Artificial intelligence tools are increasingly marketed as indispensable digital assistants, promising to effortlessly research, code, summarize, write, and unlock vast stores of knowledge. Yet, as many are discovering, sometimes these sophisticated systems can be hilariously, and worryingly, stumped by the simplest of questions. Google’s flagship AI Overviews feature recently provided a stark reminder of this, getting fundamentally confused about what year it actually is.
Over several days, users reported a peculiar temporal hiccup: when asked “what year is it?”, Google’s AI Overviews confidently (and incorrectly) stated the current year was 2024. This reporter experienced the same chronological confusion on Thursday morning, when a query about whether it was 2025 right now was met with a firm “no, it’s 2024” from the AI.
While Google managed to rectify this particular temporal tangle late on Thursday, the incident raises eyebrows. When questioned about the cause, Google offered no specific reason for the gaffe, instead providing a general statement: “As with all Search features, we rigorously make improvements and use examples like this to update our systems. The vast majority of AI Overviews provide helpful, factual information, and we’re actively working on an update to address this type of issue,” a spokesperson commented.
This isn’t the first time Google’s AI Overviews, a feature developed over several years, has produced notable blunders. Previous “highlights” include advising users to consume “one small rock per day” for its vitamin and mineral content, and bizarrely suggesting the addition of glue to help cheese adhere better to pizza. Furthermore, when the feature was launched in Hindi in India, TechCrunch investigations found its responses were often inconsistent and perplexing.
The persistence of such errors is particularly noteworthy given Google CEO Sundar Pichai’s recent triumphant announcements regarding AI Overviews’ adoption. He touted that the feature is now utilized by 1.5 billion users across more than 100 countries and is driving over 10% of search engine usage for related queries in key markets like the U.S. and India.
Clearly, Google is making a significant bet on steering users towards its AI-powered search and Q&A functionalities. However, incidents like the “year bug” and the infamous “rock-eating” advice demonstrate that while AI can perform complex tasks, its grasp on basic facts and common sense can sometimes be surprisingly fragile. As AI continues to integrate into our daily information-seeking habits, these episodes serve as crucial reminders of the ongoing development and refinement still needed to ensure reliability and, well, accuracy about what century we’re in.
