Gemini Pro brings AI-generated images to Bard

  • Google has announced a large expansion of Gemini Pro in Bard, with it now available in more than 40 languages across more than 230 countries worldwide.
  • Along with the expansion comes the ability to create AI-generated images in Bard thanks to Gemini Pro.
  • Google is calling this the, “fastest and most capable version of Bard yet.”

After landing in the zeitgeist at the beginning of last year, every technology company is playing catch up with OpenAI’s ChatGPT platform. Aiming to close the gap as quickly as possible is Google, which this week announced a massive expansion for Gemini Pro in Bard.

Gemini Pro debuted to developers and enterprises late last year, and now some of its capabilities are being used to supercharge the AI experience within Bard, which is the people-facing platform that Google built to try to rival ChatGPT.

For the expansion, Google confirms that more than 40 languages are now supported on Gemini Pro in Bard, with it accessible in more than 230 countries across the globe, including South Africa of course. On the languages side, some localisation still needs to happen, as only Swahili has been spotted as a native African language among the aforementioned 40.

Part of the language expansion is to empower the double-check feature for Bard to better scrutinise responses for validity. This is especially important as hallucinations by generative AI platforms continue to be cited as an issue that needs addressing and refinement.

“Since we know people want the ability to corroborate Bard’s responses, we’re also expanding our double-check feature, which is already used by millions of people in English, to more than 40 languages,” the company explained in a blog post.

“When you click on the “G” icon, Bard will evaluate whether there is content across the web to substantiate its response. If it can be evaluated, you can click the highlighted phrases and learn more about supporting or contradicting information found by Search,” it added.

The other significant element of this update relates to AI-generated images, as Bard will soon get the ability to create images based on prompts. It should be noted, however, that this is limited to English for now, with no mention of when the 40-plus languages supported on Bard will be able to do the same.

“This new capability is powered by our updated Imagen 2 model, which is designed to balance quality and speed, delivering high-quality, photorealistic outputs. Just type in a description — like ‘create an image of a dog riding a surfboard’ — and Bard will generate custom, wide-ranging visuals to help bring your idea to life,” the post continued.

It is from here that Google unpacks how it will tackle the contentious issue of copyright and what content its AI trains on in order to create said AI-generated images.

“To ensure there’s a clear distinction between visuals created with Bard and original human artwork, Bard uses SynthID to embed digitally identifiable watermarks into the pixels of generated images,” Google pointed out.  

“Our technical guardrails and investments in the safety of training data seek to limit violent, offensive or sexually explicit content. Additionally, we apply filters designed to avoid the generation of images of named people,” it concluded.

With Bard now expanding its accessibility and some of its features, it’s time for the other tech companies chasing ChatGPT to do the same.


About Author


Related News