Google is pushing out its most important Bard update yet – the ability to generate an AI photo. The search company has officially announced that its own chat-based AI tool is now able to generate images from text prompts, finally joining the likes of what Bing can do with Microsoft’s Co-pilot.
Read more
Google Challenges Chat GPT with Gemini AI Model
ChatGPT Alternative? Google Bard Available in the Philippines
Bard AI photo generation is simple and should be familiar to everyone familiar with Midjourney – you simply have to make a description or prompt of what you want to see, plug it into Bard, and wait for the results. The more descriptive you are the better, though Bard is smart enough to process additional requests based on the image you just generated.
Google also put some backstops to the images that Bard can make. It can’t, for example, generate images that depict situations that could be harmful or dangerous. It refused one of the prompts that we asked it to do, specifically “make a photo of a three-wheeled e-bike causing traffic in EDSA”, and it refused to create it, citing “Traffic congestion can be a serious issue, and I do not want to contribute to the glorification of unsafe driving practices.”
Results from our image generation look good, though again, it all depends on how descriptive you are when it comes to the image generated. As a reference, all of the photos in this article aside from the screenshots of our image generation attempts are all AI-generated.
