The way I see it is the web was supposed to transform into the Semantic Web in 2000. What that means is you can give everything an identifier and create any type of relationship, turning the entire web into a database designed for a type of reasoning. On its own this is pretty benign, and I thought it would be a Good Thing because it would engage people past pages, links, and predetermined content. I imagined it becoming interesting for most people to craft meaningful connections. I also thought it would enable people to be more usefully and grounded-ly critical of big orgs, through projects like Sunlight Foundation and eventually Wikipedia (and now Wikidata, which uses the formal Semantic Web).

But the basic idea of "things" and relationships is easy to conceptualize, in fact those physical evidence boards with people and things connected by association have been around for a long time.

But the formal Semantic Web tech proved too complicated, so it's only really used in areas like science and to a limited degree for SEO. However, companies like Google and Facebook saw the value and built their own "knowledge graphs," and eventually companies like Palintir started selling services built on the ability to create knowledge graphs about anything. There are a few companies that do this on a smaller scale, but Palintir has really taken off, and given (and perhaps because of) the personality of the founders and its policing/military focus, and the fact that everyone is incidentally connected one way or another, it has become a very dangerous tool because it ironically does the opposite of what these tools should do on a large scale, enables big orgs to track individuals, starting perhaps legitimately with crooks and terrorists but now expanding into every field and citizen, meaning everyone can be tracked, predicted, and managed (barring good societal rules). If it isn't stopped, it gives absolute control to whoever is on top.