Google rolling out Gems and Imagen 3, with people generation, to Gemini Advanced

Following the preview at I/O 2024 in May, Gemini is rolling out Gems, as well as Imagen 3, to Advanced subscribers.

Gems

Gems are “custom versions of Gemini” that you can create to “act as an expert on topics or refine them toward your specific goals.” They can “remember a detailed set of instructions to help you save time on tedious, repetitive or difficult tasks.”

You enter a paragraph or so worth of “Instructions” and then have Gemini rewrite it into a more structured format that explicitly lays out:

• Purpose and Goals
• Behaviors and Rules: You can specify the tone, maximum sentence length for responses, or even request emojis throughout replies

Everything can be further edited. After saving (and naming), it will appear in your Gem manager. The Gemini side panel is home to this new section and provides quick access to Gems. 

Gemini will offer some premade Gems:

• Learning coach helps you break down complex topics, making them easier to understand. 
• Brainstormer gives you easy inspiration — from fresh ideas for a themed party to the perfect gift for an upcoming birthday. 
• Career guide unlocks your career potential with detailed plans to refine your skills and achieve your career goals. 
• Writing editor can elevate your writing through clear, constructive feedback on everything from grammar to structure. 
• Coding partner levels up your coding skills and can help you build projects and learn as you go. 

Gems are rolling out in the coming days to Gemini Advanced, Business, and Enterprise subscribers.

Imagen 3 

Imagen 3, as seen on Pixel Studiom is Google’s latest image generation model and it’s coming to Gemini with “availability for users in all languages.” Rolling out over the coming days, examples include: 

• A photorealistic image of a breathtaking mountain vista with jagged peaks and snow-capped summits, bathed in the warm glow of a setting sun. Dramatic clouds, painted in vibrant hues of orange, pink, and purple, streak across the sky…
• An image of a ball gown made of paper napkins in an elegant showroom
• An animated image of a tiny dragon hatching from an egg in a sunlit meadow, surrounded by curious glowing butterflies. Vibrant colors, detailed scales
• A vibrant abstract painting with the words “Dream Big” splashed across the canvas in bold colors

Free users will get to use this model, but the ability to generate images of people is coming first to Gemini Advanced, Business, and Enterprise starting in English. 

After the previous removal in February, Google says it has “worked to make technical improvements to the product, as well as improved evaluation sets, red-teaming exercises and clear product principles.” 

With Imagen 3, we’ve made significant progress in providing a better user experience when generating images of people. We don’t support the generation of photorealistic, identifiable individuals, depictions of minors or excessively gory, violent or sexual scenes. Of course, not every image Gemini creates will be perfect, but we’ll continue to listen to  feedback from early access Gemini Advanced users as we keep improving. We’ll gradually roll this out, aiming to bring it to more users and languages soon.