I think client side heuristic spam filters would go a long way. Also just a working block command is necessary
nostr:nprofile1qqs9pk20ctv9srrg9vr354p03v0rrgsqkpggh2u45va77zz4mu5p6ccpzemhxue69uhk2er9dchxummnw3ezumrpdejz7qgkwaehxw309a5xjum59ehx7um5wghxcctwvshszrnhwden5te0dehhxtnvdakz7qrxnfk to build a good WoT, you will need to use the existing outbox model to find the trust drivers of the network, regardless of how you define what "trust" means for bitchat users.
Discussion
We have been down this road on nostr. Its not enough.
Sure, but does it not help at all? What's wrong with having multiple layers of spam blocking? It doesn't have to be effective, just not useless. Even removing 30% of the worst spam is better than not. And this is much more relevant to bitchat than amethyst.
Spammers are scripts running on desktops, sometimes with GPUs, wired to a power outlet. There is no way battery-based systems can beat them. All you are going to do is to waste people's batteries. We implemented pow on amethyst, but it was clearly draining way faster than what we though it would.
So have people mine the PoW tokens on wall power to use later on battery.
How many times do I have to re-explain basic concepts like PoW as a random guy who can't code, in discussions with top nostr devs?