Lots of comments about internet infrastructure being decentralised. In my opinion, there is a difference between decentralised architecture and decentralised control.
For instance, a single company could deploy a decentralised system. It may suit their use case better, be more fault tolerant, whatever. But they still control all of it. If they want to take the system down, there's nothing about it being decentralised that prevents them from doing that.
The same thing happens with decentralised internet services. Yes, many services are decentralised. But it ends up being easier letting a small number of groups control the majority of the nodes. This is just centralisation with extra steps.
Take Cloudflare for instance. It's decentralised, there are servers everywhere, it makes the internet for resilient and faster. But if Cloudflare decided it wanted to take a piece of content down, they could do so right away (cache times notwithstanding).
Solving the centralised control problem is very different to creating good decentralised architectures. Most tech people focus on the latter and think in doing so they have solved the former (they haven't).
For instance, a single company could deploy a decentralised system. It may suit their use case better, be more fault tolerant, whatever. But they still control all of it. If they want to take the system down, there's nothing about it being decentralised that prevents them from doing that.
The same thing happens with decentralised internet services. Yes, many services are decentralised. But it ends up being easier letting a small number of groups control the majority of the nodes. This is just centralisation with extra steps.
Take Cloudflare for instance. It's decentralised, there are servers everywhere, it makes the internet for resilient and faster. But if Cloudflare decided it wanted to take a piece of content down, they could do so right away (cache times notwithstanding).
Solving the centralised control problem is very different to creating good decentralised architectures. Most tech people focus on the latter and think in doing so they have solved the former (they haven't).