I don’t think there are many (if any) who understand Ethereum who aren’t concerned around a large percentage of validators running in 3rd party clouds. On the

I don’t think there are many (if any) who understand Ethereum who aren’t concerned around a large percentage of validators running in 3rd party clouds. On the bright side, if AWS has a large outage or simply shut down any nodes running validators, all of the validator operators would need to set up somewhere else quickly to avoid large inactivity leaks (which grow in relation to the number of offline nodes). The worst outcome would be that people would move from AWS to Azure and simply move the problem elsewhere, but ideally it will solve itself as validators get penalized more for supporting centralization of certain aspects of the protocol. The protocol has a built-in mechanic to dissuade centralization that hasn’t had an impact yet since there hasn’t been a large-scale outage affecting a single host. Once that happens, people will get a bit more cautious.