4 ms·

Can we know for sure that they "didn't actually carefully review anything"? Likely thousands of accounts are auto-flagged every minute(second?), and for permane

by danso 4y ago

Can we know for sure that they "didn't actually carefully review anything"? Likely thousands of accounts are auto-flagged every minute(second?), and for permanent suspension (as opposed to the much more common "locked because of suspicious automated activity/bad tweet"), maybe there's a process in which a report listing the suspicious signals is glanced at by a human auditor. I'm talking like seconds of review, e.g. the report says "In 5 minute span, account tweeted from a known U.S. AWS server and a Europe AWS server, etc. etc. Confirm suspension", and the human clicks "Confirm" and moves on to the next hundred reports to "review"

tablespoon 4y ago
> Can we know for sure that they "didn't actually carefully review anything"? We can't, but that's part of the propaganda game. > ...maybe there's a process in which a report listing the suspicious signals is glanced at by a human auditor. > I'm talking like seconds of review, e.g. the report says "In 5 minute span, account tweeted from a known U.S. AWS server and a Europe AWS server, etc. etc. Confirm suspension", and the human clicks "Confirm" and moves on to the next hundred reports to "review" That's actually a good example of not being careful.