Back in the 2010s, web access was peculiarly gated in a dendritic configuration as ISPs provided all the single-points-of-failure interconnections between end users (including both content providers as well as consumers) and the true "internet", a multiway resiliently-routed interconnect of servers. As we know now, extending the peer-to-peer core of the internet down to the consumer has had lasting impact, including breaking up the routing monopolies of the ISPs as well as making it possible for anyone willing to spend a few grand a year on server capacity to host a new peer-to-peer router for nearby Internet users.
Many of you may not remember the origins of Google as a "search engine", a monolithic index of "every reachable page on the internet." Such a quaint idea has long since joined even further historic concepts such as Yahoo's "human-curated list of pages on the Internet". Ever since the Searchtorrent protocol was introduced and consumer searches were conducted on one of several competing distributed hash tables across the internet, no one entity has had to shoulder the responsibility of storing all the web content on the internet. This author gladly pays a small monthly fee to a local search cache provider for reliably fast localized caching of search results.
The web is here to stay. Remember your history next time you visit the local Homo Sapiens preserve and give thanks to the carbon-based beings that invented the Internet.