Google's Search Generative Experience (SGE), powered by AI, is receiving significant updates. SGE will now have the ability to generate images based on user prompts, much like its competitor Bing's support for OpenAI's DALL-E 3. Additionally, users can create drafts within SGE and customize the length and tone of the writing.
This is part of a series of updates to SGE in response to the rapid development of AI technology. It has recently gained capabilities for creating AI-powered summaries, defining unfamiliar terms, travel and product searches, and more.
The AI image generation feature allows users to specify the type of image they want and receive results within the SGE conversational experience. Google's Imagen text-to-image model powers this feature. It's also integrated into Google Image search.
Google is aware of the potential misuse of AI-generated images and has implemented strict filtering policies to prevent harmful, misleading, or explicit content. The company is taking steps to ensure responsible use and is blocking images containing photorealistic faces and notable people's names.
Images generated using SGE will be tagged as AI-generated and contain invisible watermarking to identify their origin.
The draft writing feature can now provide different types of writing, allowing users to customize to a longer or shorter text and and also change the tone of the message. Both features will support export options to Google Workspace apps and Google Drive.
These features will roll out to SGE users, starting in the coming weeks, with initial availability to English-speaking users in the United States.
Google is being cautious with these AI features, providing opt-in and reporting mechanisms for users. Despite these efforts, the company acknowledges that challenges remain in ensuring responsible usage.
These developments reflect Google's commitment to harnessing AI for a more interactive and creative search experience.