3 ms·
You really don't need Youtube for that. You need infrastructure and you need monetization. The former you can achieve today using any service provider with auto
by EarlKing 5y ago
You really don't need Youtube for that. You need infrastructure and you need monetization. The former you can achieve today using any service provider with autoscaling (i.e. AWS, Rackspace, Linode, and a hundred others too numerous to list here). The latter you can achieve through any number of ad exchanges that handle pre-roll ads and banners. Youtube is handy primarily in being a one stop shop for both of those things and more. It allows the technically illiterate a place to vent their spleen without having to know anything about the internet works... including, I might add, how to actually market their works so they get eyeballs on them. If, however, you have an IQ higher than your shoe size and can manage the above, including figuring out how to circulate your content to places where potential viewers are so you don't need to rely on Youtube's algorithm, then you can make a go of it yourself and win. Youtube is betting, quite correctly, that most of you will never do that.
- LinuxBender 5y agoI agree with all of this. That's why a small number of people and their circles won't be a threat to Youtube at all. Even the handful of popular Youtubers moving to self hosting probably won't get their attention for some time. And yeah, most non technical people won't be going down this path any time soon. That said if each small circle of people had a couple geeky friends and pool together some money they can do it. It's happening albeit it very slowly. Too slow for YT to notice or care about the insignificant churn-and-burn. I would not be surprised if there were also a chunk of youtubers that don't even want monetization.
- EarlKing 5y agoHonestly, it's better if YouTube doesn't recognize the threat. Let them crash and burn with no survivors. And yeah, there's a lot of people who talk a big game about not wanting monetization (where monetization = ads) and tout peercasting as a universal solution to the problem, not realizing just how much bandwidth video consumes. There are peercasting solutions for video out there (like Bitchute), but they're by no means a replacement for Youtube. In any case, probably the best development in the last fifteen years has been the emergence of people willing to pay directly for content. Whether it's something like paying for a network like Hulu, or patronage on Patreon or Twitch, it's proved that you don't have to sell your soul to advertisers to make money. The next few years are going to be very interesting for sure.