The main problem with this idea that is probably the only thing preventing it from happening is that even in 2023, this is still too expensive. You can't run al

The main problem with this idea that is probably the only thing preventing it from happening is that even in 2023, this is still too expensive. You can't run all of a user's browsing sessions through a system like that for the ~$20/year income advertising generates on average. (A reminder of the drum I like to beat; these companies do not make anywhere near as much money per person as people tend to think. ~$20/year is a reasonable estimate of the higher end based on Facebook and Google earnings reports; companies that are not Facebook or Google make even less. That's what they're breaking the internet for. $20/year.) If you don't take the raw income from ads but instead look at the profits, the bar is even higher. It's a minimum of an order of magnitude too little money to pay for that (and I think a very solid case could be made for 2 orders of magnitude), so it's going to be some years before it would be feasible even at the current level of web capability. (Add more 3D and heavier weight UIs based on WASM and some raw canvas or other and it gets even worse.) So if it is the future, it's still far enough away that I wouldn't try to found the startup today; you're still way too soon. As big as the cloud is, it is absolutely dwarfed by the amount of computing power not in the cloud. It's not even close. You'd have to get people to pay for it. For that, there has to be something very compelling. I can't think of what that would be. Apparently neither can anyone else, because otherwise like I said this would be a present reality and not a blog post. However, this is just an analysis of this idea from the 2023 perspective. I'm not saying it isn't the future. It's clearly a future a lot of players want. I'm just explaining why, in light of that, it doesn't already exist.