Now Reading
Hands-On with Google Search’s AI-Powered Generative Feature

Hands-On with Google Search’s AI-Powered Generative Feature

AI

In a world driven by information and instant solutions, Google Search has long been the go-to platform for finding answers to our questions. From simple inquiries to complex research, Google has provided us with a wealth of knowledge at our fingertips. Now, Google is taking its search capabilities to the next level with the introduction of its new generative AI feature. In this hands-on review, we explore the functionality, benefits, and limitations of Google’s Search Generative Experience (SGE).

Google’s SGE aims to revolutionize the way we interact with search results. Unlike the traditional search results page, the SGE infuses the search box with conversational functionality, allowing users to engage with a chatbot-like interface. By encouraging users to dive deeper into topics and ask more challenging questions, Google hopes to provide a more intuitive and comprehensive search experience.

To access the SGE, users can sign up for Google’s Search Labs. Once granted access, the search box appears unchanged. However, when a query is entered, a new section takes up a significant portion of the screen, pushing down the conventional list of links. Within this area, Google’s large language models generate a paragraph or two of information, similar to what you might find from popular chatbot models like ChatGPT or Microsoft’s Bing Chat. Buttons at the bottom of the section lead to the chatbot interface, where users can ask follow-up questions.

While the concept behind Google’s generative AI feature is promising, the execution falls short in terms of speed and efficiency. In tests conducted by WIRED, the AI-generated responses took an average of three seconds to appear, compared to the nearly instantaneous display of traditional search results. Although Google has made efforts to optimize the speed of its AI software, the sluggishness of the generative feature remains a significant drawback.

Cathy Edwards, a Google Search vice president, acknowledges the need for ongoing speed optimizations. Despite recent updates that claim to have doubled the AI’s speed, there is still room for improvement. The delay in response time often leads users to lose interest and disregard the generative AI’s output, instead opting to rely on the faster traditional results.

Speed is not the only concern when it comes to Google’s generative AI feature. The accuracy and reliability of the AI-generated responses also leave much to be desired. In the case of queries regarding the number of stamps required to mail an 8-ounce letter, the generative AI provided a response that included apparent errors in multiplication and subtraction. Furthermore, the suggested follow-up questions failed to consider crucial variables such as shape, size, and destination when calculating shipping costs.

Google acknowledges the experimental nature of its generative AI and warns users of potential variations in information quality. This disclaimer is particularly relevant when considering the conflicting stamp price recommendations provided by the AI. Navigating to the official US Postal Service’s online calculator proved that the generative AI’s suggested values were outdated and inaccurate.

Despite its shortcomings, Google’s generative AI feature represents an exciting frontier in search technology. The ability to engage in a conversation with a search engine opens up new possibilities for information retrieval and exploration. While the current limitations hinder its effectiveness, the potential for improvement and refinement is vast.

According to Google’s Edwards, the technology is still in its early stages, and pushing its boundaries is part of the ongoing development process. As the AI continues to learn and adapt, its accuracy, speed, and comprehension of complex queries are expected to improve. The generative AI feature may not yet be a replacement for traditional search results, but it certainly has the potential to enhance the search experience for users.

Another aspect worth considering is the balance between simplifying information for general audiences and catering to specialized knowledge. In some cases, the generative AI feature struggled to provide accurate and up-to-date information when asked for specific details. For instance, when inquiring about the price of a stamp, the AI responded with an outdated figure. Only by specifying the desired information as of the current month did the system accurately reflect the recent 3-cent cost hike.

While the training data limitations of models like ChatGPT may explain similar shortcomings, it is important to recognize that generative AI features like Google’s SGE are not intended to replace search engines. Instead, they aim to augment the search experience by providing more conversational and personalized responses. Striking the right balance between simplicity and specialization remains a challenge for AI-powered search features.

Context is a critical component of any search experience, and it is an area where the generative AI feature currently falls short. When users pose complex queries, the AI-generated responses often lack the necessary context to provide accurate and relevant information. This limitation emphasizes the need for ongoing advancements in natural language processing and understanding.

See Also
Google Content

Google’s Edwards acknowledges the importance of context and states that further improvements in the AI software powering the generative feature are underway. As the technology evolves, users can expect a more contextualized and precise search experience that caters to their individual needs.

Google’s Search Generative Experience represents a significant step forward in the evolution of search technology. While the current implementation has its limitations, the potential for more intuitive, conversational search interactions is undeniable. The sluggishness and inaccuracies of the generative AI feature are challenges that Google is actively working to address.

As the AI software undergoes speed optimizations and improvements in information accuracy, the generative feature has the potential to provide users with a more comprehensive and personalized search experience. By enabling deeper dives into topics and encouraging challenging questions, Google aims to empower users with knowledge and insights that go beyond traditional search results.

While the generative AI feature may not yet rival the speed and precision of conventional search results, it represents an exciting frontier in the quest for more intelligent and conversational search experiences. As Google continues to refine its AI technologies, the future of search looks increasingly conversational, informative, and personalized.

FAQs

  1. What is Google’s Search Generative Experience (SGE)?
    Google’s SGE is a new generative AI feature within Google Search that aims to provide a more conversational and intuitive search experience. It allows users to engage with a chatbot-like interface and ask more challenging and comprehensive questions.
  2. How do I access Google’s SGE?
    To access Google’s SGE, users can sign up for Google’s Search Labs. Once granted access, the search box appears unchanged, but a new section with AI-generated information appears below the traditional search results.
  3. Is Google’s SGE faster than traditional search results?
    No, the generative AI feature is slower than traditional search results. In tests, the AI-generated responses took an average of three seconds to appear, compared to the nearly instantaneous display of conventional search results.
  4. How accurate are the AI-generated responses in Google’s SGE?
    The accuracy of the AI-generated responses varies. In some cases, the generative AI provided inaccurate information, outdated prices, and failed to consider crucial variables when answering queries. Google acknowledges that the generative AI is experimental, and information quality may vary.
  5. Will Google’s SGE replace traditional search results?
    Google’s SGE is not intended to replace traditional search results. It aims to enhance the search experience by providing more conversational and personalized responses. The generative AI feature is still in its early stages, and further improvements are expected to make it more accurate and efficient.
  6. What are the limitations of Google’s generative AI feature?
    The generative AI feature currently lacks speed, accuracy, and contextual understanding. The responses can be slow, inaccurate, and may not consider all relevant variables. However, Google is actively working to address these limitations and improve the overall search experience.
  7. How does Google plan to improve the generative AI feature?
    Google is continuously optimizing the AI software underlying the generative feature to improve its speed, accuracy, and contextual understanding. As the technology evolves, users can expect a more precise and personalized search experience.
  8. Can the generative AI feature cater to specialized knowledge?
    While the generative AI feature aims to provide more conversational responses, it may struggle to cater to highly specialized knowledge. The balance between simplifying information and providing precise details remains a challenge for AI-powered search features.
  9. What is the future of search technology?
    The future of search technology is expected to be more conversational, intuitive, and personalized. As AI technologies continue to advance, users can anticipate a search experience that goes beyond traditional results, empowering them with deeper insights and knowledge.
  10. Is the generative AI feature available across all search queries?
    Currently, the generative AI feature is still in its experimental stage and may not be available for all search queries. Google is actively developing and refining the feature to expand its capabilities and make it more accessible to a wide range of queries.

First reported by WIRED.

Scroll To Top