What you need to know
- Google is rolling out a couple of updates today (August 28), beginning with Imagen 3 in Gemini apps.
- Users can create “photorealistic” and high-quality images of animals, objects, plants, and more with Google stating that generation of people will come later.
- Gemini also picks up custom Gems, which lets users create a tailor-made version of the AI for a specific need.
Google is rolling out some updates today (August 28) for its AI model, Gemini, for image generation and custom Gems.
Users seeking assistance through Gemini apps are being treated to an Imagen 3 upgrade. Google reiterates that Imagen 3 is its new “standard” for “generating images in just a few words.” The company states users can create photorealistic landscapes and images that mirror oil-based paintings.
Additionally, Google touts Imagen 3’s ability to understand a user’s natural speech to understand the intent behind their description.
Imagen 3’s rollout in Gemini apps will remain fixed on animals, plants, and other objects at the start. Google states that “in the coming days,” it will work to bring the ability to describe and have Gemini create images of people. This is slated to arrive in early access for Gemini Advanced, Business, and Enterprise users in English.
Despite Imagen 3 giving the user full control over their text-to-image generation process, Google has set limits for photos of people. For example, the software will not support photorealistic photos or “identifiable individual” creations.
Image 1 of 2
The final update, which involves Gemini Advanced, Business, and Enterprise users, brings “custom Gems” for the AI model. The press release states that these users can soon begin “creating and chatting” with their custom-made Gemini variant. This was a feature Google teased during I/O 2024, which lets users fine-tune Gemini for a specific need.
As it rolls out over “the coming days,” Google says users will find a few pre-made custom Gem options for Gemini. “Learning Coach” is said to help break down information for users who are learning a new or difficult topic for the first time. The “Brainstormer” option is geared toward offering inspiration for a range of topics.
The remaining pre-made options are as follows:
- 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.
More importantly, those creating a custom Gem should be as descriptive as possible. Google explains that users can instruct the custom Gemini variant on its purpose and “how it should respond.”
Users can head into the Gemini app on desktop or mobile to find Gems and the ability to create their own. Most languages are supported as Google kickstarts its global rollout.
Those interested in trying what’s new for Gemini before anyone else can sign up for Gemini Advanced or Gemini for Workspace.
