🤔

1) There are Instagram accounts posting content today that would prompt many Mastodon admins to defederate immediately.

2) Those accounts are not being moderated by Instagram. User reports are closed without action. So their Threads version of those accounts will likely also not be moderated.

3) So many Mastodon admins will defederate from Threads shortly after federation.

So I'm not sure why folks are spending so much time debating whether defederating before the inevitable harm is bad?

I'm not talking about any hypothetical "embrace and extend" fears that may or may not happen in the future. I'm also not talking about deeply nuanced or complex moderation decisions.

I'm talking about basic, moderation 101 decisions that would happen today. Like, the simplest Fediverse moderation case: "How do we feel about transphobic hate posts and calls for hate-based pile-ons?"

An instance that allows this would be defederated by most admins.

@mekkaokereke did you see woof.group’s post about federation with threads? i thought it was a thoughtful exploration of what the pros/cons could be for their community (it talks about whether bigotry on the platform is an issue if it doesn’t result in harassment of anyone in their community and if nobody in on the masto instance follows the problem accounts) https://blog.woof.group/announcements/considering-large-instance-federation
Considering Large Instance Federation

Over the last year Tumblr, Wordpress, Medium, Mozilla, and Meta have announced plans to explore or implement federation via ActivityPub, ...

Woof.group Announcements
@b0rk No I hadn't seen it! But reading now...
@b0rk Well written. Really lays out many of the issues as they see them.

@mekkaokereke the real issue is that the #Fediverse 'manual moderation' will prove impossible at scales of hundreds of millions or billions of people.

This has already been proven by all the central commercial for-profit social media platforms.

Yes, they profit from the bots and extremists traffic, and many of us are skeptical about whether they really want to stop it, but I honestly I believe they really can't, and they have thrown a lot of effort to algorithmic and manual moderation... it just cannot be done very effectively. And, BTW, I think #Meta is actually the best at doing it.

Moreover, one person's extremism is another person's normalcy...

The large Fediverse sites will need to implement some programmatic/algorithmic/hybrid moderation that could talk to the #Meta counterpart or the whole federation idea with them will fail within minutes.

The only viable use case I see (at the moment) are small instances where all the users are prepared to deal with 'as-is' Meta content... they may remain federated with them for the benefit of their users to engage with the Meta crowd... and that could be perfectly fine too.

It is difficult to predict this collision 'out of the blue', but I think it should be possible to be modeled and researched... and I am sure there must be some folks doing it as we speak. 🤔

@b0rk

#Moderation #MastoAdmin

@mekkaokereke @b0rk I think the benefits of broader engagement shrink when you read that essay together with this post on algorithm creep. (No, not a sobriquet for Zuckerberg, sadly enough) https://mstdn.patatas.ca/@smallpatatas/110707812951733786 It seems... optimistic to assume threads.net will always send other AP servers unfiltered, unmanipulated feeds in strict chronological order without algorithmic manipulation of what is seen and not seen and when. This is Facebook's entire business model we're talking about.
small patatas (@[email protected])

Content warning: (Long Post) Threads undercuts Mastodon's core mission with 'algorithm creep'

PATATASTODON

@b0rk @mekkaokereke

I read that post as "Yes, they throw poop, but until we get hit we'll hang with them."

Maybe I'm old-fashioned, but I won't hang with people who throw poop.

@b0rk @mekkaokereke i still don't understand *why* threads is federating. seems like understanding their motives might help predict answers to some of these questions (like, why are they motivated to answer reports from fedi instances?)

@kellogh My current best guess is anti-anti-monopoly stuff. They're already a huge social network, have been known for buying out the competition, and if they do manage to steal Twitter's customer base, they don't want to be seen as monopolising. So, be open with a non-commercial alternative a tiny fraction of the size, and say people still have choice.

Interested to hear of other theories.

@sgf that’s seems like a solid theory

@kellogh i’m no expert but think it’s an effort to make themselves look like less of a monopoly to regulators.

with twitter crashing, and meta having almost 3 billion people on insta, fb, and now threads (covering all of the most popular social media styles), there’s a potential threat of regulators stepping in to try to break them up.

federating can take some of the pressure off, by allowing them to claim “well, you don’t HAVE to use our products, we embrace open standards” while they still get the data they built their business on.

@b0rk @mekkaokereke

@b0rk @mekkaokereke I'm really struck more and more by the fact that the groups which have been rejected by main stream social media and ended up shaping Mastodon, truly have carefully thought all this stuff through already.
I'm not on that server or part of that scene, but I'd be very happy to have a moderation crew and community run by anyone so aware of the issues.

