I'm going to paste here something I wrote a while ago. This is an technological thought experiment and an allegory for fake news and modern propaganda. I use a

I'm going to paste here something I wrote a while ago. This is an technological thought experiment and an allegory for fake news and modern propaganda. I use a technological hypothetical to remove the question of if people have any agency in falling for it and the question of if it is necessary to put up with fake news as a form of free speech. With those two questions settled in the premise, we can think clearly about the core issue of the threat of effective propaganda to democracy. Suppose neuroscientists and neural network researchers found an adversarial image which worked on humans. Not all humans, but at least 15% of them. The technology is developed such that by including a specially crafted banner ad image next to an article you could put an arbitrary thought in the persons mind. This opens up a very serious hack in a democracy. The technology is hard to build, but a nation state actor like China or Russia could easily replicate the technology themselves and use these banner ad images to sway elections. Like a psychological atom bomb, it is just a matter of time before every major player has it. 15% of the population is larger than the margin on nearly every election ever. A democracy which wants to survive only has two choices: 1. Criminalize the adversarial images and censor them from the web. Congrats. You are now in an arms race with one side making better adversarial image detectors and the other making better hidden adversarial images. In the mean time you've built an insanely good machine learning based censorship apparatus which sits on top of all internet traffic. Eventually the technology gets to a point where the cognitive basilisk isn't obvious to the unaffected and people are worried that some images are being censored for their substance. The adversaries intentionally hide the patches in posts / ads about wedge issues like guns or abortion so that taking it down looks like political censorship. While everyone agrees censoring the thought breaking images is necessary, people are rightfully scared about this apparatus being abused. Maybe the cognitive basilisks are de facto censored in the sense of being de-platformed by the media (both traditional and social variants). Maybe they are state-capitalism censored in the sense of de-platforming under threat of regulation if they aren't. Maybe we go full China with it and have a top level federal internet censor. The only real difference between these mechanisms is the technical issue of legality. For our purposes the implementation does not matter. 2. Don't let people in the 15% vote. People are again rightfully scared that it could easily expand beyond 15%. A very small, very educated, very detached from the average class of people (neuroscientists) is now trusted to decide who gets to vote. All of the 15% (and quite a few in the 85%) are suspicious of them abusing this power to forward their personal political preferences. Disenfranchising people based on inherited neurological traits is extremely contentious. The question of if it is the right thing to do becomes the defining social issue of a generation. A political fissure opens up. Some people prefer option 1 because they don't consider the adversarial patches to be "speech" and thus don't think there's a free speech risk in censoring them. Some people prefer option 2 because they don't think it is possible to win the arms race in the first case and so long as we don't let the disenfranchisement expand beyond 15% the sin is forgivable. At the very least they think option 2 is a less slippery slope than option 1. Some people like option 1 and 2 at the same time because just do whatever it takes to fix the problem as fast as possible. The people in the 15% think the cognitive basilisks are just a made up conspiracy by the elites. Of course they think this because the cognitive basilisks told them so. Then comes the disingenuous part of the fissure. Some people like option 2 but their unstated reason is because they looked at how the 15% currently votes and they think disenfranchisement will finally lead to their political party winning. Some people secretly want to keep the exploit in place because their political party happened to benefit from interference in the last election. They don't actually have a solution but instead constantly remind you how risky and undemocratic the two options are and insist we do neither. None of the disingenuous advocates are ever moved by any arguments because their purported beliefs were always just a fig leaf over their unstated wants. The debate over how to deal with this technology takes on an impassioned, repugnant, culture-war tone that we have all grown to recognize by now. What happens next I don't know. What I am proposing is that if cognitive basilisk's exist then there is a contradiction of wants within Democracy. You must pick at most 2: -Everyone gets to vote. -Nothing is ever censored. -The internet has no borders. The only remaining question is do fake news and AstroTurfing bots count as cognitive basilisks?