I mean that, ideally, facebook, medium, and youtube wouldn’t be websites, because the web is a poor solution to the problem of building responsive user interfaces.
Sometime in the late 90s, we all collectively decided that all networked applications were supposed to run off port 80 and sit inside a web browser. This leads to massive engineer labor (software engineers end up spending enormous amounts of effort trying to do what would be trivial as a native app inside a web browser, software engineers spending an enormous amount of effort trying to make web browsers capable of doing everything web designers are trying to do with them without creating security flaws), massive resource waste (cycles and packets wasted en masse because nobody’s using more efficient methods, even when those methods would also be easier to implement, because crappy solutions have become the default), and ultimately we screw the pooch on security too because we’ve layered everything important on top of a simplified version of a 1992 demo intended to explain Enquire to suits instead of ensuring that our technical decisions make technical sense.
We’re locked in now. But, were we to make sensible decisions from the beginning, our hypertext systems would have permanent content-based addressing with automatic replication (instead of temporary machine-based addressing that fails to uniquely, consistently, or permanently correspond to any given piece of information), no scripting (since scripting is totally unnecessary for hypertext), no association with the domain name system (because domain names are broken), no association with the certificate chain system (because root certificates get leaked all the time), no ad-based revenue (because ad-based revenue encourages a race to the bottom in terms of content quality), and no embedded markup (external, offset-based markup is easier to implement and avoids most common markup-related problems).
As a web developer, your job is to make objectively poor decisions about web development in ways that are profitable to your employers. That doesn’t mean that you need to believe those decisions are good; you just need to implement them.