@clhendricksbc @Gargron To be honest though the question is whether and how something like the fediverse can defend itself against these sorts of uses. It seems like this open environment is extremely susceptible to this kind of attack.. Perhaps its a vulnerability and intrinsic part of social media in general..
@amsomniac @clhendricksbc @Gargron Yes the culture is a helpful buffer but for how long? If we can take the 'corporate takeover' of pawoo/pixiv as a reference (documented in detail here http://ansuz.sooke.bc.ca/entry/335 ) what happens to the culture of the fediverse if it suddenly doubles or triples in size? I think intuitive ways of using multiple accounts (pseudonyms etc) would be helpful as a start.
@sakrecoer see above. I think same applies to the media literacy argument..
@rra sure thing! Knowledge is probably the best tool against any propaganda really. At the end of day, digitalized and botnet troll enhenced or not, that is what it is: propaganda. My tinfoil hat whispers that the CA breakdown is part of it. I mean, CA have been informed that their office will be searched, leaving plenty of time to clean up.
@Gargron @clhendricksbc @amsomniac @sakrecoer by the way these concerns can come across as negativity but its really the opposite, a love for potential of the platform. I think its important though to have the difficult discussion on how to make sure 'we' don't implement tech bro ideas with more ethical (decentralized, etc) tech, just to repeat the mistakes. Rather how can we use this to shift to a completely different model?
@rra @Gargron @clhendricksbc @amsomniac I like this thing: https://datproject.org/ And i can't wait for this to happen: https://enoki.site/
@amsomniac @rra @clhendricksbc @Gargron Thatβs still a issue that they will only get caught if they run the malicious parts on the same domain as the main instance or make themselves obvious with excessive follow bots.
@duck57 @rra @clhendricksbc @Gargron what does a system resilient to untrusted instance admins/followers look like? is that worth trying for on a platform like gnu social/mastodon/diaspora/etc or can we just patch this in culture by using encrypted group chats in smol circles and aliases with exclusively public posts in the larger, public sphere?
@amsomniac @rra @Gargron Yes, good points. Also part of the issue is feeding people particular posts/stories based on that data in order to influence their political decisions (among other things). And so long as one is on an instance that doesnβt have that capability then at least that part wonβt happen. Hereβs to getting posts from people you follow as they post them (if they allow you to), in chrono order, with no weird algorithms.
@rra @clhendricksbc @Gargron good point. Probably the best tool against these attacks is better education regarding the implications and consequences of sharing information about oneself. Ironically, in my 1st internetclass in 1996 the teacher emphasized the importance of never using a real name.
@rra @clhendricksbc @Gargron One aspect is the creepiness of the real name policy. If you don't have to identify yourself or your friends, that's one way of resisting detailed profiling by corps and political parties.
@rra @clhendricksbc @gargron This. As soon as you make private data available to friends, whether on a centralized system like Facebook or something more decentralized, you're vulnerable to your friends being hoodwinked into running malicious software.
@rra @clhendricksbc @Gargron I take some solace in the idea that since posts are public and easily accessible, no one would want to buy that data, and if anyone made a third party app that tried something like that with private posts, or ran a malicious instance, or even tried to sell demographic data to sell ads, it seems like the culture here wouldn't tolerate it and that instance would be shunned.
hiveway was an interesting experiment, I expect we'll see more weird things like that.