Google is bringing new changes to its Search Generative Experience (SGE), an AI-powered experimental version of the company’s search engine that was launched in May.
On Wednesday, the tech giant announced SGE will be getting a few updates over the next few weeks that will enhance users’ search experience and improve the tool’s overall performance.
Google’s AI Search Engine To Display More Videos And Pictures With Results
People who have signed up for SGE in Google’s Search Labs will soon start to see more images and videos related to the topic they are looking for.
Google has optimized SGE by merging multiple search modes – image, video, and text – into a single block that will definitely enhance the users’ experience. The intention is to deliver more multimedia to help illustrate the topic searched by the AI.
For example, If you ask it for suggestions on how to chop potatoes into fries, SGE will not only provide you with text information scrapped from across the internet but also include images and YouTube videos that elaborate the process to help you get to the answer more quickly.
This is not a new feature for Search. The aim here is not just to find you good links, but to try to synthesize and generate relevant and accurate information. Google has been integrating more and more YouTube-oriented results to search over the past couple of years.
With SGE, people will be directed to a specific portion of a video that will help them with the query “How to chop potatoes into fries” instead of having to skip through the entire video.
SGE Adds Dates To Source Links To Help Users Understand If The Information Is Relevant
Another area of focus for SGE is surfacing and contextualizing links. The AI is now going to display publishing dates next to the top three articles showcased in the summary box.
The Silicon Valley giant says this is an effort to help users understand how recent the information from any web page is and if it still remains relevant on the day of the search.
Android-related news blog 9to5Google discovered that SGE was also starting to show prominent in-line source links with certain queries. Along with the search result, the new sourcing method also showed the name of the publication or website and where the data is coming from.
This was apparently a test feature that Google tried out with some SGE users and there is no confirmation whether it will be widely applied. For the time being, Google continues to display short paragraphs without any sources.
Another issue faced by the company is the time taken by SGE to generate answers. All of the major language-model-based AI tools, like Bing, ChatGPT, Brad, and SGE, take a few seconds to compile the necessary information and answer questions. Meanwhile, normal search engines can deliver results within milliseconds.
Google Cuts Time Taken To Deliver Search Results In Half
Google announced it cut the loading time for SGE in half and made it faster than ever before. By improving the performance of AI overviews, the tool can deliver results much faster, easing pressure on the company’s servers and making it easier to expand SGE to more users.
AI-Powered Language Models Are The Future Of Search Engines
However, after reviewing the tool, the speed difference is barely noticeable. SGE is definitely not slow but when compared to Google’s own search engine, it is miles behind. Generally speaking, SGE is fast enough and provides users with detailed information about queries.
With more refinement, the tool can easily become the defacto Google Search, giving users a much more immersive experience.
At the most recent Alphabet earnings call, CEO Sundar Pichai said that SGE really allows the company to “think outside the box” and not always be constrained to the way search engines used to work before. He anticipates that over time, language model-based tools will be how a search engine works.
SGE is a look into the not-so-distant future of search engines.