Google’s Bard chatbot gets an AI image generation feature with its recent update along with its language model Gemini Pro now available in over 230 countries in over 40 languages.
You can now create AI images by entering text prompts describing the features of the images within Bard’s conversational chatbot accessible through https://bard.google.com
The image generation capability is powered by Google’s updated Imagen 2 text-to-image model which is designed to deliver high-quality realistic photos.
Google introduced Bard last year and is continuously upgrading its capabilities. First, it integrated LaMDA (Language Model for Dialogue Applications) into Bard, then upgraded to PaLM and currently it’s working on top of the Gemini Pro language model.
With the new update, Bard is giving strong competition to the popular OpenAI’s ChatGPT Plus which uses the DALL.E 3 text-to-image model to generate stunning images.
Microsoft has also forayed into image generation by integrating the same DALL.E 3 model into Bing Image Creator which is also a very popular AI image generation tool nowadays.
With the advent of Artificial Intelligence (AI), most people are looking at it as a revolutionary innovation for the benefit of mankind. But as the technology is advancing, it is signalling some challenges that must be addressed clearly.
Being consistent with its principles, Google has designed Bard’s image generation feature ensuring a clear distinction between visuals created with Bard and original human artwork. For this, Bard is using SynthID to embed digitally identifiable watermarks into the pixels of generated images.
They have also made strict guidelines to limit violent, offensive or sexually explicit content. Google also promises to continue investing in new techniques to improve the safety and privacy protections of their models.
