There was a moment — sometime around 2012, give or take — when typing a question into Google actually felt like a superpower. You'd get a list of real websites, written by real humans, pointing you toward real answers. It was imperfect, sure, but it worked. It was useful.
Then someone in Mountain View looked at that utility and thought: what if we monetized the hell out of it?
Fast forward to today. Type in a question about, say, the best running shoes for flat feet, and here's what you'll encounter before you see a single organic result: four sponsored listings, a Google Shopping carousel, an AI-generated summary that confidently tells you something subtly wrong, a "People Also Ask" accordion that exists purely to keep you on Google's real estate, and — if you scroll far enough — an actual website. Maybe.
The Slow Creep of Sponsored Everything
Google's ad revenue in 2023 topped $237 billion. That number doesn't happen by accident. It happens by gradually, methodically reshaping what users accept as a "normal" search experience.
The company has spent years expanding the definition of what counts as a result. Ads used to be clearly labeled, visually distinct, and confined to a sidebar. Then the sidebar disappeared, the labels shrank to near-invisibility, and the ads moved directly into the result stream. Studies have repeatedly shown that a significant chunk of users can't reliably distinguish paid listings from organic ones — and Google has shown zero interest in fixing that confusion, because that confusion is the product.
In parallel, Google began colonizing search results with its own properties. Want flight info? Google Flights. Weather? Google's widget. Local restaurant? Google Maps. Business hours? Google's knowledge panel. The actual website of the business you're looking for? Somewhere below the fold, if it's lucky.
This isn't a bug. It's a business model.
SEO Spam and the Content Farm Renaissance
Here's the thing about a system that rewards ranking over relevance: it creates an entire industry dedicated to gaming it.
The SEO content mill is one of the web's great unacknowledged disasters. Entire networks of websites exist not to inform, entertain, or help anyone — but to capture search traffic through keyword-stuffed articles written to satisfy an algorithm rather than a human reader. You've seen them. The recipe blog that makes you scroll through seventeen paragraphs of personal memoir before reaching the ingredient list. The "how-to" article that answers every possible adjacent question except the one you actually asked. The product review site where every single product gets a 4.7 out of 5.
Google has periodically claimed to be cracking down on low-quality content. Its "Helpful Content" updates were supposed to demote this garbage and surface genuinely useful material. The results have been... mixed, to put it charitably. In practice, many small independent websites with real expertise saw their traffic crater, while large content aggregators with armies of mediocre writers continued to dominate. The algorithm, it turns out, is better at detecting signals of authority than it is at detecting actual usefulness.
The Reddit Gambit
Somewhere along the way, users started appending "reddit" to their search queries. Not because Reddit is always the best source of information — it frequently isn't — but because it tends to surface actual human opinions rather than algorithmically optimized marketing copy.
Google noticed. And rather than fixing search to surface genuinely helpful content, it struck a deal. Reddit threads now appear prominently in results for all kinds of queries, creating the ironic situation where a forum known for misinformation, brigading, and years-old advice is now effectively subsidized as a Google search product. Reddit signed a $60 million annual data deal with Google in 2024. The appearance of organic community results is, in part, a licensing arrangement.
This is the search engine's answer to the authenticity problem: buy the thing that looks authentic.
AI Summaries: Confidently Wrong at Scale
Then came AI Overviews — Google's gambit to keep users from ever leaving its pages at all.
The pitch was reasonable enough: let a large language model synthesize information from multiple sources and give you a quick answer up top. The execution has been a minor catastrophe. Within weeks of launch, AI Overviews were recommending users put glue on pizza to help cheese stick (sourced from a satirical Reddit comment), suggesting people eat rocks for nutritional benefits, and hallucinating medical advice with the calm confidence of a doctor who definitely did not go to medical school.
Google has patched the worst offenders, but the fundamental problem remains: these systems generate plausible-sounding text, not verified facts. And plausible-sounding text, displayed prominently above all other results, is deeply, structurally bad for anyone trying to find accurate information.
The cruelest irony is that AI Overviews don't just risk giving you wrong answers — they actively suppress the source websites that could give you right ones. Publishers have reported traffic drops of 20 to 60 percent for queries now captured by AI summaries. The web's content gets consumed to train the system, and then the system replaces the web.
What You're Actually Left With
The result of all this — the ads, the self-preferencing, the SEO spam, the Reddit laundering, the AI hallucinations — is a search engine that has become genuinely difficult to use for anything that requires precision or trust.
Need to troubleshoot a specific software error? You'll wade through a dozen SEO-optimized tech blogs that paraphrase each other before finding a Stack Overflow thread that actually helps. Looking for a local small business? Good luck distinguishing the real listing from the three competitors who bought ads against its name. Researching a health question? The AI summary will give you something that sounds like it came from a medical professional but was synthesized from a Reddit post and a content farm article from 2019.
The web's primary navigation tool has been optimized — relentlessly, brilliantly — for everything except navigation.
And the worst part? There's no obvious off-ramp. Bing is Microsoft's version of the same problem. DuckDuckGo is earnest but limited. The independent search engine space is a graveyard of well-intentioned projects that couldn't survive without Google's index or Google's advertising dollars.
Search is broken. Google broke it. And every quarter, it reports record profits for doing so.
The internet is working exactly as intended — just not for you.