> These seem like solvable problems with decentralized systems. In both cases, someone has to go through the work of manually identifying the bad content, right

> These seem like solvable problems with decentralized systems. In both cases, someone has to go through the work of manually identifying the bad content, right? In a centralized system, that's someone working for the system - in a decentralized system, that's a random user. It's solvable in the same sense that email spam is solvable. Much like adblockers, spam is "solved" by re-centralizing. > If the volume of spam is truly extreme, then what's to prevent you from having distributed blocklists that are fed by automated processes (as opposed to manual additions), and users just subscribe to the ones that they trust? Most users are unequipped to evaluate that and disinterested in putting in a bunch of work to defend themselves against the flaws of the system at hand. Like adblockers or email, most members of the general population want it to be easy, automatic, and require minimal effort beyond clicking the button that gets them going. Users generally want things to work for them. Investing deeply in protecting themselves because the system's designers didn't consider abuse is rarely towards the top of the priority list. People like things that just work, and the further a thing is from that the more adoption will struggle.