[Groxx]

There are tons of sybil-proof systems if you don't include signal from sources by default, but instead opt them in by hand. E.g. use a web of trust and then choose who you trust. It doesn't matter if there are trillions of accounts you don't trust, because you don't trust them.

[blamestross]

Which isn't a web of trust. it is just an "allowlist". Humans are vulnerable to sybil attacks too.

[Groxx]

An allow-list with transitive trust is a web of trust. And I said "use a web of trust", not "use an allow-list", because I really did mean "web of trust".

And sure, they can be, if they adopt patterns that allow it. I can also find plenty of counter-examples. I don't think "humans are less vulnerable to sybil attacks than automated systems" is a weakly-defensible stance at all.