And yet we can't turn this off
byu/Independent-Wait-873 ingoogle
What year is it? A simple question, right? You’d think a cutting-edge AI from a tech giant like Google would nail it. Yet Google’s AI Overviews seem to be having a bit of a temporal tangle. A recent screenshot shared online showed a user asking “is it 2025?” only for the AI Overview to confidently, and confusingly, state, “No, it is not 2025. The current year is 2024, as of today, May 27, 2024. The year 2025 is in the future, and the current date is May 27, 2024.”
This isn’t even a one-off. We tested it ourselves and Google’s AI Overview gave a similar response.
This isn’t an isolated hiccup generating head-scratches and frustration. The AI-generated summaries, which Google has been aggressively rolling out at the top of search results, are meant to provide quick answers. Instead, they sometimes offer responses that are baffling, incorrect, or just plain weird, especially for seemingly straightforward queries. Matter of fact, there are so many such examples that accounts like @Goog_Enough were keeping tabs on all the blunders, as highlighted by Search Engine Roundtable last year.
The public’s patience is wearing thin, particularly as there’s no easy off-switch for this feature. One commenter said:
I hate AI Overview with a burning passion it wouldn’t be so bad if it wasnt forced but the fact I can’t turn it off is the human equivalent of hell.
Some have resorted to adding “-ai” to their searches or installing third-party extensions. Others change their default search engine, but none of those fixes come from Google itself. We, too, have highlighted multiple ways to disable AI Overviews, however, having to manually tinker with stuff just to get correct information shouldn’t be something that we need to tolerate.
Google has tried to put a positive spin on AI Overviews. In a blog post at I/O 2025, the company touted the feature’s roll-out to over 200 countries and support for more than 40 languages. It claimed people were searching more and clicking deeper into results. While the company is gloating about how many users use AI Overviews, they seemingly left out the fact that users don’t even have the option not to use it.
“Billions of people are using AI Overviews every day”
— Lily Ray 😏 (@lilyraynyc) May 22, 2025
Uh yeah my dude, you made it virtually impossible to turn off 😅 https://t.co/rkDQGZMhEV
And even if we take Google’s word, when the AI itself gets basic facts wrong, that kind of momentum feels built on shaky ground. Users who relied on a quick summary now have to scroll past it or ignore it altogether.
The introduction of ads within AI Overviews and the more immersive “AI Mode” further complicates the picture. While AI Mode is pitched as a more powerful search experience, publications like New York Magazine have argued that these AI-driven changes are “burying the web alive.” The concern is that by summarizing content directly on the results page, Google reduces the need for users to click through to the original websites. But that’s not all. When the AI cites businesses, it tends to link back to their Google Business Profiles rather than the website. This could starve content creators and businesses of traffic and revenue, ironically diminishing the very web that these AI models are trained on.
When AI Mode cites your business, there’s a great chance Google is only linking to your Google Business Profile, not your homepage or any other page on your site.
— Lily Ray 😏 (@lilyraynyc) May 26, 2025
Sorry about that shiny website you built… and your analytics 😅 pic.twitter.com/W1nksxPIXA
And if you think the evil ends there, I have more bad news. Publishers can’t even prevent Google’s AI Overviews from scrapping data from their sites unless they opt to completely remove themselves from Google Search itself. You can read more about this in our coverage from earlier this month.
But this doesn’t end with AI Overviews. AI in general isn’t trustworthy at the moment by a long shot due to how prone these models are to hallucinations. We’d assume these AI bots will be far smarter than the average Joe, but that’s far from the truth in certain situations. At times, simple tasks also render these AI bots useless. Take this instance for example, where Gemini couldn’t grasp the concept of tossing a coin randomly. Rather than randomly outputting either heads or tales, Gemini just kept alternating between the two.
At the end of the day, I’m sure this isn’t going to be the last fumble by Google. It feels like silly AI responses to the most basic queries will soon be the norm. It’s hard not to laugh at the irony. Google’s chasing the AI dream, but it can’t nail down the current year. Until they fix these hiccups, AI Overviews will keep sparking more groans than awe. Users just want the truth, not a tech experiment that’s still figuring itself out.