As we approach the U.S. election, misinformation and abuse are on the rise and more people are fleeing X in search of something safer and more trustworthy. This is an incredible opportunity for the #fediverse but only if we are able to maintain and expand the effectiveness of our #moderation as we scale.
In this latest episode of Dot Social I go deep with two well known trust and safety experts who are advancing decentralized moderation for the fediverse: @samlai.bsky.social and @jaz. We discuss strengths and weaknesses today as well as what the future could hold.
Check out the conversation on our PeerTube instance or wherever you get your podcasts.
@mike can't wait to listen when I have time. I've been gravely concerned with the recent discussions not mentioning misinformation as a HUGE problem when discussing limiting replies.
It's very concerning and reckless and my concerns have gone ignored. I hope you touch on it here.
Always appreciate your opinions, Mike.
@BeAware yes it's super important we build in a way to protect people from bad actors who bomb their replies or quote post malevolently.
@mike definitely. Though, misinformation should be discussed as well. Especially in an election year. I haven't had the chance to listen yet. Will do tonight.
Cool! @samlai.bsky.social I've really appreciated your work on securitying federated platforms, looking forward to hearing more!
@mike @miaq is there a transcript and/or set of internal links to different segments?
@errhead @jdp23 @samlai.bsky.social @mike @jaz I’ll be adding a transcript today to the podcast version (where there is a field to add it) and I’ll talk to our flipboard.video instance owner about this feature. I wasn’t aware we were on an older version of PeerTube, so thanks for letting me know! The chapter list is on both PeerTube and YouTube in the video description.
Ooops, how embarassing, I hadn't clicked on the "see more" for the chapter list! Thanks all for pointing me to what should have been obvious to me!
@jdp23 @errhead @samlai.bsky.social @mike @jaz No worries, Jon! I've also just added the transcript. It's imperfect (AI!) but I did do some cleanup so I hope it's still useful https://dot-social.simplecast.com/episodes/samantha-lai-and-jaz-michael-king/transcript
Thanks much, greatly appreciated! @miaq
@mike @samlai.bsky.social @jaz
Listening to this now, another great episode!
@mike @samlai.bsky.social @jaz
will listen but so far I'm not sold on moderating for misinformation being a good idea. absolutely for harassment.
@wjmaggos @samlai.bsky.social @jaz what are your thoughts on why moderating misinformation might not be a good idea?
1. adds a lot to admin workload
2. inherently political and divisive
3. makes choosing a server more difficult
4. impossible to do well
5. we are the open web, not an editorial page
I'm a dissident lefty. the culture here imo is very establishment left. Dem party basically. my hope here is to have the greatest public forum ever but we just had people very upset NYT was calling Biden too old post debate. if we were more diverse, bringing that up pre debate could easily get moderated away.
1. admin & moderator workload is real. reducing harms to users is real.
2. moderating false information which may mislead or increase likelihood of harm is not inherently divisive, and is sometimes required by law
3. only for people who want to consume misinformation
4. impossible to do perfectly, but certainly able to be done well
5. the server you administer publicly blocks 2 servers. this suggests you, along with most service providers, have applied editorial standards.
I'm not arguing for consuming misinfo. I block for harassment. And yes, misinfo can lead to real harm.
But the way we best figure out what's true is dialogue and sharing good journalism/science. Not admins being referee. What I fear they will usually do is take the current consensus position, with a bias towards their user base. Alternative but sometimes correct views will struggle to find a place they can be posted that is well federated to go viral and correct misunderstanding.
>I fear they will take the current consensus position
Some will, some won't. Some will focus not only on misinformation and harassment, but also on the numerous other issues that also warrant intervention to reduce harm.
Here's 34 of them: https://connect.iftas.org/library/iftas-documentation/shared-vocabulary-labels/
You are free to operate your service exactly as you please.
So is everyone else.
1/2
@mike @samlai.bsky.social @jaz
Absolutely loved this! The work Jaz & IFTAS are doing across Fedi blows us away.
@mike @samlai.bsky.social @jaz
Great episode!
@mike @samlai.bsky.social @jaz
what a great talk.
I am a little concerned regarding your discussion of the sharing of Moderation incidents. It seems like ya'll would like it to be "exclusive" for some reason. Though, I believe every admin should have the same tools available to them.
If I don't know about an incident which might cause me to want to block an instance, I can't block them.
Every admin should be able to share their experiences with each other easily enough so that we can all make informed decisions on what is best for our instances.
I do agree that Fediblock is a shit show.
It needs to be more evidence based. But the info should be available to all admins.
I don't want to feel left out just because I'm on my own instance by myself. I need to moderate for myself and want to help other admins as well with my own experiences.
@BeAware @samlai.bsky.social @jaz Thanks! I don't think we want moderation to be exclusive. It's more about trying to establish a place where a) moderation professionals can collaborate and b) moderation decisions can be made in a way which don't invite abuse of the individuals making the decisions.
I totally agree that moderation decisions should be evidence based and publicly available to other admins.
@mike @samlai.bsky.social @jaz Thanks for clarifying this, Mike.
Much appreciated. Love the work everyone is doing, and can't wait to see where it leads.
It's a bit of a worrisome, yet exciting time here on Fedi. I trust the leaders in this space will get it right eventually.