Gemini is coming soon to AI Agent Gems and Imagen 3 capabilities
Gemini apps are getting two new advanced capabilities, Google announced Wednesday. The Mountain View-based tech giant’s in-house artificial intelligence (AI) chatbot is getting the AI agent Gems and image generation capabilities from the recently released Imagen 3 AI model. While the former will only be available to Gemini’s paying users, the latter will be rolling out to all users, including those on the free plan. Those on the free plan, however, may notice some additional limitations on image generation.
Gemini to get gemstones, picture 3 possibilities
Google has announced in a blog that it will integrate Gems and Imagen 3 into the Gemini apps after. Both features were first previewed at Google I/O earlier this year. Gems, in particular, has already rolled out and will be available to Gemini Advanced, Business, and Enterprise users. The company said that Imagen 3 features will begin shipping to Gemini, Gemini Advanced, Business, and Enterprise users in the coming days.
Gems are essentially miniature versions of the chatbot with a limited data set. They can be customized to focus on a specific set of topics, allowing the AI model to generate more specific and accurate information. Google said: “With Gems, you can assemble a team of experts to help you brainstorm ideas for a challenging project, brainstorm ideas for an upcoming event, or write the perfect caption for a social media post.”
Users can also add specific instructions to a Gem to further refine their answers. Once the feature is available to users, they will also find a set of pre-made Gems created by Google. These include Learning coach, Brainstormer, Career guide, Writing editor, and Coding partner. Gems are available in multiple languages on desktop and mobile devices in over 150 countries.
Imagen 3, the company’s latest AI image generation tool, is also rolling out to Gemini apps. It can generate images in a variety of styles, including Nikon DSLR, GoPro-style, wide-angle lens and more. Google says it can also generate “photorealistic landscapes, textured oil paintings, or whimsical clay animation scenes.”
A major upgrade with Imagen 3 is that the AI model also lets users generate images of people, something that was removed after many users noticed that Gemini was generating biased and harmful images of people. To reduce the risk of deepfakes, the company says it has added built-in protections. Additionally, SynthID has been used to watermark the images as generated by the AI.
While the company didn’t specify, it did indicate that Imagen 3’s capabilities may also include inline editing of the generated images. However, it appears that editing can only be performed using text prompts. Google specifically says that Imagen 3 “does not support generating photorealistic, identifiable people, images of minors, or excessively gory, violent, or sexual scenes.”
