[randomran01234]

Is there any evidence that almost 100% of transactions use Infura? There are several other popular nodes as a service since the cited 2018 article was written: Alchemy, Quicknode, Moralis. Anybody running a full node, including home stakers, can point to their own RPC endpoint.

When one of the popular nodes as a service goes down, many dApps and users will need to change their endpoint to another. I think the author might be overstating the concern of this centralization.

The solution to RPC centralization is well known: light clients. It is on the developers roadmap but obviously they are prioritizing proof of stake and scalability.

[davidgerard]

In 2012, you could presume cryptocurrency users had some idea what, say, cryptography was.

In 2022, you can't presume a damn thing.

As I noted in another comment, users could do all manner of things they don't do. In practice, defaults are overwhelmingly important.

When Infura blocked all of Venezuela in March, due to an error in over-applying sanctions against Russia, some wondered if the Ethereum network was being blocked. No, it was just the app everyone depended on for almost everything.

In the fabulous future, I'm sure it'll be great. In present-day reality, things are as I described them.

Nobody actually cares much about Infura or other points of centralisation because approximately 100% of Ethereum participants in 2022 have a firm and unshakable ideological commitment to being in it for the money, on an extremely short-term horizon.

[randomran01234]

So your main argument is that people are too stupid to understand that Ethereum is not the same as Infura, and so when one goes down you may as well call them both down? Lol.

Lack of education can be fixed. You can even help: instead of spreading poor information, try to educate your readers about the difference between a node as a service and liveness of the blockchain.