When Google was founded in 1998 it solved a big problem: spammy and chaotic web. Does this sound familiar? Because it's happening again. The World-Wide-Web is still a place for everyone, but if it doesn't have good SEO nobody will find it. AI webs and images are surrounding the internet.
How is Google handling that? By deleting blue links and replacing them with AI agents. How are other search engines handling that? By creating blacklists and scanning websites using AI detectors. That will never work.
Well, I have an idea...
[ ? ] Basic concept
Basically, imagine Wikidata (a big database of lots of things filled with data, similar to Wikipedia), but not for objects, but for websites. Websites would have their own pages with information about them - what kind of website it is, whether it has ads, a short description, when it was created/who founded it, how big it is, etc. It would be all open-source and managed by the community.
[ * ] Features
Sounds weird, right? Why would someone do it? Well because of the advantages:
- You can filter your search in hundreds of ways! One example: you want to find something on a small forums, that are ad-free and founded more than four years ago for no AI garbage. On google? Impossible. In this idea you would just type it into the search. Super useful. Do you know any other search engines where you can do this? I don't.
- What about the history of websites? Normal search engines only work with how the pages look at the moment. But that's not always efficient. With a database of website history, a search engine can find out that a website has been changed (all its content) four times in a month. That's not how normal websites look, is it?
- Another thing is SEO. SEO requires experts and a lot of money. In this system, you just enter data into a database - anyone can do it!
[ @ ] Image search
These days, some AI images are that that realistic, that is impossible to tell if the image is AI or not. So we have to do it differently, not by checking when the image is already taken, but by when the image is being taken - insert cryptic metadata into the photo. And I'm not talking about that garbage EXIF metadata that can be overwritten, but about C2PA.
It's still not mainstream, but it's groundbreaking. It’s also the only way to make sure that the photo was taken by a human.
[ # ] Potential Challenges
As every project in it early stage, it have potential challenges. But they are not impossible to solve, they only require some effort. The database can be filled it can be full of spam from random people, from bots from anyone. To solve this problem, we would need to implement a system similar to the one used by Wikipedia.
Another thing, MONEY. I don't have any idea, how much it can cost, but it will be not for free.
The internet is changing rapidly. It would hard to keep the database fresh, hm? Well, bots can help a lot, by checking important info (if website is only or not and countless other details).
[ / ] Final though
I am releasing this here, because I want someone to make it real. I'm not planning to do it myself, I don't havr enough skills to do it. I just want to make search fun again.
If Wikipedia could build the world’s biggest encyclopedia from scratch, it's possible to build the world’s biggest web directory.
