Microsoft Bing is set to unveil a groundbreaking feature known as "Deep Search," leveraging the capabilities of OpenAI's GPT-4. This innovative addition aims to provide users with more relevant and comprehensive responses to intricate search queries, marking a significant enhancement to Bing's existing web search functionality.

Unlike a replacement, Deep Search is positioned as a complementary feature, allowing users to delve deeper into their search topics without altering the familiar Bing interface. According to Microsoft's official blog post, Deep Search builds upon Bing's current web index and ranking system, incorporating GPT-4 to transform search queries into detailed descriptions that guide the system towards more informative results.

For example, if a user seeks information on loyalty programs in Japan, a traditional search might yield basic results. However, Deep Search would expand the query to offer insights such as explaining various loyalty card programs in Japan, detailing benefits and limitations, presenting examples from diverse categories, and providing a comparison with alternative payment methods. In instances of ambiguous queries, Deep Search excels by identifying possible intents and presenting comprehensive descriptions for each. Users can then choose the most relevant intent, allowing Bing to pull relevant results that may not typically surface in conventional searches. Deep Search's prowess extends further by conducting multiple variations of a query, broadening the search spectrum. For instance, in the loyalty points example, it might explore variations like "best loyalty cards for travelers in Japan" or "managing loyalty points with phone apps," ensuring a more exhaustive coverage of the user's intent.

Microsoft emphasizes that Deep Search is an optional feature, as its comprehensive approach may take up to 30 seconds to complete. Users who prefer instantaneous results can continue to enjoy regular Bing search outcomes. Notably, Deep Search employs a refined ranking system, considering factors such as topic relevance, details, trustworthiness, and recency when presenting results to users. This advancement reflects Microsoft's commitment to delivering more informative and specific search results, surpassing the capabilities of conventional search methods.