yup. how is that not obvious to anyone is beyond me… some of those workers have contracts that would require amazon paying severance in case they would just fire them like so many other companies do. better make them leave on their own.

it could and did happen. I remember reading about 6 year old kid getting lost at the airport gate (i think it was in germany) only to be discovered when they finally looked for their parents in italy. just went away while parents were preoccupied with something else and looked like someone else’s kid while boarding the plane. as a parent travelling with small kid I can totally see that happening anywhere - kids can just walk anywhere with no questions asked as long as they are next to an adult who looks like they could be their parent.

that is an interesting idea. still… you can create an account (or have a troll farm of such accounts) that will mainly be used to trust bots and when their reputation goes down you throw them away and create new ones. same as you would do with traditional troll accounts… you made it one step more complicated but since the cost of creating bot accounts is essentially zero it doesn’t help much.

you are right - it doesn’t have to be one or the other… I just assume that for social media to work as I expect I don’t know most of the people on the platform. given that assumption and the lowering price of creating bots and ability to onboard them I expect that eventually most of the actors on the platform will end up being bots. people that write them are often insanely motivated (politically or financially) and creating barriers for them is not easy.

I was thinking about something like this but I think it’s ultimately not enough. You have essentially just two possible ends stages for this:

1. you only trust people that you personally meet and you verified their private key directly and then you will see only posts/interactions from like 15 people. the social media looses its meaning and you can just have a chat group on signal.

2. you allow some length of chains (you trust people [that are trusted by the people]^n that you know) but if you include enough people for social media to make sense then you will eventually end up with someone poisoning your network by trusting a bot (which can trust other bots…) so that wouldn’t work unless you keep doing moderation similar as now.

i would be willing to buy a wearable physical device (like a yubikey) that could be connected to my computer via a bluetooth interface and act as a fido2 second factor needed for every post but instead of having just a button (like on the yubikey) it would only work if monitoring of my heat rate or brainwaves would check out.