The difference is in the past most discourse was by mouth - impossible to censor and eavesdropping was costly because you needed a person to do it. These days

The difference is in the past most discourse was by mouth - impossible to censor and eavesdropping was costly because you needed a person to do it. These days more and more discourse is online - heavily censored by platform rules which are much more strict than laws, partially because it makes the moderator's job easier to err on the side of caution, partially because it teaches people to self censor which then extends into real life - censorship can be automated and even invisible (shadowbanning, fake degradation of QoL, ...) - eavesdropping is also automated and omnipresent The real danger of AI or LLMs these days is not they machines will want to kill us but that powerful people will use them to mass-profile everyone and stop any dissent from spreading. Do you think China will have a revolution and restore democracy? How? How will people organize? And every currently democratic state will be the same - democracy only needs to fail once and it will be almost impossible to restore. --- EDIT: Even word of mouth is in danger, sufficiently powerful organizations are or will be able to access microphones near you. Or cameras for lip reading. Again, en-masse, automated by AI.