In a significant update today, Google has introduced a transformative feature to Bard, its advanced AI system. Notably, Bard can now seamlessly integrate with Gemini Pro, Google’s latest multimodal AI model, across all languages and territories where Bard is accessible.
Unlike before, where Gemini Pro was exclusively integrated into Bard for English, this enhancement empowers Bard with “more advanced understanding, reasoning, summarization, and coding abilities.”
A notable breakthrough is Bard’s newfound ability to generate images, initially available in English across most countries worldwide at no cost. This capability is driven by Google’s cutting-edge Imagen 2 model, specifically engineered to strike a balance between quality and speed, ensuring the delivery of high-quality, photorealistic outputs.
Utilizing this feature is straightforward – users can input a description into Bard’s text box, receiving customized, comprehensive visuals to bring their ideas to life. To ensure authenticity, Bard incorporates SynthID, embedding digitally identifiable watermarks into the pixels of generated images, providing a clear indication that they were AI-generated.
Moreover, users can now cross-verify Bard’s outputs with Google search results in over 40 languages by clicking on the G icon. Bard will assess whether there is corresponding content on the web supporting its response, presenting highlighted phrases for users to explore additional information, whether supporting or contradicting, found through Google search.
According to the Large Model Systems Organization, recognized for evaluating language models and chatbots across languages, Bard with Gemini Pro stands out as “one of the most preferred chatbots available,” irrespective of cost. They emphasize the remarkable progress Bard has made, corroborated by blind evaluations conducted by Google’s third-party raters, positioning Bard with Gemini Pro as “among the top-performing conversational AIs when compared to leading free and paid alternatives.”