Nostr solves this with unique nPubs, the users just have to build reputation to their addresses. That's better than paid verification. This new era of Large Language Models do however make spam a tricky thing, because we now have clever spam. (paid relays & zap infused notes might be the best bullwork)

Deepfakes with the purpose of influence, is an unsolved problem worthy of contemplating. Don't believe everything you see on the internet is easy to say, hard to implement. But a good rule nevertheless.

I'm sure we will see interesting solutions from the nostr eco-system on these problems. It's like a jungle, with a greater evolution modifier than anything we seen in this domain before.

thx for the video clip share, good one.