Google’s AI Overviews have undergone a remarkable transformation from laughingstock to legitimate search tool in just over a year, according to a candid assessment from a journalist who famously tested the feature’s early failures. When Google AI Overviews launched in spring 2024, the feature quickly became infamous for providing absurd and dangerous recommendations, including the viral suggestion to add glue to pizza sauce to prevent cheese from sliding off—advice apparently scraped from a Reddit joke.
Business Insider’s Katie Notopoulos, who gained internet notoriety by actually making and eating a glue pizza to test the AI’s recommendation, now admits that she regularly uses AI Overviews instead of traditional search results. This shift represents a significant evolution in both the technology’s capabilities and user acceptance. The early days of AI Overviews were plagued by what became known as the “you can’t lick a badger twice” problem—where users discovered they could type any nonsensical phrase followed by “meaning” and receive confident AI explanations of non-existent idioms.
Testing this vulnerability in April 2025, Notopoulos created fake phrases like “you can’t fit a duck in a pencil” and “the road is full of salsa,” and the AI dutifully provided explanations as if these were established sayings. Google spokeswoman Meghann Farnsworth explained at the time that the AI systems were attempting to provide helpful results even when faced with “nonsensical or ‘false premise’ searches,” acknowledging the limitations when users intentionally tried to game the system.
The improvement has been substantial. When recently testing the same type of fake idiom (“you can’t tell a yak not to dance”), the AI provided a much more accurate response, correctly identifying it as “not a common, established idiom, but rather a playful or poetic expression” while offering potential interpretations. This demonstrates Google’s refinement of its AI systems to better distinguish between legitimate queries and nonsense.
The article raises important concerns about the broader implications for content creators and publishers. As users increasingly rely on AI-generated summaries rather than clicking through to original sources, websites that employ human writers face reduced traffic and revenue. Despite these ethical concerns, Notopoulos acknowledges that AI Overviews have become genuinely useful, with users developing better intuition about which queries are suitable for AI-generated answers.
Key Quotes
When people do nonsensical or ‘false premise’ searches, our systems will try to find the most relevant results based on the limited web content available. This is true of Search overall — and in some cases, AI Overviews will also trigger in an effort to provide helpful context.
Google spokeswoman Meghann Farnsworth provided this explanation in April 2025 when addressing the AI’s tendency to generate confident explanations for made-up idioms. This quote reveals Google’s acknowledgment of the AI’s limitations while defending its attempt to be helpful even with problematic queries.
Google AI Overviews are getting less laughably bad, and just, well … useful. I’ve gotten more used to seeing them, and as I get used to them, I’ve also gotten better at predicting if the type of query I’m posing is the kind of thing an AI Overview can answer.
Katie Notopoulos, the journalist who famously ate glue pizza to test early AI recommendations, admits to the feature’s improvement and her own behavioral adaptation. This reflects how users are learning to work with AI tools even as the technology improves, representing a two-way evolution in the human-AI relationship.
Of course, the fact that people (such as myself) are willing to use the AI Overview answer in Google instead of clicking on a link and sending traffic to a website that perhaps employs humans to make that content (such as myself) is not necessarily great.
Notopoulos acknowledges the ethical dilemma of using AI Overviews as a content creator herself. This self-aware observation highlights the tension between user convenience and the economic sustainability of human-created content, a central concern as AI increasingly mediates information access.
Our Take
This honest assessment from a former AI skeptic provides valuable insight into how quickly user behavior adapts to improving AI technology. The normalization of AI Overviews despite their catastrophic early failures demonstrates remarkable user forgiveness when technology shows consistent improvement. What’s particularly striking is the journalist’s self-awareness about the economic implications—she’s essentially admitting to using a tool that undermines her own profession.
This represents a broader pattern in AI adoption: initial resistance gives way to pragmatic acceptance as capabilities improve and users develop better mental models for when AI works well. Google’s strategy of rapid iteration despite public embarrassment appears vindicated, though the long-term consequences for web publishers remain deeply concerning. The real question isn’t whether AI Overviews have improved, but whether the traditional content ecosystem can survive their success.
Why This Matters
This evolution of Google AI Overviews represents a critical inflection point in how billions of people access information online. The shift from unreliable, meme-worthy failures to functional search tools demonstrates the rapid pace of AI improvement and adaptation. For the broader AI industry, this validates the approach of launching imperfect products and iterating based on real-world feedback, even when early failures generate negative publicity.
The implications for content creators and publishers are profound. As AI Overviews become more reliable and users increasingly accept AI-generated summaries over clicking through to original sources, the traditional web traffic model faces existential threats. This could fundamentally reshape online publishing economics, potentially reducing revenue for human-created content while concentrating power with AI platform providers.
For businesses and marketers, understanding when and how AI Overviews surface information becomes crucial for SEO strategy. The technology’s maturation also signals broader consumer acceptance of AI-mediated information, suggesting that similar AI features across other platforms will face less resistance. This normalization of AI as an information intermediary represents a fundamental shift in human-computer interaction that will influence product design, content strategy, and digital literacy for years to come.
Related Stories
- Google’s ‘Ask for Me’ AI Phone Tool: A Game-Changer for Time Management
- Reddit Sues AI Company Perplexity Over Industrial-Scale Scraping
- Google’s Sundar Pichai Responds to Microsoft CEO Satya Nadella’s ‘Dance’ Comment on AI in 2024
- Google Founders’ $511B Fortune Soars on AI Breakthrough Success
Source: https://www.businessinsider.com/google-ai-overviews-search-answers-improved-why-explained-2025-12