It’s interesting because there’s an approach in Nostr being explored so that relays can require proof-of-work. The idea is to make running spam expensive. What I don’t like about it is that it’s adding inefficiency and waste, but it’s hard to find alternative solutions; spam management (spamd) for email on my little mail server consumes a constant 3% of my CPU and 30% memory. I’m paying for that, and it’s wasted CPU cycles already. Shifting the cost to spammers is a great idea.

The Nostr NIP doesn’t suggest this, but having a scaling cost would be handy. Your first message is free, with an increasing complexity per relay cost.

AP could extend to include something like this, but at this point, it would be a surprising development. Although the POW approach has been around for much longer than Bitcoin, it never really caught on with email, and is now associated with crypto so strongly it’s tainted by association.