Is this conjecture true or false?

Any computing system that goes from mostly closed and permissioned to a mostly open and permissionless system will see its compute workload go from 1% security assurance to 99% security assurance and therefore need to 100x the size for the same available functional bandwidth.

Or 10-90 to 90-10 or similarly dramatic phase change.

Delver at contraptions.venkateshrao.com, running summerofprotocols.com, host of /gloom ☀️

Concretely, to go from a system designed to serve say 5 friends to 10 random strangers you have to scale the infrastructure 10x-100x

“Open” is a “now you have two problems” solution to any problem but in a good way that’s worth taking on if you have resources to waste

Governance Public Utilities for @zksync at @thezknation

I wonder if this translates to buildings / bridges and airports

Sonething like that…for eg iirc most of the EF’s budget can be seen as going towards Ethereum infrastructure security assurance

This feels correct but I don’t think so. There’s elegant security solutions that enable broad functionality, but maybe it’s more about energy input into the system (eg security tests and design?) and that’s what you mean by compute workload?

Directionally right, the numbers seem overblown. IMO you could safely approximate that by human transaction costs. Not quite zero but low intrafirm. 50% or so (for legal costs; startup budget numbers flew by couple of days ago) interfirm.