Re:XCheck at Meta: Why it exists and how it works
by DisjointedHunt on Sunday October 16, 2022 @01:51PM
(#33225597)
This is par for the course at Facebook (ahem, Meta). The blog post completely misses the point about why such a system is bad.:
- Facebook created a differentiated system of treatment in its enforcement systems.
- Can this be used for preventing mistakes? Sure. The problem of "mistakes" in Facebook enforcement is much deeper and is rooted in systemic apathy (But thats for a different topic)
- Facebook allowed access to this system in a non transparent manner and to this day does not maintain transparency on who's on it, allow external review or submissions to this very powerful shielding logic while under criticism for SELECTIVELY allowing access to partisans.
- The other angle of criticism is that such a badge in Integrity systems is inherently used in classifiers or dumb rules that take down anything reported by these special accounts.
Writing a long ass blog post speaking about generic issues with Integrity systems without acknowledging the criticism on transparency and access to a system of such power is only making things worse. With or without official blessing, if some random engineer can simply push a rule saying "Take all reported accounts by shielded accounts in India down, no questions asked", that circumvents the official takedown processes that are simply ornamental at this point.
tl;dr:
Facebook has a non transparent system of power and its not the specific code word for this tool that is bad. Allowing selective access to such immense power without transparent review and reporting is always susceptible to corruption and favoritism.
- Facebook created a differentiated system of treatment in its enforcement systems.
- Can this be used for preventing mistakes? Sure. The problem of "mistakes" in Facebook enforcement is much deeper and is rooted in systemic apathy (But thats for a different topic)
- Facebook allowed access to this system in a non transparent manner and to this day does not maintain transparency on who's on it, allow external review or submissions to this very powerful shielding logic while under criticism for SELECTIVELY allowing access to partisans.
- The other angle of criticism is that such a badge in Integrity systems is inherently used in classifiers or dumb rules that take down anything reported by these special accounts.
Writing a long ass blog post speaking about generic issues with Integrity systems without acknowledging the criticism on transparency and access to a system of such power is only making things worse. With or without official blessing, if some random engineer can simply push a rule saying "Take all reported accounts by shielded accounts in India down, no questions asked", that circumvents the official takedown processes that are simply ornamental at this point.
tl;dr:
Facebook has a non transparent system of power and its not the specific code word for this tool that is bad. Allowing selective access to such immense power without transparent review and reporting is always susceptible to corruption and favoritism.