This isn’t directly Bubble-related, but scratches an itch that I’ve had for some time: how exactly does a vector database work?
I’d open an explainer, read the words “high-dimensional embedding space,” and close the tab, thinking this was on par with quantum physics. Every description I found was either four words long and useless or forty pages long and worse.
But throughout my entire career, I’ve suffered from a particular compulsion: I cannot leave a tool unexamined. So I did the thing I do. I tried taking it apart until it made sense. This is the result. My notes, polished. And it turned out to be the most elegant and fascinating piece of technology I’ve come across in a long time, and the story runs from a 19th-century librarian to the memory inside every modern AI.
The core idea: meaning is distance. Your brain already works this way. Thoughts are found not by spelling, but by resemblance. A psychologist proposed in 1987 that this might be a universal law of how minds organize meaning. Decades later we built one out of arithmetic, and it obeys the same law.
It’s long. Grab a cup of coffee. But if you’ve ever wanted to actually understand what’s under the hood of the AI tools everyone’s using, jump into the rabbit hole with me and let’s find out.
Feedback is very welcome!