@mekkaokereke This, exactly!
And we shouldn't cut million dollar-heavy corporations some slack because they shoehorned a Twitter clone onto their Instragram thing and had the oh-so-generous idea to make it compatible to ActivityPub.

Quite the opposite: Theyy should use those shedloads of money to hire proper moderators, give them the necessary training and oversight, and make sure that Threads doesn't just integrate with the Fediverse on the protocol layer, but also on Layer 8.

@mekkaokereke

Right?? LibsOfTikTok is on Threads. Do you federate with a service that has chosen to allow LibsOfTikTok an account? It shouldn't take more than a minute to answer that and move on.

@mekkaokereke I read the blog post from woof.group that another responder linked, and one thing it highlighted is that the tools for moderation with Mastodon are pretty limited and will struggle to scale. Individually reviewed reports, and the only response being limited or blocking at the account or instance level. Not much in the way of granularity or scalability there.

I wonder if the tooling could be built to do collective blocklists and community moderation, to supplement the current tools. As anyone with a pihole or ublock origin installed can attest, collective blocklists can make a material difference to one’s experience. Perhaps some of those tools could be adapted to the fediverse, and help alleviate the baby-in-the-bath water problem that is blocking large instances.

@mekkaokereke Federation should be contingent on a clear plan and commitment from Meta to protect fedi users from harassment campaigns. Minimum.
@mekkaokereke (ideally in conjunction with plans from our side for stronger protections, which we should have anyway) https://mastodon.social/@misc/110698437172375433
@misc @mekkaokereke Exactly. Which is why the Fedipact makes sense, since any request of that kind would be met with Pointing And Laughing at the person making such a request. Meta has never and will never moderate to Mastodon standards.

@WhyNotZoidberg

Wildly different incentives!

They moderate to the legal minimum to keep advertisers getting eyeballs. Users are the product.

We moderate to keep users happy because servers are pricey and people want their friends to be happy and help out with the costs.

In a very, very real sense, centralised social media depends on captive audiences not having a better place to go.

@misc @mekkaokereke

@Homebrewandhacking @misc @mekkaokereke

Which is why they tend to ban people for nudity, and let genocide talk slide.

@WhyNotZoidberg
And which is also why I think they will try to introduce protocol incompatibilities, difficulty migrating accounts, interchange/peering requirements, etc
@Homebrewandhacking @misc @mekkaokereke
@WhyNotZoidberg @Homebrewandhacking @misc @mekkaokereke Nah, they ban people for breastfeeding and let genocide talk slide.

@misc from what I've read, Meta actually are intending to have Fedi instances adhere to the rules set by Meta, rather than Meta adhere to existing Fedi rules.

So it isn't really Meta joining the Fediverse, but Meta trying to claim and control the Fediverse.

@81732bit @misc Yeah, that’s definitely how it will go, instances wanting to federate with Meta will probably have to apply and in the application will be an agreement to follow their moderation rules.
@Kyleric @81732bit guess we can’t say we live in the dumbest timeline as long as Twitter still exists
@81732bit @misc Which is funny because anyone who's been using Mastodon for long knows how that's gonna go: "ban breastfeeding but allow transphobia" gives them Parler and Poast but not Mastodon flagship or Art.
@mekkaokereke i’d say two reasons: 1) some people REALLY like corporate social media 2) some people REALLY like to oppose forthright protest

@mekkaokereke threads will federated with gab and vacuum up the connections to those users. Their behavior really doesn't line up with a chance of mastadon acceptance.

If so, the PR blitz about how insular mastadon is will be amazing. We are all a woke cancel culture (the right wing insult versions) over here.

@ATLeagle @mekkaokereke AFAIK Gab doesn't federate, but I guess that is a threat - Facebook building the naziverse. This is why we need to be fighting to make them kick off the nazis rather than meddling in each others' admin decisions. 🤦
Dear @mekkaokereke,
You bring up good points and I largely agree. That said, there *may* be a place in the Mastodon ecosystem where @threads.net makes sense. The other thing I want to say to the dear brothers and sisters here is, either way, avoid being sucked into the low quality discourse that Twitter, Insta, FB and others have fostered over the years. Some of that exists here, especially from celebrity accounts seeking audiences on Masto, but it's easy to avoid.
With love,
ESM

@esmichelson

I don't understand the connection between "low quality discourse" and what I said.

Can you say more?

Dear @mekkaokereke,

I am not saying what you wrote is low quality. On the contrary. Please forgive me for any misunderstanding.

