Meta & Children's Online SafetyChat: You'd think we JUST started working on integrity
Thread View
████ name redacted. Different bars may be the same person or different people.
quoted in an excerpt in the archive.
Friday 9 June 2023
11:22████Kabbabe Exhibit 2/4/26 Jenn Dunn RMR CRR so disappointing but none of it is surprising Melissa
11:22████I'm not surprised we have bad hashtags. our search safety team is a few employees who look at a list of popular hashtags every week to decide which to filter out: (it's super low tech Melissa
11:22████we also do very little to make sure handles are appropriate which is so creepy and problematic Melissa
11:23████Damn, Stamos finally figured out the most effective path to influencing Zuck Melissa
11:23████it's a really well written article, really frames and contextualizes the issue quite well Melissa
11:23████The fact that we recommended the pedo account that activist viewed once to all her followers is really disappointing Melissa
11:23████Content moderators not enforcing the rules is such a big problem. Could be the biggest problem on right now Melissa
11:23████It's a really big problem that you view one media and then we start recommending many like it. We gotta solve the rabbit hole issue Melissa
11:24████I feel like this should be mandatory reading for all of Relevance on how recommendation systems can go.
It's really fair and well written by people who clearly understand the issues well
11:25Tomas Kabbabedang
11:25Tomas KabbabeI was not aware that this was an issue
11:26Tomas Kabbabethis = the WSJ article?
Melissa
11:26████yes
11:27Tomas Kabbabecan you tell me more about this?
Melissa
11:28████we think rater recall could be as low as 50%, and the non-rec labelers only know English... they label 0 media/week as hate speech, bullying, or suicide and self injury even though we know the prevalence is higher. and they frequently get easier topics wrong that don't require language comprehension, like disturbing content https://fb.workplace.com/notes/812741413296770 Melissa
11:29████this is a more recent check on rater recall that paints a rosier picture, but with the methodology
11:34Tomas Kabbabeman
11:35Tomas KabbabeYou'd think we JUST started working on integrity with issues like this one
11:35Tomas KabbabeClearly this is hard, I don't want to dismiss the good work from the integrity teams strongly disagrees PLTF EX exhibitsticker.com
11:35Tomas KabbabeBut it feels like we're so behind where we should be Melissa
13:49████I mean you see the chaos of the way products are shipped at Meta Melissa
13:49████we don't do comprehensive, holistic overviews often enough Melissa
13:50████we don't hold enough people responsible for metric monitoring. we only staff teams when we want to build, not maintain Melissa
13:51████the central integrity org is pivoting to focus on "voice" aka false positives and overenforcement. their other big new focus is "integrity value": trying to justify how integrity work adds value to the business. as a result, they've stopped investing in SWOP, which is our Meta metric for content violations!
Melissa
13:52████integrity is an adversarial space. there's millions of people who will profit from finding ways to exploit it, so they're always trying. product teams change the way things work all the time and don't move in lockstep with integrity Melissa
13:53████and there's just SO MANY TYPES OF HARM. it's hard to think of all of them for everything always. just because there's no harm on an issue type or surface today doesn't mean there won't be harm on hashtags tomorrow, so someone makes the call that it's not a priority and then boom, things evolve Melissa
13:59████and non-rec is owned and operated by the small team on RI, so we don't benefit from Meta's best practices on making sure raters actually label the violating content as violating