The Backlash to Google’s AI Search, Explained
Why Google Search telling people to eat rocks is just the tip of the iceberg that’s about to hit online media
Google recently introduced a transformative update to its search engine: AI-generated search answers, also known as AI Overviews. While Google hopes to streamline the search process by providing users with succinct AI-generated summaries from sources pulled in from all over the web, this significant change has sparked considerable backlash, raising concerns about the accuracy of AI responses, the impact on advertisers, and the broader implications for the digital ecosystem. Let’s take a closer look.
Google AI Recommends Eating Rocks
Despite the promise of efficiency, the rollout of AI Overviews, which is now available to all U.S. users, has been undermined by some hilariously eyebrow-raising errors. For instance, there have been reports of the AI suggesting using non-toxic glue to keep cheese on pizza, or to use gasoline when cooking spaghetti, and recommending that people consume rocks. Upon further investigation, it would seem that all these ridiculous AI suggestions were based on misinterpreted or satirical sources, such as a joke comment from Reddit or The Onion.
The public backlash against Google’s AI Overviews has been swift and intense, amplified by the numerous examples of social media posts of Google’s AI providing inaccurate or absurd answers. Besides drawing ridicule on social media, these examples have also raised serious concerns about the reliability of AI-generated information, to the point that some critics claim that Google’s AI search is worse than its regular search experience; in fact, Many articles have been published within the past few weeks advising people on how to turn it off.
For what it’s worth, Google has defended its AI search product, stating that the majority of queries are handled accurately and that the errors reported are outliers. As ridiculous as some of the answers are, most of them are so far beyond the realm of common sense that it seems doubtful how much, if any, real-life damage they could’ve caused.
Google also pointed out that the fact that some of the incorrect AI answers going viral has prompted some social media users to doctor the results to one-up each other with more outlandish answers. For example, one manipulated screenshot posted on Reddit appears to show Google saying that a good remedy for depression was jumping off the Golden Gate Bridge. Google claims that the company’s systems never returned that result, but acknowledged that it will continue to improve the accuracy of its AI search results.
That said, the fact that such outlandish answers now seem entirely plausible for users is a troubling sign that the wide rollout of AI Overviews may have been a bit premature. Perhaps, a limited rollout confined to certain types of search queries would have been a safer route. Google, of course, is just doing the best it can to stay ahead of the competition. But haste often makes waste, as the saying goes.
Impact on Brand Advertisers
One of the most significant aspects of Google’s AI Overviews is their integration with advertisements. Google has announced plans to incorporate ads within the AI-generated overviews, which would likely redefine the landscape of search advertising.
From a business perspective, the integration of AI Overviews and ads represents a strategic move for Google to maintain its dominance in the search advertising market. By adapting to the age of generative AI, Google aims to stay ahead of competitors like Bing and ChatGPT. However, as the controversy shows, the company clearly faces the challenge of balancing innovation with the need to ensure the accuracy and trustworthiness of its search results.
On one hand, this integration means that ads will be contextually placed within the AI summaries, potentially gaining higher relevance and seamlessness than the conventional “sponsored results” that sit atop the search results, which many users have now been trained to scroll right past without a glance.
On another hand, this change presents enormous challenges for brand advertisers that rely on search as a key media channel for customer acquisition. The shift to AI-generated summaries could affect the visibility and effectiveness of traditional search ads. Advertisers may need to adapt their strategies to fit into this new format, which could involve reevaluating their ad placements and messages to align with AI-generated content.
Google shared frustratingly few details about its new Overview ad format in its announcement. Ads will have the opportunity to appear within the AI Overview in a section clearly labeled as ‘sponsored’ when they’re relevant to both the query and the information in the AI Overview. AI Overview will draw on ads from advertisers’ existing campaigns, meaning they can neither completely opt out of the experiment nor have to adapt the settings and designs of their ads to appear in the feature. So far, it doesn’t seem like there’s anything that brands can do to avoid being included in AI Overview results, thus exposing brands to be potentially placed to harmful answers.
This AI Overview debacle also came at a delicate time for Google, as advertiser trust in Google has hit a new low in the wake of recent revelations from the DOJ antitrust trial, which demonstrated how Google has been systematically increasing costs for search advertisers.
Therefore, while advertisers might benefit from enhanced targeting and integration without additional effort, they must also navigate potential user skepticism and backlash towards ad placements within AI-generated answers. In response, Google will have to strike a delicate balance between maintaining its search ad revenue, as search transitions to a new AI-first interface, and maintaining a positive user experience. Failing to do so could lead to a decline in user engagement and further erosion of trust among both consumers and advertisers.
