As expected, Google is trying to use generative artificial intelligence to enhance the capabilities of its search engine. The announcement was made at the Google I/O conference, where the company showcased future updates to its search, which will initially be available as an experiment within the new Search Labs program.
The integration of generative artificial intelligence in Google Search should help users to understand topics faster, discover new points of view and search for information more conveniently. The company cites the example that if a user enters a multifaceted question such as “what is better for a family with kids under 3 and a dog, Bryce Canyon or Arches National Park,” generative AI will process that information and provide a synthesized overview of key details, as well as links for further research.
After that, the search engine will suggest the next steps, including the opportunity to ask additional questions. Thus, the user will go into dialogue mode with Google, which should allow them to dive deeper into the researched topic. Importantly, generative AI will preserve context from one question to the next, ensuring a seamless user experience. The company has already used a similar approach with its virtual assistant Google Assistant.
Generative AI also has potential applications in online shopping. It can help users understand product features, compare products, and make choices, simplifying complex decision-making. When searching for a product, the user will receive a concise overview of the important factors to pay attention to, as well as relevant offers. This feature is powered by the Shopping Graph, a database of over 35 billion product listings.
Despite the potential benefits, Google recognizes the inherent limitations of generative AI and learning from linguistic models (LLM). The company plans a balanced approach to the introduction of these technologies in Search in order to protect itself from possible scandals. Therefore, such precautions as censoring requests will be used.
In the initial phase of deployment, these features will be available as part of an experiment in Search Labs called SGE (Search Generative Experience), available in the Google application (for Android and iOS) and the Chrome desktop browser in the US (initially only in English). This should allow Google to collect user feedback and improve search over time.