, thanks to native integration with Adobe Express.
Express is Adobe's own cloud-based suite of photo editing tools and templates. Adobe says once users are done creating their images, they will "have the ability to edit and further modify it using Express right within the Bard platform." Adobe Express also offers access to a wide selection of templates, fonts, and other multimedia assets. Moreover, Express lets users post their finished AI-generated content directly to social media platforms.
Microsoft currently offers a free credit system that lets users create AI-generated images with a limited number of "boosts" on a daily basis. Once you run out of those free boosts, you will still be able to create as many images as you want, but the process will be slow. Neither Adobe nor Google has detailed any such system yet, and there is no word if an Adobe subscription will be required to access the full suite of Firefly and Express tools in Bard.