Google has relaunched its AI image generation capabilities with the new Imagen 3 model, which offers enhanced quality and built-in safeguards to address previous issues. The new features, previewed at Google I/O, are rolling out to Gemini Advanced, Business, and Enterprise users.
Custom Gems
Gems are customizable AI experts available to Gemini Advanced, Business, and Enterprise users. Users can create Gems to act as experts on various topics by writing instructions and naming them. These Gems can assist with tasks such as coding, career advice, brainstorming, and writing. Premade Gems include:
- Learning Coach: Simplifies complex topics.
- Brainstormer: Provides inspiration for events and gifts.
- Career Guide: Offers detailed career plans.
- Writing Editor: Gives feedback on writing.
- Coding Partner: Helps with coding projects.
Gems are available on desktop and mobile in over 150 countries and most languages.
Imagen 3 Model
Imagen 3 is the latest image generation model, offering high-quality images based on user prompts. It supports various styles like photorealistic landscapes, textured oil paintings, and whimsical claymation scenes. Key features include:
- Advanced Safeguards: Built-in safeguards and adherence to product design principles.
- User Control: Users can modify images if the initial result doesn't meet expectations.
- SynthID: Tool for watermarking AI-generated images.
Imagen 3 is rolling out to Gemini Apps and will soon support the generation of images of people, starting with an early access version for Gemini Advanced, Business, and Enterprise users in English. The model avoids generating photorealistic, identifiable individuals, depictions of minors, or excessively gory, violent, or sexual scenes. Google aims to expand this feature to more users and languages based on feedback. Google continues to improve its AI tools, ensuring a better user experience and adhering to clear product principles.
