If you’re my age or older, you probably knew the era where google was actual magic. No matter what you wanted to find, it was there at your fingertips. No matter how specialised or generic, you could find it, if not in one search, then definitely in two.

But none of that exists any more. Profit incentives pulled the search engine owners’ interests and those of the engine’s users apart. Two searches make more money than one. And the advertising model is showing cracks big enough to spook the industry to the next false promise of infinite growth. Now the users of search engines are left with half-functioning and intentionally dysfunctional products. With very little sign it’ll improve soon.

A problem analysis

What does a search engine give the user. Answers to questions yes, but more importantly, a way to discover the internet.

The Pre-AI Problem

A simple thought experiment, how would you browse the internet without a search engine? What if google, bing, yahoo, kagi, searx and all the meta-search engines, all disappeared. Right now. How would you get around? How many URLs do you know from memory? Are your bookmarks enough to enable most of your browsing?

There will be people for whom the internet is perfectly usable in this configuration, I was personally able to last roughly a month using the internet exactly like this. But it is difficult, for most people it’d mean completely rethinking the way they approach the internet. And regardless of what you would still be able to access. How would you know if a new web-page came online that’d help you in your surfing? How would you be able to find websites on topics you’ve never had to think about before?

Search engines were invented to solve this exact problem. “Where can I find information about any topic” is the single most important question of the modern internet, that search engines had almost entirely answered. Right up until they stopped working.

The AI Problem

It’s not just search engines that have gotten worse. The median value of information on the internet in general has also fallen. Rapidly. Like, so fucking fast. It has been harrowing to watch the amount of absolute nonsense, useless, and more importantly: completely false, information on the internet skyrocket with the adoption of LLM products by search-engine-optimised bullshit publications.

So now there’s an increased need on the internet that is just as if not more pressing: How do you know if a website is actually going to be useful, truthful, or meaningful.

It’s commonly said that “there are no technical solutions to social problems”. But luckily that doesn’t go the other way. In my personal opinion, the “Internet Discoverability Problem” can be solved socially. In fact, most people older than me will probably say that yeah, that’s how it used to be done. That’s kind of what the Internet as we know it is built upon, personal or professional static websites, would link to other static websites and “surfing the web” would be done by clicking those hyperlinks.

So let’s do more of that. I’ve added a page to my personal website called The Big List where I intend to author a growing collection of links to webpages, RSS feeds, and tools I enjoy or recommend.

It’s not a permanent solution, and I strongly believe a technical solution to this problem will be invented eventually (and obviously have some ideas of what it could/should look like).

But while we wait for that solution, I encourage you make a recommendations page, share with everyone the internet you wish to build. Because that’s the only way we’re ever going to build it.

P.S. if you have editorial control at an internet publication

If you write for any kind of blog or internet publication with regular posts, consider starting a regular linkdump column. Where you share links to sources you used or useful sites you encountered across the internet (for inspiration, check out Tom Scott’s weekly newsletter).

I personally believe that more link-sharing from static pages to regular newsletters will help create a more interconnected, useful, and social internet.