google wounded the internet and ai companies are treating the injury with mercury.

October 08, 2026

The internet allowed us to publish and link-together an enourmous amount of information, how to find anything in this vast ocean of interlinked documents is still a problem. To build a better internet we need to work on its content's discoverability.

Finding something in an information mess is not a new problem, archivists have been working on it for hundreds of years. While their work focus mostly on phisical archives the underlying problem and challenge is fundamentally the same on the digital world. Google's original mission was to "organize the world’s information and make it universally accessible and useful". Almost 30 years later, that problem is still there.

To tackle this challenge, Google built a service consisting of indexing everything it can put its hands on and provide internet's users a single entry point for it. The idea was good as it somewhow solves the issue: I have questions, there is a shit amount of information out there that I cannot possible process so I enter my query in a text box and get results of things that can satisfy it. The problem with narrowing the Internet to a single entry point is all the side-effects of it:

  • There is no clear rules on who gets to the first results,
  • It degrades the value of the internet by narrowing everything to just the first 10 resutls (to select 10, it needs to darken thousands),
  • People's behaviour start changing and writing content to be discovered,
  • You get paid results that will appear whathever it solves your query or not,
  • Publishers got entangled in a ethically-dubtious business models of advertisment that relied purely on massive amount of data collection,
  • etc.

The list go on but the situation is this: the solution that Google envisioned led to a natural enshittification.

Going into the year 2027 the problem got out of hands. Search engines (and the web itself) got so enshittified that trying to find something useful is no longer possible. Let's recap what a modern experience of browsing the web looks like: 1. We go to a search engine, 2. We enter my search text and I pray I get some decents hits 3. We get redirected to the results page. 4. We need to skip the paid ads (usually the first results), 5. We pull out our internal click-bait classifier to avoid clicking sites we know are shit, 6. Open a plausible result, 7. Instantly deal with the cookie banner (a by-product of Google's business model BTW), 8. Scroll through a couple of paragraphs of bullshit content that's only there so it get's picked and ranked by Google's algorithm, 9. Scan the website while dodging the ads, 10. Hiding the banner that invites you to subscribe to a newsletter, 11. Read a couple of paragraph that looks interesting, find out that the site does not provide what you where looking for (frustrating but can happen..). 12. Rinse and repeat until you get something usefull.

The internet's experience is just shit.

With the quality of Search Engines nowadays it is more than clear that users are going to pivot to whatever promises a better experience. We poison the internet so much that people are begging for a new experience to deal with it. And yes, here it comes AI companies with their radioactive quackery (did you know that radioactive thootpaste was a thing?) trying to sell us yet another magic text box. This one comes with a new flavor: to digest all the shit in the internet for you so you don't need to manually browse it. I sometimes find myself asking to the AI because I cannot longer tollerate yet another web search. The main problem however is that AI depends on searching! Underneath, what AI does is search the internet, read every page and summarize all the findings. If what they find while searching the internet is the garbage content we have released for engagement with search engines, the AI is going go that extra mile of effort to serve you a beautiful plate of shit.

So what's the solution you would say? No idea yet. But here are some good starting points to keep in mind:

  1. The internet is now full of garbage,
  2. Good content lives mostly in small community driven sites that usually are not findable by search engines nor AI, and
  3. We now that single entry-point solutions that promises you to organize the world's information and make it universally accessible can go terribly wrong.

The good thing is that the problem didn't change in nature (just in scale) it is still a problem of organizing a lot of information and, now more than ever, determine what has long-term value that it is worth keeping. I guess archivists are going to be more needed than ever.