Google's Imagen 2, the image generator, comes to life with Bard; can work with 40 languages

Google’s AI assistant, Bard, has received updates to improve accessibility, accuracy, and creative potential.

Previously available in select languages, the advanced features of Gemini Pro, which offer improved understanding, reasoning, and coding abilities, are now accessible in over 40 languages and 230 countries.

The popular double-check feature, which allows users to verify Bard’s responses against information from the web, has been expanded to cover over 40 languages. This enhancement promotes confidence in Bard’s answers, which, to be honest, is a big thing with LLMs as they tend to get ‘lazy.’

And the biggest of all, a new feature introduces image generation to Bard. Users can now provide text descriptions, and Bard will generate creative images based on those descriptions. This feature currently uses the updated Imagen 2 model and initially operates in English with the potential for future expansion.

Notably, the image generation feature prioritizes responsible development by employing SynthID watermarks to distinguish AI-generated images from original artwork. Additionally, technical safeguards aim to prevent offensive or harmful content creation.

These updates represent a significant expansion of Bard’s capabilities, focusing on accessibility, accuracy, and creative expression. By offering advanced features in multiple languages, providing enhanced fact-checking options, and introducing image generation, Bard aims to become a more valuable tool for individuals and businesses across the globe.

More here.