The value-add of these services is pretty substantial, though. Free hosting alone is pretty big, since even setting up two servers to host your videos (one EU, one US) won't be as good as Google's global peering and edge storage servers[0]. And the web in terms of discoverability is dead - even if everyone used RSS readers and creators hosted their own videos, how do they find new viewers that want to watch their content without resorting to traditional advertising, or hoping word-of-mouth (online forums, IRL, or otherwise) works?
this only applies to situations where your content is large in terms of byte count.
For textual content, there's barely any real cost tbh. a caching layer solves most of your problems with edge node speed (so you pay cloudflare or akamai or some such).
> The value-add of these services is pretty substantial, though.
For videos (and music to a degree), it's unique in that the content size is huge, and people are increasingly impatient. I have no doubt that youtube infrastructure cannot be replicated by a private person. Therefore, their services are not "value-adds", but the primary value, and your video content is the "value add".
"Discoverability online" is a perennial question. And it's a rabbit hole as well since there's more then one way of exploring, discovering and consuming information with the question of "relevancy" being entirely context-driven.
Back in the 90s and 00s, the Web/Internet was promoted as a break from traditional mainstream media e.g. flipping between a fixed set of television networks that curated the content you'd get to see that night. It was hailed as this new way of sharing information compared to "40 channels and nothing's on tonight".
"What is the Web supposed to be?" has always been a question. Push or pull? The age of curated feeds and algorithms hasn't answered that question definitively. On the contrary.
> how do they find new viewers that want to watch their content without resorting to traditional advertising, or hoping word-of-mouth (online forums, IRL, or otherwise) works?
The Web/Internet is a game changer when it comes to providing access to services and information anytime any place. But, I don't believe that the Web / Internet was ever going to definitively solve an essentially a challenge of interaction between humans: "How do I get noticed?"
Consider a traditional library. As far as authors are concerned, putting books together and providing catalogues or curated suggestions to patrons doesn't imply that every single one of them will ever see the same reach or get the same readership. Not unless their work gets actively promoted.
Putting your work online - whether self-hosted or via a platform - only moves that challenge from the analogue to a virtual space.
Your mentioning RSS readers is interesting because subscribing to RSS feeds implies that you went through a phase or process of discovery first. Maybe someone recommended someone else's blog. Maybe you stumbled upon a blog via a blogroll. Maybe it's a blog of a friend who gave you their domain at a social occasion. There's just more then one way for a feed to arrive in your RSS reader.
To my mind, some of the criticisms about the quality of content on platforms, or what platforms recommend in their feeds, today draw parallels to criticisms given regarding last century's TV networks. To an extent, I think it's because the recommendation feeds of these platforms work just the same way: as constant feed that's passively consumed, having consumers flicking through whatever is presented until something engaging passes along.
Not at the scale YouTube operates; the maybe 100k english creators (and a million more elsewhere) that have a full-time job doing content creation wouldn't have it without YT's discoverability algorithm.
There seems no way to prove or disprove that statement, since the evolution of the web created this reality and a different reality never got tested.
The current reality arguably depended on significant forks in the road of history, at least some of which might have been somewhat random.
For example, in a different, less monopolistic, version of evolution, there might be twice as many content creators, each making about half as much money.
0: https://peering.google.com/#/infrastructure