Recently, Google announced Gemini API and Google AI Studio.Grounding with Google SearchWe introduced a new feature called “. This feature, which may sound somewhat unfamiliar to people outside the IT industry, is simply a feature that helps AI provide smarter and more reliable answers. It’s like we’re solving a difficult problem. Just as you find answers through Internet searches when faced with a problem, AI can also use Google searches to answer users’ questions based on more accurate and up-to-date information.
The challenge of Perplexity AI: A new era of search-based AI services
To understand the “Grounding with Google Search” feature, first A service called Perplexity AIYou need to look at . Perplexity AI is an AI service that analyzes Internet search results and provides answers when users enter questions. For example, if you ask “What’s the weather like today?”, Perplexity AI will search weather websites and provide you with a summary of the weather information for your current location. In this way, Perplexity AI is receiving a lot of attention by presenting a new search method that breaks away from the existing keyword-based search and uses AI to directly find information and provide answers.
Google’s response: Expansion of the search powerhouse into the AI field
In response to the emergence of search-based AI services such as Perplexity AI, Google has chosen a strategy of integrating its powerful search engine into its AI model. The “Grounding with Google Search” feature is part of this strategy, combining Google’s outstanding search capabilities with an AI model called Gemini. The Gemini model can provide more accurate and rich answers to users’ questions by accessing real-time information through Google search. For example, if you ask “Who won Wimbledon this year?”, the Gemini model can retrieve the latest Wimbledon match results and provide accurate winner information.
Benefits of “Grounding with Google Search”: Increased transparency and trust
The “Grounding with Google Search” feature does more than just improve the accuracy of your answers. This feature also plays an important role in increasing users’ trust by clearly presenting the basis for the AI model’s answers. The Gemini model provides answers along with links to the websites they are based on and suggestions for related search terms. Users can directly check the source of the answer through the provided link and determine the reliability of the information. Additionally, it can be expected to reduce the “hallucination” phenomenon in which AI models generate incorrect information.
How the “Grounding with Google Search” feature works: Synergy between Google Search and AI
The “Grounding with Google Search” feature works by collecting relevant information about the user’s question through Google search, passing it to the Gemini model, and using it to generate an answer. The ‘dynamic retrieval’ feature allows developers to activate the evidence feature only for specific questions. For example, you don’t need to apply the Evidence feature to questions that require a subjective answer, such as “How are you feeling today?”, but you do need to apply the Evidence feature to questions that require a factual answer, such as “Who will be the 2024 Wimbledon men’s singles winner?” You can enable the feature to provide more accurate and reliable answers.
The future of Google: Innovation through the convergence of search and AI
The “Grounding with Google Search” feature is interpreted as part of Google’s efforts to not only maintain its position as a leader in the search market, but also establish itself as a leader in the AI field. Google seeks to innovate user experience and open new possibilities through the convergence of search and AI. In conclusion, the “Grounding with Google Search” feature is an important advancement that helps AI become smarter and more trusted. This feature is expected to allow Google to further develop its search services while maintaining its leading position in the AI field.
