Post #4066635
2026-07-24 15:07 UTC
RE: https://mastodon.social/@appleinsider/116975524473168017
I suspect that in the age of AI, trusting that sources are real humans will be a based on a social verification graph. I can vouch that people I personally know are human. They can vouch for people they know. If there were a way to compute the social confidence score of any given poster, it could filter out a lot of bots.
You’d think meta would be well-positioned to implement something like this. But instead they’re just going to do the idiotic face scanning thing.
Replies (2)
-
@siguza@infosec.space 2026-07-24 15:41
@shac@ioc.exchange they have no incentive to do the right thing...
-
@ewhac@mastodon.social 2026-07-24 22:22
@shac@ioc.exchange One problem with the social graph (or Web of Trust, as it was called by PGP/GPG advocates) is that The Bad Guys can stand up billions of fake identities that all vouch for each other (q.v. "link farm"). If you bias simply for link quantity, The Bad Guys win. Yes, the graph will be isolated. So bribe (or threaten) a few real people in desperate straits to vouch for some fake identities (or pwn their Windows machines to do it). Poof! The fake island now has "legit" links.