US flag
  • US flag +1 646 844-80-75

The New AI-Powered Google Search: What to Expect?

Zlata Pustovalova
Zlata Pustovalova
May 19, 2023 | 11 min. reading

Google is ready to take a new step in rethinking all current products, including search, by implementing AI’s most recent practices. This big news was officially announced at the annual Google I/O.

But what exactly can users anticipate? And how will these updates impact the overall search experience? Let’s delve into the details!

Introducing the New Google Search Experience

The Interface

The revamped Google search interface includes an AI-generated response positioned above the search results. To ensure transparency, Google explicitly identifies this response as an experimental feature. 

The AI-generated answer is neatly boxed within the interface, accompanied by a list of the sources used by Google to generate it. Clicking on these sources allows you to delve further into the provided information. Additionally, you have the option to pose more questions or explore deeper insights by toggling the button located in the top right corner.

Within the AI-generated responses, Google presents clickable boxes accompanied by images from relevant websites. Users can seamlessly navigate to these websites for more exploration.

“As for the fate of SEO? Changes in the industry and the daily work of SEO specialists will occur. In my opinion, this is not a debatable issue. An extended snippet that occupies the entire first screen when it appears in the SERP will crash the CTR of pages in the TOP, as it has happened more than once. Over time, the SEO community will find patterns for hitting the AI ​​snippet and will start optimizing for AI. Also, changes may affect the types of content that will be created for businesses. Everything will depend on the niche and market.”

Zlata Pustovalova, COO Livepage

Also, Google has implemented a catchy visual feature where the color of the generative AI answer box dynamically adjusts to reflect specific journey types and the intent behind your query. As stated by Google, this enhancement adds an extra layer of context and engagement to the search experience.

AI-Driven Vertical Search

The application of AI-powered search extends beyond traditional queries and even integrates with vertical search experiences, like Google Shopping results. Drawing upon the Google Shopping Graph, Google’s Search Experience Generation (SGE) easily gathers 35 billion product listings. More than this, it receives a staggering 1.8 billion updates per hour.

To cater to the dynamic nature of these vertical search experiences, the generative AI undergoes rapid, near-real-time updates. This ensures that users receive up-to-date and relevant answers promptly. When searching for specific types of products, Google excels at delivering reliable recommendations, helping users make informed decisions.

Dynamic Conversations Engagement

Engaging in dynamic conversations with Google’s enhanced search experience is now easier than ever. Simply provide additional details in the “Ask a follow-up” box to refine your query, and Google will quickly generate a fitting answer.

This conversational mode is also valuable when seeking more intricate and evolving information journeys. Leveraging advanced AI capabilities, Google understands when a user is searching for information related to a previous question. By preserving the context from prior inquiries, Google reformulates subsequent queries to better reflect the user’s intent, fostering a more effective and personalized search experience.

How Does the New Google Search Work?

Harnessing Cutting-Edge Technology

Google’s latest search experience leverages a diverse range of Language Model Mixtures (LLMs) to achieve its remarkable capabilities. Among these LLMs are MUM and PaLM2, which are slightly better than popular GPT4 in terms of translations and mathematical tasks, demonstrating the depth and breadth of technological advancements of the updated system

This search experience is meticulously developed and trained to identify and present high-quality web results that align with and support the information presented in the output. By combining state-of-the-art LLMs and focused training methodologies, Google has created a search experience that delivers reliable and corroborated information.

Focusing on Safety and Reliable Information

Similar to Google’s commitment to preventing the display of harmful, offensive, or explicit content in its ranking systems, the Search Experience Generation (SGE) is designed with the same objective in mind. 

The emphasis lies in generating informative responses corroborated by reliable sources, ensuring that users receive trustworthy and verified information. In cases where Google’s systems have lower confidence in the source or encounter information gaps, the aim is to avoid presenting AI-generated snapshots. Furthermore, explicit or dangerous topics are deliberately avoided to maintain a responsible and secure search environment.

According to Liz Reid, VP of Search at Google, the new version of Google Search exercises caution by not providing answers to every query. Google’s emphasis is on ensuring a safer search experience, particularly for sensitive inquiries.

For instance, Google refrains from displaying answers related to medical matters and queries in the financial domain. This cautious approach aligns with the concept of YMYL (Your Money, Your Life), an expanded category that now encompasses civic information.

Balancing Trust and Accuracy

Liz Reid highlighted that there is a preference among users for answers presented in a more fluid style. However, Google prioritizes providing factual responses over fluidity, as it instills greater trust among users.

In the context of generative AI, avoiding false or inaccurate information is of paramount importance to Google, particularly when addressing YMYL topics such as health and finance. AI hallucinations, confident but wrong responses given by an AI, are a concern in generative technology and are carefully mitigated to ensure the utmost accuracy and reliability of the information presented.

Google also crafted the conversationality of the Search Experience Generation (SGE). It offers factual responses supplemented with relevant resource references. This deliberate approach guarantees a search experience that is not only informative but also maintains the trust and confidence of its users.

Google’s Comprehensive Approach to Generative AI in Search

Google adopts a five-point approach to enhance generative AI within its search ecosystem:

  1. Information needs. Google strives to streamline the search experience by minimizing the number of steps required for users to get the right response that helps to accomplish their tasks. 
  2. Information quality. Emphasizing the importance of reliable information, Google ensures that the responses provided by AI are of high quality. Particularly in health and financial domains, Google takes a cautious approach when determining the appropriateness of answering queries, prioritizing accuracy and credibility.
  3. Safety constraints. Google conscientiously evaluates whether to provide first-person responses and considers the challenge of balancing users’ trust in fluid answers with the accuracy verification limitations. Also, Google takes into account the potential risks associated with providing information without full verification.
  4. Ecosystem. Google aims to foster a thriving content ecosystem by acknowledging and crediting the sources that contribute to the search experience. By providing traffic and credits to these sources, Google encourages users and searchers to explore their content.
  5. Ads. Google seeks to deliver relevant ads that provide additional value to users. Careful consideration is given to optimizing the presentation of ads, ensuring they offer relevant information in a non-intrusive manner.

What about Citations and Links?

In the new search experience, explicit answers are composed of information extracted from specific websites. Notably, these websites are prominently displayed within the response, complete with a clickable thumbnail image, title, and URL, leading users directly to the publisher’s website.

Yet, Google does not directly cite or attribute individual pages. Instead, its AI model synthesizes information from a diverse range of sources. To ensure factual accuracy, Google seeks corroborating evidence across multiple sources before presenting the answer and associated citations. These citations predominantly originate from reputable online sources.

Publisher Site Links

To provide a visual depiction of the improved approach, here is a screenshot showcasing the inclusion of publisher websites within the answer:

Toggle deeper

You can also click, at the top right, on the toggle button to do a deeper dive into more sources shown by the generative AI.

With search results below

Furthermore, you can scroll down to view traditional search results presented in a more concise and easily digestible format.

What does this all mean?

At Livepage, we always strive to be on the wave with the latest updates of everything that impacts business promotion. And today, we wanted to tell you about the new Google Search era. This post covers the most crucial questions. 

To conclude, there is nothing to add but to wait for the real possibility of seizing all the new Google Search enhancements on our own. They are expected to change the well-known search experience for the better by making it more effective, personalized, and transparent.

“Google presented a new issue model, and this is not Bard in its purest form but a really new Google user interaction model.

What conclusions can be drawn from what you see?

Google itself becomes a full-fledged site, which, in the case of the issuance of AI, will monitor whether the answer of Google itself was useful, or whether the person followed the link of the publisher, or made another request. 

– First of all, sites such as Wikipedia and publishers with evergreen content will suffer. 

Google never gets into the topic of Your Money, Your Life. That is, he does not take responsibility for health and money. 

In general, all gray niches seem to remain safe from AI. AI follows the principle — everything that is not added to Google Ads is not shown by AI.

Google will always try to look for trusted resources and show facts rather than some dynamic and time-changing information.”

Mikhail Shcherbachev, CEO Livepage

And what about you? Are you ready for the release? Feel free to share your thoughts in the comments!

Leave a comment

Your email address will not be published. Required fields are marked *