Google’s new Search Generative Experience (SGE) allows users to create AI-generated images directly from its search bar and has implemented a DeepMind-developed system, SynthID, to watermark generated images.
Key Points
- Google’s SGE enables users to input natural language prompts in the search bar and receive AI-generated images created with their Imagen text-to-picture AI.
- Users can refine the four alternative images provided by the system and even export them to Google Drive or download them directly.
- A feature also allows users to utilize Google Lens on generated images to locate real-world products resembling the AI-created visuals.
- Restrictions include a necessity to be opted into Google Labs and the SGE program, availability only in the US and for English-language applications, and limited to users over 18 years old.
- Google has implemented SynthID, a system designed by DeepMind, to embed a visually imperceptible watermark on images, indicating they are AI-generated and providing creation data.
Key Insight
Google integrates a unique blend of AI imaging and natural language processing, expanding its generative capabilities to provide users with innovative image-searching experiences, albeit with notable limitations and embedded meta-watermarks.
Why This Matters
This advancement signifies a pivotal shift in user interaction with search engines, intertwining visual creativity with information retrieval, thereby transforming mere search functionality into an engaging, imaginative, and personalized user experience. However, the limitations and restrictions embedded ensure a controlled use of this technology, balancing innovation with ethical and practical usability, which may pave the way for new norms in digital content creation and consumption.
