Remember the first time you used ChatGPT? That moment felt like magic. I remember typing a clumsy prompt and getting back something that sounded… human. I laughed, I gasped, and I immediately thought, “This is going to save me hours of work.”
And it did. For a while.
But lately, I’ve started to feel a creeping unease. It’s not about AI taking our jobs—that’s a whole other existential can of worms. It’s about something more subtle, more insidious. It’s about how the very tools we’re falling in love with are quietly, invisibly, changing the internet we grew up on. And here’s the kicker: we’re paying for it with something far more valuable than a subscription fee.
Let me explain.
The Great Bargain: Convenience for Context
We’ve all heard the old adage: “If you’re not paying for the product, you are the product.” It’s been true for Google Search (ads, your data), Facebook (ads, your soul), and now… AI.
But this isn’t the same bargain. With AI, you’re not just giving up your personal data. You’re giving up the texture of the internet.
Think about it. When you ask a chatbot a question, it gives you an answer. It’s clean, concise, and often correct. But it’s also flat. It removes the journey. You don’t get the weird, tangential forum post from 2008 that solves your problem in a completely unexpected way. You don’t get the passionate blog rant that teaches you more than you asked for. You don’t get the serendipity of clicking through five tabs and landing on a random Wikipedia rabbit hole about the history of the stapler.
OpenAI and Google are betting you don’t care about the journey. They’re betting you only want the destination. And maybe they’re right for most things. But here’s the hidden cost: the web is becoming a library of summaries, not a library of sources.
The SEO Apocalypse Nobody’s Talking About
Let’s get into the weeds for a second. I run a blog. You’re reading it. That’s cool. But I’ve noticed something terrifying in my analytics: traffic is dropping. Not because my content is worse, but because Google is now answering your question directly in the search results.
You know those featured snippets? They’re about to get a turbo boost. Google’s Search Generative Experience (SGE) is already rolling out, and it’s basically an AI summary that sits at the top of your search results. You type “how to fix a leaky faucet,” and instead of clicking my detailed guide with photos and a video, you get a three-paragraph summary written by a machine.
The result? Content creators like me lose traffic. Traffic means ad revenue. No revenue means no new content. No new content means the internet becomes a ghost town of old, stale information, all curated by a handful of AI models.
Here’s what most people miss: AI models don’t create new knowledge. They remix existing knowledge. If we stop producing new, original, human-driven content, these models will start cannibalizing themselves. They’ll train on AI-generated summaries of AI-generated summaries. The quality will degrade. It’s already happening in some corners of the web.
I’ve found that the most insightful articles I read today are the ones that don’t look like they were written by a bot. They have typos. They have opinions. They have a soul. And those are becoming endangered species.
Your Privacy Is Being Farmed by a Different Machine
We all know Facebook tracks us. But AI? It’s a whole different beast.
When you use a “free” AI tool, you’re not just feeding it your queries. You’re feeding it your thinking. Think about the prompts you type. They’re personal. “Help me write a breakup text.” “What’s a polite way to tell my boss he’s wrong?” “Generate a business plan for my side hustle.”
Every single one of those queries is data. It’s emotional data, strategic data, personal data. And it’s all being hoovered up by companies that are notoriously opaque about their training practices.
I’m not saying they’re reading your emails. But I am saying that the boundary between “training data” and “surveillance” is getting blurry. OpenAI and Google are in a race to build the most “intelligent” model. To do that, they need the most diverse, human-like data possible. Where do you think that data comes from?
It comes from you. For free.
The Creativity Tax: Are We Outsourcing Our Brains?
This is the one that keeps me up at night.
I’ve started to notice a pattern in my own writing. When I hit a block, I used to stare at the wall, take a walk, or read something unrelated. Now, my first instinct is to open a chatbot and ask for a “hook” or a “better way to phrase this.”
It works. It’s efficient. But am I getting better at writing, or am I getting better at prompting?
There’s a subtle difference. The AI is doing the heavy lifting of creativity. I’m just the project manager. Over time, I worry that my own creative muscles will atrophy. Why struggle through a tough paragraph when the machine can do it in three seconds?
This isn’t just a writer’s problem. It’s an everything problem. Artists using DALL-E. Coders using Copilot. Marketers using Jasper. We’re all outsourcing the process of creation, not just the task.
Let’s be honest: The internet is about to get a lot more boring. Not because AI is dumb, but because it’s too smart. It averages out the weird, the brilliant, and the broken. It gives you the most probable answer, not the most interesting one.
So, What Can We Actually Do?
I’m not saying throw your laptop in the river and go live in a cabin. AI is here to stay, and it’s genuinely useful. But we need to be intentional. Here’s my shortlist for staying sane:
- Use AI as a junior editor, not a ghostwriter. Let it polish your work, not create it from scratch.
- Reward human content. Click on the blog post. Leave a comment. If you find a weird, passionate article that clearly wasn’t written by a bot, share it.
- Limit your “pure query” use. If you need a fact, use a search engine and click a link. Save the chatbot for brainstorming or summarizing, not for replacing the discovery process.
- Pay for the good stuff. If you use an AI tool heavily, pay for the subscription. It’s a tiny signal that you value the service, and it reduces the chance you’re the product being harvested.
But efficiency isn’t the same as enrichment. And I, for one, don’t want to live in a perfectly curated, soulless library of machine-generated summaries.
I want the rabbit holes. I want the typos. I want the human mess.
The question is: do you?