Bard generates photos now, finally – The Verge
By Emilia David, a reporter who covers AI. Prior to joining The Verge, she covered the intersection between technology, finance, and the economy.
Google’s Bard chatbot is adding AI image generation, catching up on a feature that rival ChatGPT Plus has had for months.
Users can prompt Bard to generate photos using Google’s Imagen 2 text-to-image model. Bard, now powered by Google’s Gemini Pro large language model, was always going to have image generation. It was assumed the more powerful Gemini Ultra model would power it; however, that model remains in development.
Google has been positioning Bard as a worthy competitor to OpenAI’s ChatGPT Plus, which runs GPT-4 and lets users generate images thanks to DALL-E 3 integration. Both chatbots perform well, but Bard’s lack of text-to-image features gave ChatGPT Plus a bit of an edge. People can use the updated Bard with Imagen 2 at no cost, unlike ChatGPT Plus, which relies on a paid subscription.
AI image generation came under fire recently after sexually explicit fake photos of Taylor Swift flooded X. While it’s still unknown which platforms people used to generate those images, tech companies have started to lock down their image generating platforms.
Google said Bard’s image capability was “designed with responsibility in mind.” The company’s watermarking feature will be baked into the pixels of the photos, helping to indicate that the images are AI-generated. Google also said it had implemented technical and safety guardrails to avoid generating images of well-known people and limit violent, offensive, or sexually explicit content.
Image generation will not be limited to Bard. Google released a new experimental photo tool powered by Imagen 2 called ImageFX. ImageFX lets users create images with simple text prompts so people can play around with generative AI, with the feedback going directly back to Google engineers. As with Bard, all photos generated through ImageFX will be marked by SynthID and follow Google’s AI principles and technical guardrails. It joins Google’s other AI creation tools, MusicFX and TextFX, which launched last year, and is available for testing on Google Labs.
Bard is also getting a handful of other expansions today. While image generation is only in English for now, the chatbot has expanded to over 40 languages in more than 230 countries. The previous version only worked in English and was available in 170 countries. Some languages it supports include Arabic, Bengali, Tamil, and Urdu.
Google also expanded the double-check feature to 40 languages, which lets users evaluate responses by automatically searching the internet to support or contradict content.
/ Sign up for Verge Deals to get deals on products we’ve tested sent to your inbox daily.
The Verge is a vox media network
© 2024 Vox Media, LLC. All Rights Reserved
