Google's AI-Powered Search: Generating Images & Drafts
Google’s AI-powered search feature, SGE (Search Generative Experience), is continuously evolving and introducing new skills. In this blog post, we’re going to explore two exciting enhancements to SGE: image generation and draft writing. These advancements demonstrate Google’s commitment to leveraging artificial intelligence to enhance the search experience. Whether you’re looking for visually appealing images or need assistance in crafting written content, SGE has got you covered. Let’s dive in and discover how these new features can benefit tech-savvy individuals who want to stay at the forefront of AI technology.
Google’s AI-powered search feature, known as SGE (Search Generative Experience), is evolving with new capabilities. This AI feature, renowned for its conversational mode in Search, is set to enter the realm of image generation directly within SGE. This aligns with the capabilities offered by competitors like Bing, which supports OpenAI’s DALL·E 3 for image generation. Furthermore, SGE is enhancing user experience by enabling the drafting of content within the platform. Users can now customize the length of the output, adjusting it to be longer or shorter, and even alter the tone of the writing, toggling between more serious or casual styles.
These latest features arrive in the wake of a flurry of rapid updates to SGE, reflecting the accelerated pace of AI technology development. In recent months, SGE has seen a significant expansion of its capabilities, including the ability to generate AI-powered summaries, provide definitions for unfamiliar terms, as well as notable coding enhancements and the introduction of travel and product search functionalities, among various other improvements.
The recently introduced AI image generation feature allows you to input a specific prompt, defining the type of image you desire, whether it’s a drawing, a photograph, or a painting, for instance. SGE promptly generates four results within the conversational interface of SGE. Users can interact with these results by tapping on the images, which are downloadable in .png format. Additionally, you have the flexibility to edit the initial prompt to generate a fresh set of images. At its core, this feature harnesses the capabilities of Google’s Imagen text-to-image model to deliver these results.
This same feature will soon be accessible within Google Image search. While browsing through image search results, if you can’t find the specific photo you’re looking for, you’ll have the option to generate a new image using prompts provided in a convenient box that appears within the search results.
It’s worth noting that due to instances of inappropriate content resulting from AI image generation, Google is taking precautionary measures. As a result, this new image generation feature will be restricted to users aged 18 and older, despite SGE having recently expanded access to teenagers in the U.S. between the ages of 13 and 17.
Moreover, Google is actively considering the responsible introduction of this technology, a process that has led to the implementation of rigorous filtering policies within the product, as conveyed by the company. These filters have a clear mandate: to inhibit the generation of images that may be “harmful, misleading, or explicit,” as stipulated by Google.
They also serve to ensure adherence to Google’s strict policies governing the appropriate use of generative AI. Furthermore, the technology proactively restricts the creation of images featuring photorealistic faces and curtails prompts that reference notable individuals. These measures have a dual purpose, aiming not only to prevent the generation of inappropriate content but also to counteract the spread of misinformation.
Naturally, people have shown a degree of ingenuity when it comes to tinkering with AI image generation tools, often producing obscene or unconventional results. This was evident earlier this month when both Meta and Bing’s AI image generation tools gained notoriety for their misuse in generating inappropriate or startling images, even in the face of the existing company policies.
Google acknowledges that its tools may not be flawless. As a precautionary measure, they remain opt-in, initially accessible through Google Search Labs. Additionally, Google has incorporated a feedback mechanism that enables users to report instances when the technology malfunctions or is used inappropriately. Despite these safeguards, the company is confident in its efforts to implement robust filters and anticipates that users may encounter more blocks than they might initially expect. This seems like quite a challenge!
Furthermore, the images will include embedded metadata in the files, clearly identifying them as AI-generated content, in addition to concealed watermarking that remains imperceptible to the human eye. This technology is facilitated by SynthID, a system jointly announced by Google Cloud and Google DeepMind at the end of August.
The latest addition enhances SGE’s role as a writing assistant. Initially, SGE could generate a draft, a feature Google introduced with the idea that users might want to compose emails to companies whose advertisements or links they came across while using Search. However, this draft-writing feature has been expanded to create various types of content, allowing users to select longer or shorter pieces and adjust the tone to be more serious or casual.
Both of these new features, the option to create different drafts and generate images, will come with export functionality in SGE. For the former, you can export your written content to Google Workspace apps such as Gmail or Google Docs. As for the images you generate, they can be conveniently saved to your Google Drive.
Starting today, both features will be progressively introduced to a subset of SGE users and will subsequently be made available to a broader user base over the following weeks. These enhancements will be accessible to users who have chosen to utilize SGE through Google Search Labs. It’s important to note that, for the time being, these features will be exclusively offered in English in the U.S., even though SGE has been recently introduced in India and Japan.
Scenario 1: A Transformative Shift in Content Creation
With Google’s SGE introducing image generation and draft writing features, the landscape of content creation undergoes a significant transformation.
- Content creation becomes more accessible to individuals and smaller businesses.
- Traditional roles in graphic design and content creation may evolve.
- A rise in user-generated content reshapes industries like marketing and advertising.
- Increased automation leads to a higher volume of creative work.
- Enhanced user experience through tailored content and visuals.
Scenario 2: The Evolution of Online Interaction
Google’s SGE, now equipped with image generation and draft writing capabilities, redefines how people interact online.
- Enhanced engagement in social media and online communication.
- Online discussions and debates may involve AI-generated content.
- Broader use of AI-driven personal assistants and chatbots for dynamic conversations.
- Users collaborate with AI in diverse online spaces.
These scenarios demonstrate that Google’s SGE with new AI capabilities will not only alter content creation but also reshape online interactions, possibly leading to fundamental shifts in various digital domains.
As Google’s AI-powered search, SGE, continues to evolve, the introduction of image generation and draft writing features brings exciting possibilities for users. The ability to generate images directly within the search experience provides a convenient and creative way to find the perfect visual content. Additionally, the draft writing feature offers flexibility in creating written content, allowing users to customize the length and tone of their drafts. These enhancements not only showcase the capabilities of AI technology but also highlight Google’s commitment to providing innovative and user-friendly search experiences. Stay tuned for more advancements as SGE continues to evolve, empowering users to explore and create in new and exciting ways.
How do you envision integrating AI image generation into your daily life or work? Are there specific applications that come to mind? Which of the introduced features, image generation or draft writing, are you most excited to explore and why? Share your insights below.