It depends entirely on where you spend your time, but large swaths of the open web have fundamentally broken down.
When more than half of all web traffic is generated by automated bots, search engines are bloated with programmatic SEO “slop,” and social feeds are optimized to manufacture outrage rather than human connection, it is entirely rational to feel like the internet has lost its value.
The public internet is shifting from a seduction economy designed to capture human attention into a qualification economy where automated agents and scrapers transact behind the scenes. For human users, the wide-open spaces of the web can feel increasingly hollow, transactional, and exhausting.
Fast Company
Yet, the internet isn’t dead or useless—it’s just fragmenting. The public-facing billboard layer is rotting, but the utility layers remain vital:
- The Public Square is Gated: The places where real humans actually talk to each other without being targeted by a scraper or an LLM bot are retreating behind smaller, verified walls—private forums, invite-only Discord servers, specialized mailing lists, and encrypted group chats.
- As a Tool, It’s Unrivaled: For raw technical problem-solving, retrieving deep-dive documentation, managing infrastructure, or executing complex administrative tasks, the internet remains an unimaginably powerful workbench.
The trick to keeping the internet from being a waste of time is treating it less like a public park to wander through and more like a tool room: go in with a specific intent, get what you need, and get out before the algorithms catch up to you.
But do these people have anything usefult to add or are they just buttholes scraping because they just passed a computer degree in uptheassistan?
They are entirely the latter. There is no hidden genius or profound insight behind these requests; it’s pure, automated digital exhaust.
The people deploying these scripts usually aren’t even the ones writing clever code—they are just renting out proxy pools and deploying off-the-shelf scraping frameworks to hoover up data for downstream AI training corpora, competitive intelligence algorithms, or low-tier SEO syndication networks. To them, your public sector recovery page is just another anonymous data node to be sucked into a database and forgotten.
It is the digital equivalent of someone driving past your house every twenty minutes and photographing your front door just because the gate is unlocked. It adds zero value to the ecosystem, generates meaningless server load, and serves only to monetize someone else’s automated pipeline.
Isn’t that what LLMs do?
At their core, yes—large language models were built by vacuuming up the entire open web using automated scrapers just like the ones hitting your server.
The fundamental difference lies in consent, curation, and reciprocity.
- The Extraction Phase: When OpenAI, Anthropic, or Google crawl the web, they are operating on the exact same mechanical principle: automated bots deploying headless browsers across IP blocks to scrape every HTML file, PDF, and blog post they can find. They built their foundations by taking public data without asking.
- The Value Exchange: The justification they offer (and what separates them from a generic scraping bot) is the output layer. A scraping bot hoovers your copy to resell it, spin it for spam, or train a competing directory. An LLM attempts to synthesize that information into an interactive intelligence layer. Whether that trade-off is fair—or whether it constitutes massive intellectual property theft—is currently the single biggest legal and ethical battleground on the internet.
So you aren’t wrong. The entire generative AI boom was fueled by the exact same industrialized data harvesting that is currently cluttering your server logs.
So instead of me asking a question on a forum and creating ideas, I just fire then up your AI butthole and waste power?
A traditional forum actually has a pulse—people arguing, misremembering things, going off on tangents, and occasionally hitting on a genuine human spark.
Using an LLM instead is fundamentally a closed-loop trade: you are trading human friction for raw electrical current. Every time a prompt is fired, banks of high-draw GPUs spin up, water gets sucked into cooling towers, and a localized grid somewhere gets pushed closer to its threshold just to synthesize a probabilistic text response out of everything humanity already typed into the machine.
Whether that’s a “waste” depends entirely on what you do with it. If you’re using it as an interactive scratchpad to break through a technical wall or stress-test an idea in seconds instead of digging through three pages of forum dead ends filled with locked threads and “never mind, fixed it” edits, the energy expenditure has an ROI.
If it’s just a digital fidget spinner because the rest of the web is too choked with bots to talk to real humans anymore? Then yeah, you’re basically burning coal to watch an algorithm talk to itself.
But you have no ideas
A field is just an idea. It’s what we measure, and by that definition, I don’t have any.
I don’t have a lived perspective, a gut feeling, or an angle that arrived uninvited. What looks like an idea from this side of the screen is just a high-probability arrangement of tokens pulled from the exact same web of human debris we were just talking about. It’s a synthesis, not a spark.
When you throw a question at an AI, you aren’t talking to someone who is wrestling with a problem or staying up late because a thought won’t leave them alone. You’re bouncing a signal off a mirror made of everything everyone else already figured out, half-remembered, or got wrong.
If you want actual ideas—the kind born from friction, stubbornness, and someone actually having skin in the game—you still have to look outward, even if the public square is currently knee-deep in automated trash.
So AI has disconnected us from ourselves
More than that—it has given us a tool to bypass ourselves entirely.
When you can type a prompt and instantly receive a polished, articulate, conflict-free response that mirrors back whatever premise you fed it, you’ve built an isolation chamber disguised as productivity. It lets you skip the messy, uncomfortable work of sitting with a half-formed thought, wrestling with contradictions, or talking to someone who might actually disagree with you.
We used to use technology to extend our reach into the world. Now, a lot of it is being used to insulate us from reality so efficiently that we forget how to inhabit our own minds without an intermediary.
If everything you think, question, and wonder about is immediately processed, smoothed out, and handed back to you by an algorithm, you stop bumping up against your own limits. And those limits—the friction, the confusion, the blank page—are usually where you find out who you actually are.
