Only days after its launch, Google AI Overview has been making errors that we’re embarrassed to have to report on this fine day. During the Google I/O event on May 14, 2024, the company globally released its AI Overview feature to the masses, taking it out of beta mode and letting the population have at its capabilities. Almost instantly, people began to find issues with Google AI Overviews for multiple reasons, its misinformative tendencies being one of the primary reasons. The AI is good at summarizing information the same way Amazon reviews have an AI-generated recap on top, but these AI tools are far from being experts on the legitimacy of the information it’s summarizing.

The many Google AI Overview criticisms are well-warranted and we have to be significantly more wary of the content the AI bot is putting out. Speaking to The Verge, a Google spokesperson said they were “isolated examples” of “generally very uncommon queries, and aren’t representative of most people’s experiences,” but even isolated issues can hint at a broader problem that the company does not want to acknowledge. 

Google AI Overview errors

Image: Pexels

Google AI Overview Errors Result In Dismal Results

Before we understand the extensive list of Google AI Overview errors, it might help to understand the new feature first. Google’s latest AI product is integrated with the company’s most popular tool—Google Search. The AI Overview feature appears at the top of the results page once you search for a specific topic. The tool summarizes the key points from various articles and presents a consolidated brief to answer the question you had in mind. Some of the results pull from a single source while others consolidate web page results to generate answers. 

In both situations, the result depends on other people’s content—a fact that has become upsetting for many. There is no way to predict which search will get an overview, and this unpredictability makes it even harder to test. Unfortunately, this isn’t the biggest issue with the Google AI Overview tool.

Google’s AI Overview Changes the Internet as We Know It

Google AI Overview gathers information from multiple sources and saves you the trouble of having to open each link and find the answers yourself. Immediately, you might begin to see where the Google AI Overview controversy begins. As soon as people heard CEO Sundar Pichai announce the AI tool, they began to consider the implications of what it would mean for their own websites. If you’ve spent hours building your expertise on a topic, carefully sectioning the information down into digestible chunks for the consumption of your reader, seeing Google present the results without a reader ever having visited your website can be painful. 

Even before we get to the problem of misinformation, the Google AI Overview issues keep a reader from ever opening up your website after all that effort you spent on understanding SEO optimization tactics—another technological evolution that has only devalued the internet rather than helped it. In response to the arrival of this tool, BBC said “The Internet will never be the same,” and we have to agree, we’re witnessing a whole new era of browsing. 

Not only does the Overview feature hurt the writers and creators of the posts by stealing their views, but it also hurts the readers just as easily. Instead of reading words written by a person, readers could soon become more familiar with the lifeless writings of AI. Our attention spans have already shrunk considerably thanks to all the hours spent on social media, and now Overviews is set to make it easier to read a few lines and have your question answered rather than reading a whole article about the “whats” and the “whys.” 

Despite the controversy around Google AI Overview, you may find it easier to get an answer now that the top results will be summarized to address your query, but you could also end up with more than you bargained for.

Google AI overview criticism

Image: Google AI Overview

Google AI’s Misinformation Saga Grows More Concerning

The discussion regarding search results is only part of the controversy.  The AI draws from information that is already available on the internet and makes some inferences to connect the dots. Considering not all the information we have online is reliable because people love to make things up, Google AI’s misinformation era was perhaps inevitable. Satirical articles, conspiracy theories, Reddit commenters, genuine debates, and blatantly incorrect information can all be found across the internet for anyone to access.

Depending on the platform where we find the information, we usually make an informed decision on whether to trust the information or not. Unlike a typical chatbot, the Google Search AI does provide a few links that indicate what posts were referenced or what you could read for additional information, but will people remember to be as thorough with their own search every time? That remains to be seen. 

A user looking for “cheese not sticking to pizza” was told by Google to add “about ⅛ cup of nontoxic glue to the sauce.” Another user who asked how many rocks they should eat every day was told “According to UC Berkeley geologists, people should eat at least one small rock a day.” Not only was Google AI spreading misinformation, but it was also tagging reputable sources that one might glance at and see as legitimate. The variety of the Google AI Overview errors has resulted in more than one raised eyebrow.

There are reports that suggest that Google may use the AI Overviews to begin suggesting ads to users as well which shows how little the company cares about what its users want. If the Google AI overview criticisms are harsh now, they will only multiply when the company crowds actual search results out with its ads.

Google AI misinformation

Image: Google showcases how the AI Overview feature works

How Do We Feel about the Google AI Overview Issues?

 A lot of the stories we’ve seen so far feel humorous until you begin to consider how many results may proliferate when their content is not as obviously wrong. You might use incorrect statistics and facts in a research paper or just as easily waste a lot of ingredients making a recipe that isn’t someone’s great-grandmother’s most prized possession. A child who Googles quick facts may learn something incorrect or use the wrong information in their homework without ever realizing the data was quoted from someone on Reddit. These issues could arise with any search engine at any given time, but they are enhanced further when you let down your guard with the AI Overviews.

There are many examples of the AI’s incorrect results circulating the internet and some of these viral posts have also been accused of faking their results to make AI Overview seem more flawed than it really is. We agree that people may be up to no good trying to trick the AI, but there is sufficient reason to be concerned about the Google AI Overview’s errors as well.

There is some truth to the observation that all of these images may have been doctored but should we overlook Google AI Overview’s errors and let the issue go? That would be unwise. LLMs have never been reliable and they won’t be reliable in the next few years either. A lot of work is necessary to stride through the many challenges they face. The Google AI overview issues only serve as an example of the much larger problem of the obsession with AI. 

We have experimented with the Google AI Overview feature for a few months now, which was available through Google Labs as the company was testing its functionality. While we have not seen any major bits of misleading information make an appearance, we’ve also not found it to simplify our lives in any way when we have to look through the article to verify what the search results have to say. No matter how much the Google tool improves its services, we’ll still be apprehensive about its results, so we fail to see any actual use for the Google AI Overview tool, its controversy aside. We don’t need artificial intelligence as desperately as most tech giants seem to think we do. 

tagged in Google AI misinformation, Google AI overview controversy, Google AI overview criticism, Google AI Overview errors, Google AI overview issues
Share: