Google AI search experience Update
Google’s AI-driven search innovation, known as SGE (Search Generative Experience), is evolving and broadening its capabilities. One notable update is the introduction of image generation directly within SGE, similar to the functionality offered by Bing, which leverages OpenAI’s DALL-E 3. With this enhancement, SGE now enables users to prompt the creation of images, further enriching the conversational search experience.
Additionally, SGE has empowered users to craft drafts within the platform, allowing for customized outputs in terms of length and tone. Whether you require a more serious or casual style of writing, SGE adapts to your preferences, enhancing its versatility.
These latest enhancements reflect the rapid evolution of SGE as Google AI technology advances at an accelerated pace. In recent months, SGE has been enriched with various features, including the ability to generate AI-driven summaries, provide explanations for unfamiliar terms, and even improve its coding capabilities. It has also expanded its functionality to encompass travel and product search capabilities, among other improvements.
With the new AI image generation feature, users can input prompts specifying the type of image desired, whether it be a drawing, photograph, or painting. SGE promptly furnishes four image results directly within the conversational experience. Users can easily interact with these images, downloading them as .png files or modifying the initial prompts to generate a fresh set. Behind the scenes, this feature leverages Google’s Imagen text-to-image model, enabling seamless image generation.
This innovative image generation capability is not limited to SGE alone; it’s also being extended to Google Image Search. When browsing through image search results and unable to locate the specific image you require, a convenient box with prompts will appear, allowing you to create a new image tailored to your needs.
However, it’s worth noting that AI image generation has raised concerns related to inappropriate content. In response to these concerns, Google is implementing certain restrictions. The new image generation feature will be accessible exclusively to users aged 18 and older. This decision comes even after SGE recently expanded its services to teenagers in the United States, specifically those between the ages of 13 and 17.
In an effort to introduce this technology in a responsible manner, Google has implemented stringent filtering policies within the product. These filters are designed to prevent the creation of images that could be deemed harmful, misleading, explicit, or in violation of Google’s prohibited use policies for generative AI. Furthermore, Google has taken steps to block the generation of images featuring photorealistic faces and prompts containing the names of prominent individuals. These restrictions aim not only to curb inappropriate content but also to combat the spread of misinformation.
Nonetheless, individuals have demonstrated a knack for creatively utilizing AI image generation tools to produce unconventional and sometimes inappropriate outputs. Recent incidents involving the misuse of Meta and Bing’s AI image generation tools went viral, despite existing policies and safeguards.
Google acknowledges that its tools may not be infallible. As a result, these features remain opt-in, accessible through Google Search Labs. Users are also provided with a feedback mechanism to report instances where the technology misfires or is misused. Despite potential imperfections, Google is confident in the effectiveness of its filter implementation and believes users might encounter more restrictions than they initially anticipate—a challenge of sorts.
Additionally, the generated images will contain embedded metadata that clearly identifies them as AI-generated. An invisible watermark, powered by SynthID, a project announced by Google Cloud and Google DeepMind in August, further enhances their authenticity.
Building upon SGE’s existing capabilities as a writing assistant, a new feature now allows users to create various types of written content. Users can modify the length of the output or adjust its tone to be more serious or casual, thereby enhancing the versatility of this draft writing feature.
Both of these exciting new features, namely the ability to create alternate drafts and generate images, come with convenient export options within SGE. For those utilizing the draft writing feature, your content can be easily exported to Google AI Workspace applications such as Gmail and Google Docs. Meanwhile, any images you generate through SGE can be effortlessly saved to your Google Drive, providing seamless access to your creations.
These new features are set to launch, with an initial rollout beginning tomorrow for a select group of SGE users. The availability will then gradually expand to a broader user base over the course of the coming weeks. These features are accessible to users who have opted in to use SGE through Google Search Labs. Initially, they will be offered exclusively in English within the United States, despite SGE’s recent expansion to India and Japan.
