Google used to be the kind of relentlessly ambitious tech giant that pursued moonshots reminiscent of science fiction; think internet delivered via giant balloons and even an effort to try and cheat death. Today, however, a Google Search user might literally die if they followed some of the insane advice of Google’s AI Overviews, the new iteration of Search that frequently serves up machine-generated slop with the same misplaced confidence as a drunk guy explaining crypto at a party.
I couldn’t help but shake my head in disbelief at Google co-founder Sergey Brin’s recent admonition to the company’s workforce that 60-hour workweeks should be the norm as the company continues to build out its AI products. As if the fact that last night, when I asked Google Search, “is it 2025,” Google’s AI Overview confidently told me that, no, we’re still in 2024 — that, somehow, such an error is an employee timecard problem, and not a product of a once-great company drinking too much of the AI hype cycle’s Kool-Aid.
To be more specific about that query that I asked Google Search: Surely, I thought, asking what year we’re in is the lowest possible bar for an AI system to clear. Either it is 2025, or it isn’t. This is the kind of binary question that you’d think ought to be AI’s bread and butter. Instead, Google’s new AI Overview confidently told me: “No, it is not currently the year 2025. The current year is 2024” (image below).
The kicker: The AI cited a Wikipedia article to back up that nonsense — an article whose very first sentence begins with “2025 is the current year.” In other words, Google’s AI (that users supposedly love, Google tells us) plagiarized Wikipedia and still managed to get the answer wrong.
It’s all incredibly sad to watch, because Google Search used to be the closest thing the internet had to an oracle. You could type in the vaguest half-formed thought and instantly get exactly what you were looking for. Now, it’s a glorified Reddit comment section, albeit with extra steps — confidently hallucinating false information while plastering links underneath to make the mistakes and hallucinations feel more credible. Why should anyone trust Google’s machine-generated summaries when the AI can’t even figure out what year it is?
I should add: I first noticed that error, above, last night. As of this afternoon, it remains unchanged.
There’s something darkly poetic about watching the world’s smartest company accidentally turn its own search engine into a bumbling idiot. Google spent decades building the most powerful information retrieval system in human history — only to break it in record time, all because the company convinced itself that speed and convenience trump accuracy.
The tragedy is that Google didn’t lose its edge because someone built a better search engine — it lost its edge by forgetting what made it great in the first place. The old Google was a portal to the web’s infinite complexity, a tool that rewarded curiosity. The new Google feels like a vending machine for mediocrity. No wonder a recent report found that around one-third of active web users don’t use Google all that much anymore. And I’m actually a good example of why that’s the case.
Search for my name, and Google gives you some correct links and details about me, but also a ton of irrelevant details that have nothing to do with me. Which means, if you were probing me, you’d probably want to try your search again with something more specific — and, voila. No wonder Google can crow it’s now up to several trillion Google Searches executed per day. When you’re a monopoly, you can get away with enshitifying your service, and people will counterintuitively use it more.
Brin’s recent criticism of the company’s workforce for not spending enough hours in the office working on AI feels like a misplaced attempt to scapegoat employees. The real problem isn’t remote work; it’s leadership. If Google is falling behind in AI, the blame rests with the people at the top, not rank-and-file workers.
Brin and Larry Page voluntarily stepped away from the CEO’s office in 2019, leaving the company in the hands of Sundar Pichai — a capable operator, but hardly a visionary leader. Under Pichai, Google has transformed from an innovation powerhouse into a risk-averse bureaucracy, more focused on maintaining its search monopoly and cutting costs than pushing boundaries. While OpenAI and others sprinted ahead with generative AI breakthroughs, Google hesitated — despite having many of the foundational AI technologies first. It’s not that Google’s engineers aren’t working hard enough. It’s that they aren’t being led.
If Brin wants to know why Google lost its edge, he should ask himself and Page why they handed the company over to a caretaker CEO instead of staying in the fight to define the future themselves.
