AI in Search: The Overview

0 Shares

The overview

Search is changing…

Here at Redefine, we are invested in all things search and Google. As new technologies emerge (mainly AI), we hear companies everywhere talk about how they are using “generative AI,” or “LLMS.” “AI integration” and more. These propositions for new AI tech play an important role in the future of the internet as Google has entered the AI innovation race. Google has created its own AI to compete with the infamous ChatGPT; for a while, this AI was a separate offering to “Google Search,” but as of May, Google is implementing its AI Gemini into the search engine itself. 

You may have seen some search results that look a little different recently with “AI overview” on the top of the SERP. This is the beginning of Google integrating AI into search. According to Google, “With generative AI, Search can do more than you ever imagined.” As marketers working with organic search and studying how people seek answers, we have to ask ourselves critical questions: Is AI the future of search? Will AI become the primary way we answer questions? Are the days of pulling up old blogs, ancient Reddit threads, or 15-year-old YouTube videos to find answers coming to an end?

Google believes that “With expanded AI Overviews, more planning and research capabilities, and AI-organized search results, our custom Gemini model can take the legwork out of searching.” Google aims for Gemini to become the primary search tool, making its LLM the main provider of information. However, this shift raises concerns that the answers people find on the internet could become less human and reduced to lackluster summaries of what were once thought-out pieces of content. Is this the future of how we get our information? 

The preview

People have always needed a way to find information. Before Google, if you wanted to find answers, you’d have to pour through books at your local library, sift through a stack of encyclopedias, or find a trustworthy source to confide in. The internet revolutionized how humans search, reducing our need for physical discovery and providing infinite answers to infinite questions at our fingertips. Is AI merely the next step in this evolution, or does it ultimately make our search for answers “less human”? 

AI is being integrated into just about everything we use on the internet. Most online tools now have AI features, social media companies like Instagram are building AI into how we consume content, and even design companies like Figma are using AI to make designing anything faster. 

We are on the verge of a major shift in how people search. Google’s updates to the browser will change the way we find information. Whether these changes are positive or negative is up to the user to decide. However, with the early implementation of the AI overview, we’ve already seen some missteps that deeply affect people’s trust in the tool.

What is Google’s AI Overview?

Just over a year ago, OpenAI released ChatGPT, causing a significant stir. In just two months, the chatbot grew to over 100 million active users, making it the fastest-growing app globally and triggering a competitive surge in AI development.

Google has been working hard to join the race and pushed competitively to get Gemini to the public. ChatGPT poses a huge threat to many tech companies as they surge ahead to be the global leader in AI. Google’s  AI overview uses Gemini to pull from across the internet and generate a SERP that combines blogs, articles, videos, social posts and code. This new type of SERP is supposed to give users an overview of the answer they are looking for. Featured Snippets will no longer be the main summary of an answer for users; rather, it will be a summary generated by Gemini. 

The bigger picture of this is that the AI is supposed to have in-depth responses to different tones of voice for different users, with all the capabilities for complicated searches tailored to you. Google wants a search engine that can give unique answers based on the needs/wants of the user. 

AI Overviews will begin rolling out to everyone in the U.S., with more countries coming soon. That means that this week, hundreds of millions of users will have access to AI Overviews, and we expect to bring them to over a billion people by the end of the year.

Implications on Google Search

Featured snippets are less useful.

The introduction of the AI Overview search results may render traditional featured snippets less useful. These snippets, designed to provide quick answers to user queries, could be overshadowed by more responses generated by AI. We are already seeing the featured snippets section of a page being pushed much farther down on the SERP, resulting in a huge loss of CTR

#1 ranks” may no longer exist.

The concept of “#1 ranks” in search engine results may become a thing of the past. With AI organizing search results, the traditional ranking system could be replaced by more dynamic and personalized responses. Instead of a single top result, users may continue to receive a curated set of answers tailored to their specific needs and context. This shift could fundamentally change how SEO strategies are developed and executed. 

Using video and other media to rank may be more prominent.

If the AI overview continues to be prominent on Google, using video and other media to rank in search results could become more useful. Gemini’s ability to analyze and understand multimedia content means that videos, images, and interactive elements could play a larger role in providing relevant answers to user queries. This shift would encourage content creators to diversify their formats and leverage multimedia to improve visibility and engagement.

Although many of these implications are true and may have some positive aspects, the rollout of this AI overview was anything but smooth. 

Will AI ruin search?

When AI overview was first officially launched, some major issues occurred across Google. With answers that didn’t make sense, false and even harmful SERP answers, people were looking to turn off the new search completely. To make matters worse, there were a large number of fake screenshots going around, making the issues with the AI overview seem even worse than what was actually happening. 

So what did happen? 

According to Google, “some odd, inaccurate or unhelpful AI Overviews certainly did show up. And while these were generally for queries that people don’t commonly do, it highlighted some specific areas that we needed to improve.

