When you pay your water bill, you aren’t just paying for the upkeep of the pipes that brought the water to your house – you’re also paying for the production of that water. The internet should be no different.
Besides paying a fixed monthly cost to your ISP for the physical connection, there should be a tiny monetary amount – a fraction of a cent – attached to each HTTP request you make, that can go towards covering server costs. Currently sites have no choice but to pay for their upkeep with advertising. Replacing this with direct payments would drastically curtail the data broker and surveillance industry that currently lives off of it.
How server costs would be measured, and whether sites would be allowed to charge a premium on top of that (eliminating paywalls, but also making web browsing a much more price-weary activity) is up for debate.
But currently using the internet is like paying for a car, without paying road tax.
I agree with most points but IMO CDNs are not just a potato, it is a ton of potatoes all easing the load on your server. The individual endpoint is not much but one endpoint does not a CDN make. You’d need the network to be able to lower the load on the server significantly (depending on the data type and caching method of course, but generally I’d say this is true).
Also agree with data caps in the current age. In the early mobile internet transfer was expensive but now it isn’t. I cannot put words on how grateful I am for being able to have no data caps on 1Gig (up to 2.5Gig with same carrier, just in a location where they have upgraded the equipment) as well as no data caps on 4G/5G mobile.
From the perspective of the provider yes it’s millions of potatoes. But from the perspective of the consumer, it’s one potato. Their HTTP request is served by the dumbest, cheapest piece of equipment feasible. The fact that there is a complex network of millions of them only means that each HTTP request made is guaranteed to be served by the closest potato.