this bit was the cherry on top

Omg it’s too much I’m dying. Thanks, comrade. Thank you very much. Double-nice good find.

Did the article in question make it peer review? If so, wow, standards are lower than I expected for those who support the empire.

Even down to basic logic. If a ‘tankie’ community forms by ‘siphoning’ users from other ‘far left’ online communities, can one really conclude that the number of ‘tankies’ has grown? Surely that’s only evidence that Marxists move around (let’s not mention that one reason is they keep getting booted from mainstream platforms, while extreme right wingers get algorithm’d to the top of everyone’s feed).

The number of ‘left extremists’ didn’t go back to zero e.g. when Chapo Trap House got quarantined. They just couldn’t congregate on they sub anymore. So they didn’t magically grow again when Hexbear or GenZedong were created. They just moved. It seems obvious that numbers have grown. But that’s not directly supported by the argument.

Can’t wait for the next project where they follow up with their suggestion and compare left wing ‘extremists’ with right wing extremists:

“Left wing extremists don’t commit terrorism and are frequently found arranging protests on behalf of the most vulnerable people. They hold outrageous views such as the right to basic health and safety precautions at work and the right to access a doctor when ill. If they are involved in violence it is as subject to the violence of the state or of right wing extremists. Nonetheless we conclude that these people are dangerous. Because they make our brains feel mushy with cognitive dissonance by pointing out that the good guys^TM get up to some downright bad guy behaviour.”

Edit: I’m obsessed. It gets worse:

We’re not learning if something is racist or not – we’re learning which is more racist. Who knows what we’ll find … [?]

Jfc. Only someone who never has to worry about being subject to racism could say something like this with such hubris. What a redacted. At least they’re honest that with liberals—the majority of most online spaces—it’s only a question of degree.

If anyone wants an example of how technology, including AI, is designed to be intentionally racist, this is it. The results of this project could be a machine learning moderation algorithm that doesn’t try to prevent racist speech but only to moderate the more racist of any two utterances. A deliberate intention to build into AI the exact kind of techno-structural racism enacted by liberal humans up to now. Liberals just can’t help themselves, can they?

Holy shit, they seriously use word “tankie” in a fucking research paper?

pretty funny way to scam the NSF out of half a mil