Korva is a scroll of the internet’s best ideas.
200 accounts pushed data center backlash
The disclosure came Thursday night, after X removed a lot of bot accounts, and Axios ran it under the headline China Is Secretly Fueling America's Data Center Rage.
A small number of accounts can shape public opinion by pounding influential people in their replies and building what readers wrongly take for consensus. Here the accounts told influential people that Americans dislike data centers, and that was the picture the bots supplied.
A reply-filtering bot that flags any account under a year old or under 200 followers and surfaces what it tweets shows the same pattern: constant brigading from fake accounts. The same tactic appeared during the elections, on both sides.
How does astroturfing work?
Astroturfing is a coordinated campaign that imitates grassroots support. Operators run many accounts that post, reply, and amplify one another so a view looks widely held. Research on social proof, including a 2006 experiment in Science by Matthew Salganik, Peter Dodds, and Duncan Watts, found that people's choices shift toward options that others appear to have chosen.
How do platforms spot bot accounts?
Platforms look at signals such as account age, follower counts, posting rhythm, and whether many accounts share devices, networks, or near-identical text. Coordinated accounts often show up as tight clusters in the graph of who replies to and amplifies whom. Researchers at Indiana University built a tool called Botometer that scores accounts on more than a thousand such features.
Why do replies shape perceived opinion?
Replies appear directly beneath a popular post, where large audiences read them. Readers tend to treat what they see there as a sample of public opinion, even when it is a small, unrepresentative slice. Social psychologists call the belief that your own view is more common than it is the false consensus effect, described by Lee Ross and colleagues in 1977.
More from All-In with Chamath, Jason, Sacks & Friedberg
