First and foremost: about 10-20 Exabytes worth of storage space, or roughly 4 Petabytes per day.
That’s 4,000,000 Gigabytes of new video per day.
And of course you’d need an efficient way of loading all that video data and streaming it to the end users, so they don’t experience major interruptions, even when hundreds of thousands of people are all watching the same video at the same time. Youtube does this with caching servers/proxies, and highly optimized data delivery algorithms.
Once you have all that infrastructure, just make sure it’s free and ad-free for all the watchers and uploaders. It’s not like you need to pay for all those servers and storage… right?
Best I can do is three over worked tech drones in a trenchcoat who’ll work on this project in their spare time.
What about 2 Raspberry Pi 3s?
Honestly? Hundreds of billions (maybe trillions) of dollars to invest in engineering, infrastructure, staffing, creator incentives, marketing, etc.
You’re talking about taking on a household name that has 2.75 billion monthly active users. That’s more than a quarter of the humans on this planet, and probably some in space too. They make over $50 billion dollars a year to make sure their competition is absolutely crushed.
Realistically though? Enforcing antitrust laws.
That people becomes willing to pay for the services they need - and that someone NOT greedy makes that service…
Coordinated attack on multiple data centers.
By whom? Even the largest botnet in existence wouldn’t be a blip on their radar. I used to work as an SRE at Facebook. The egress graph didn’t even show ddos attacks. That’s how insignificant they are at that scale of infrastructure. There were things that affected the egress graph, like the superbowl.
By whom?
There are only five countries that have enough nukes to take out all YouTube data centers at once: USA, Russia, UK, France and China.
Ok, fair enough. I don’t know exactly what effect nuclear weapons would have on the egress graph, but I’d imagine it wouldn’t be good.
A big bunch of money(trillions)
I believe the biggest hurdle (even bigger than the mountain of tech and capital needed) is people having the will to try to migrate to something else.
Everyone talks big about stuff beong crap but almost nobody puts an ounce of effort into trying, even when there are viable alternatives, the slight amount of discomfort of actually making the switch is enough of a deterrent to actually do it.
Think of when everyone were doing that reddit blackout where they would solemnly swear they would come back after pretending they would abandon it, lemmy existed, forums existed, nothing was one-to-one with reddit so barely anyone actually did leave or even tried replacing it with something else.
Engaging algorithms is the reason people stay on proprietary services. Fediverse users love to shoot themselves in the foot talking about how algorithms and AI sucks then wonder why people don’t want to switch
I think human lazyness as you say is the biggest obsticle. My kids school sent out a Google form the other day to collect family stats. A google form requires a google email and i had degooglised the family 5-6 years ago. My wife was concerned that not providing the info would singke out my son but i was determined to not fill that form, thus passing my data to Google. So i called up the school and sought a second method. The school secretary eventually agreed to record the info over the phone. I lost 30 mins of my day. Basically, if we don’t have the will to mske change, change will never come. But then we have no right to complain, do we?
Engaging algorithm, no paywall, funding or monetization strategy, viewpoint diversity, and a large amount of people already using it
YouTube is popular because it gives creators a way to get paid. So a new service would need to offer some of that.
Ease of use, content creator tools, wide reach, free, less ads, more reasonably placed, no tracking, donation buttons, I’m sure there’s more.
A Great Firewall so that domestic video hosting services can be developed.
52% of the current Youtube viewers to choose another service
Oh, so just 1.4 billion people.
That’s how monopolies work yeah.
