Google has announced a new tool for creating images using artificial intelligence. The company introduces the new Semantic Generative Exploration (SGE) search experience, allowing users to generate images directly from written texts.
Microsoft had previously offered a similar capability using OpenAI's DALL-E model. Additionally, users can now compose drafts within the SGE search experience and customize output properties, such as length, style, and desired writing format.
These new features come after a series of rapid updates to the SGE search experience, enabling it to perform various tasks such as summarizing content, generating definitions for unfamiliar terms and code instructions, and providing useful features in travel and product searches.
By utilizing artificial intelligence to generate images, users can specify the type of image they want, such as photographs or artistic illustrations. The SGE search experience yields four downloadable results as PNG files, or users can modify the query to create a new set of images.
This feature was implemented using Google's text-to-image transformation model known as Imagen. The feature can also be used when searching for images through Google's search engine.
If the desired image is not found during browsing, users can input a new query to create a new image. It is mandatory for users interested in using this feature to be adults, due to the potentially inappropriate content that can be generated using artificial intelligence.
Google has implemented strict filtering policies to control this content, preventing the creation of harmful or misleading images. Additionally, it prohibits references to specific individuals or the creation of realistic images depicting recognizable faces. These steps aim to prevent the spread of misinformation.