The “Managed Decline” of the Internet
Coined by analyst Casey Newton, the “managed decline” of the internet refers to the gradual deterioration of digital user experience and the internet’s open ecosystem, driven by the increasing dominance of a few major players, particularly Google and Meta, given their massive global user base. As Newton puts it in his Google I/O event report:
Over the past two and a half decades, Google extended itself into so many different parts of the web that it became synonymous with it. And now that LLMs promise to let users understand all that the web contains in real time, Google at last has what it needs to finish the job: replacing the web, in so many of the ways that matter, with itself.
As a result, the internet, once a vibrant and diverse ecosystem where anyone could create content and compete for visibility, is now increasingly controlled and shaped by large corporations that prioritize their own platforms and services. This is also often correlated to the ongoing “enshittification” of online services.
In this context, Google’s introduction of AI Overviews signifies a broader shift in how information is consumed and monetized online. This change has implications not only for advertisers but also for the overall ecosystem of content creators, publishers, and users.
Content creators and publishers are likely to experience a significant impact from AI Overviews. Traditional search results drive traffic to websites, supporting the digital advertising revenue model that many content creators rely on. With AI-generated summaries providing answers directly on the search page, the need for users to click through to individual websites may decrease, potentially reducing web traffic and ad revenue for publishers.
This reduction in traffic could disproportionately affect smaller websites and independent content creators who rely heavily on search engine visibility for their audience reach. Large, established brands might adapt more easily due to their existing recognition and resources, but smaller players could struggle to maintain their online presence and profitability.
Case in point — Casey Newton recently interviewed the owner of Retro Dodo, a small UK-based publication focusing on retro games. From 2020 to 2023, Retro Dodo enjoyed nearly 2 million pageviews a month, mostly driven by Google search. However, in September 2023, Google introduced a significant ranking algorithm update called the “helpful content update,” which aimed to reduce the visibility of content created primarily to rank high in search engines rather than to provide genuine value to users. This update led to a dramatic decline in traffic for many small, independent publishers, including Retro Dodo.
On top of that, Google’s newfound focus on pushing AI search also seems to leave little room for sites like Retro Dodo, as AI Overviews often pull content from sites like Retro Dodo, reducing the need for users to visit these sites directly. But if niche interest publishers like Retro Dodo can no longer survive in the new AI-driven era of the internet, then who will be there to supply the niche content to be regurgitated by AI?
No wonder publishers like News Corp., Atlantic, and Vox Media are now all signing licensing and product deals with OpenAI. If your whole business model is about to be upended, then you might as well make a lucrative deal with your №1 enemy’s biggest competitor.
And it’s not like the deal is any better for the users either. Putting aside the aforementioned inaccurate and potentially harmful AI answers, a deteriorated web is not going to inspire much user engagement either. The web is at its best when it inspires everyone to contribute to it — that’s why web 2.0 has been ruled by UGC and the algorithms that can serve them to the right audience. Yet, if the web is about to become a frictionless experience of being spoon-fed by AI-generated content, then what is fun in contributing to it?
In other words, the managed decline of the internet implies that users are being funneled into more passive roles, consuming pre-packaged information instead of actively engaging with diverse sources, making our own judgment calls, and participating in content creation via comments or replies. In the past, Google was telling you somebody else could answer your question. Now Google’s AI is doing the Googling for you and directly answering your question.
Moreover, the integration of ads within AI summaries could affect the user experience. While contextually relevant ads might enhance the search process, there is also the potential for users to feel overwhelmed or misled by commercial content interwoven with informational summaries.
In conclusion, Google’s controversial AI Overviews is part of a broader trend of managed decline in the internet ecosystem. The shift towards AI-driven search brings an unintended consequence of diminishing the diversity and richness of the web, making it harder for independent niche publishers to thrive. As more and more publishers turn to deals with AI giants like OpenAI to secure their future, the web risks becoming a more homogenized and passive space, where user engagement and the spirit of open contribution could be greatly diminished.
For brand advertisers, the integration of ads within AI-generated summaries presents both opportunities and challenges. While it offers enhanced targeting and relevance, it also necessitates adaptation to new formats and the potential for increased user skepticism. Google must carefully balance its innovative AI-first approach with the need to maintain trust and accuracy to prevent a decline in user engagement and advertiser confidence.
At the end of the day, however, whether AI search takes off or not depends heavily on just how useful AI Overviews will get. For most of the online searches, people are just looking for a quick answer, and if those answers are 98% accurate and seamless to get, then I imagine most users will unwittingly take the trade-off — the larger digital ecosystem be damned.