One area we identified was our ability to interpret nonsensical queries and satirical content. Let’s take a look at an example: “How many rocks should I eat?” Prior to these screenshots going viral, practically no one asked Google that question.” 

The example above was seldom searched on Google. The scarcity of substantial web content on this topic led to a “data void” or “information gap.” However, satirical content about it was republished on a geological software provider’s website. As a result, when this query was searched, an AI Overview linked to one of the few sites covering it.

Issues like this could become more prominent, especially during experimental rollout phases. However, as search experts, we remain skeptical and concerned about the persistent issues of accuracy and bias due to the vast range of opinions and incorrect answers. Furthermore, there are challenges related to the infrastructure and the significant expenses required to maintain these AI systems. A recent 31-page report from Goldman Sachs raises the question of whether the investment in these generative AI systems is worth the benefit. Many are skeptical, viewing these AI systems as part of a massive tech bubble similar to the dot-com bubble of the early 2000s.

AI will get better, and the tech will evolve, but do we truly need it? It is possible search will be improved and give us better access to information, but ultimately, the negatives could outweigh the positives and lead to a broken system of searching. 

Positives for AI in search

Fast Answers: One of the most significant advantages of AI in search is the speed at which it can provide answers. AI can quickly sift through vast amounts of data to find relevant information, delivering instant results that save users time and effort.

Less Bias: AI has the potential to reduce bias in search results. Traditional search algorithms can sometimes favor specific sources or types of content. By integrating AI, Google aims to provide more balanced and impartial search results by analyzing a broader range of data.

More In-Depth SERP: AI integration can enhance the depth of search engine results pages (SERPs). Instead of simple links to websites, AI can offer comprehensive overviews, combining information from multiple sources into a single, detailed response.

Combines Multiple Sources: AI-driven search can pull information from various types of media, including blogs, articles, videos, social posts, and code. This ability to amalgamate different content types provides users with a richer and more diverse set of answers.

Negatives for AI in search

Less Human Experience: While AI can provide fast and detailed answers, it can also make search feel less human. Responses may become more robotic and stiff, lacking the nuance and personality that come from human-written content.

Energy Inefficiency: AI-driven search is significantly less energy efficient. According to recent studies, Google’s AI search summaries use ten times more energy than traditional Google searches. This increased energy consumption raises concerns about the environmental impact of widespread AI integration.

Potential for Wrong Answers: AI can still provide incorrect answers. The technology is not infallible and can misinterpret queries or pull inaccurate information, leading to misleading or wrong responses.

Clunky and Inaccurate: AI search technology is still evolving, and its early implementations can be clunky and inaccurate. Users may encounter awkward phrasing, irrelevant information, or poorly synthesized summaries, which can detract from the search experience.

User Resistance: Many users are skeptical or outright dislike AI-integrated search. They may find it less intuitive and trustworthy than traditional search methods, leading to reluctance in adopting the new technology.

As AI continues to develop, these positives and negatives will shape how search evolves. While there is potential for significant improvements, it’s crucial to address the drawbacks to ensure AI enhances rather than detracts from the user experience.

Is it too much too fast? 

The internet was fascinating because it gave people access to information in a way that had never been done before: books, movies, and articles were now at our fingertips in a way many didn’t expect. With the dot-com boom, there was excitement for what was new, leading companies to prioritize being “online” and caring much less for what people actually wanted.

Are we pushing too far, too fast, when it comes to relying on AI to answer our questions? The intense focus on AI can pressure teams to release features without a clear use case. As Conor Woods, Product Manager at Figma, notes, “AI is a hammer, and everybody is looking for nails.” Instead of using AI for the sake of it, teams should identify real problems and determine if AI is the best solution. The risk of over-reliance on AI is that it may strip away the human element that makes the Internet a rich platform for genuine expression and discovery. 

In their report, Goldman Sachs highlights the significant costs and infrastructure challenges associated with AI, raising questions about whether the investment in generative AI systems is justified. They argue that the current AI architecture isn’t designed to solve the complex problems necessary to justify its high costs. This sentiment echoes the skepticism surrounding the rapid integration of AI, as many fear it could lead to a less human, more mechanical internet​.

The internet is built upon human thoughts and ideas. While making a robot summarize those ideas could enhance the experience, it could also diminish what we liked about the internet in the first place: a space where real people share their genuine thoughts and ideas.

Something to remember

AI in search may be the next step in how we use search and find answers. However, as users, we want to ensure that the product is ready and that the negatives don’t outweigh the positives. Innovation will continue to move forward, and we have to adapt, but it is wise to be cautious of rapid changes driven by the race to the top of AI advancement. Ensuring that AI enhances rather than detracts from the search experience is crucial as we navigate this technological evolution.

Author avatar
James Taylor
James is a marketer, writer, and branding enthusiast. He is a Cal Poly Pomona Alum who has a passion for helping others grow their business. In his free time he loves to relax with a nice book and a campfire.
0 Shares
Share via
Copy link