The ad model of the internet is very fragile. Some even question if it works for the most part currently. Most sites that use it are speculative including the rather large ones like youtube. The fact that this speculation has been able to go on for so long with no one saying, um, yeah, ad model doesn't work really is a testimate to how far out the time preference window can be extended with the Fed's involvement.
But my other broader prediction, though less specific, of which this one depends, is that many services of many types are going down or will be reduced, or will be less reliable. This is a necessary consequence of the depression that is coming. If it will be hard to maintain all of the services involved in food production and distribution, then some of the more luxurious and pointless services are definitely going to go. A period of self reliance is coming, or at least less consistent external services is coming. And that's going to include the service of watching videos for free that you didn't host yourself. It's going to be hard for these companies to continue to hire the programmers needed to maintain these systems and continue with active development needed to convince themselves that their is some probability of that service may be a net gain.
Programmers are expensive yo. And nobody is going to be paying them for some speculative shit. And why keep a project alive at all if it only costs you money and you have no plans leave that portion of the project's life-cycle.