I was speaking broadly about the baiting and trolling that occurs on ad supported platforms and sincere wish that Mastodon remains a place of relative tranquility.

Humbly,
ESM

@mekkaokereke MediaMatters has an excellent article out today documenting how hate speech is proliferating on Threads, which very clearly illustrate your points.

Content warning on this, because there are a lot of screenshots of extremely vile hate posts in it.
https://www.mediamatters.org/facebook/mark-zuckerberg-claims-he-wants-metas-threads-app-be-friendly-so-far-unmoderated-hate
Mark Zuckerberg claims he wants Meta’s Threads app to be “friendly.” So far, unmoderated hate speech and misinformation are proliferating.

Media Matters for America
@laurenshof @mekkaokereke
Ugh. Ew. 🤢
This is exactly why I left monetized social media behind. It has been really, really peaceful not having to see that kind of disgusting rhetoric on a daily or even hourly basis.
Why would we ever invite that into Mastodon?
@mekkaokereke that's exactly why the admins on my instance decided not to federate with Threads with an option to change this later if they proved their moderation was up to scratch, basically saying: we know they're bad, onus on them to prove us wrong.

@brunogirin @mekkaokereke

the time to give FB the benefit of the doubt was gone way before Cambridge Analytica

in 2011, during the #ArabSpring people reported FB were giving the Egyptian government the location of activists that were mysteriously disappearing.

FB e-suite gave half-assed excuses for why they did; notwithstanding Egypt having no jurisdiction over them at the time.

FB cannot be trusted.

period.

@mekkaokereke I have personally received death threats on Facebook that Facebook ruled did not violate their community standards, I am not a fan of their offerings.

@mekkaokereke

I don't expect to defederate over meta's business, although I fully agree with Cat Valente (https://catvalente.substack.com/p/mark-fking-zuckerberg-is-not-your).

I 100% expect to defederate over their crappy moderation.

I have not proactively blocked, because I don't know the actual federation hostnames and I don't want to let down my guard.

Gab is welcome to them.

Mark F***king Zuckerberg Is Not Your Friend

You know, I had other things I wanted to yell about.

Welcome to Garbagetown

@mwl @mekkaokereke

"Musk is just so brain-serratingly bad at running an actual business that isn’t two grifts and a government subsidy in a trenchcoat that he managed to essentially brick a site that had functioned without serious issues for a decade and a half the same week his mother had to tell him not to fist-fight another billionaire"

Oh I am loving this

@mekkaokereke

"bUt mY fRiEndS aRe oN tHeRe!!!"

@mekkaokereke there are so many people - and importantly organisations - on there, that if/when it federates, I'd view it as probably too big to defederate. Silencing so many people because of the actions of a few, or because of inadequate regulation of their megacorp owner, seems like snatching defeat from the jaws of victory.

@georgelund @mekkaokereke Oh the other hand, on a federated network, it's accepted that if your instance turns shitty you're simply supposed to register on a new one and move.

#ActivityPub in particular is designed in a very instance-centric way, so that part really has to be accepted for it to work at all.

@mekkaokereke I suggest reading this very well considered piece on defederating threads.

https://www.timothychambers.net/2023/06/23/project-and-the.html

IMO, we need to figure out how to defederate users, not instances, or risk becoming irrelevant.

Threads and the Fediverse - A Smarter Battle Plan to Protect the Open Social Web

With the #meta #Project92 or #Threads Fediverse offering, there has been a, well, robust discussion of how to avoid threats looming. Those advocating mass-preemptive defederation make three cases for it. ➡️ To avoid data mining … However, defederation does virtually zero to avoid any big tech entity scraping all the Fediverse public social graph today - Want proof? See here: is.gd/q8U2pv But what if they merge that Fediverse data with their own internal data from IG isn’t that worse than just scraped data?

Are folks from #NetrootsNation on the Fedi? If so like or comment here to find each other.

@mekkaokereke I’ve been in the “wait & see” camp but having watched Threads, I’m pessimistic that federation would last longer than 15 minutes. I’m still “modified wait and see” in case they’ve cleaned up their act when the hypothetical Federation Day approaches. But as things stand now, little point trying to make it work.

I wonder if leadership over there knows that their much-trumpeted alignment with the Fediverse will evaporate instantly, given current circumstances?

@timbray @mekkaokereke I'm curious if leadership would even care? I have a feeling that they only even deigned to mention Mastodon because it got some popularity with the migration from Twitter.
@agiorgio @timbray @mekkaokereke it has the feel of a fig leaf or misdirection; the way they said "to federate with _us_ you'll need to abide by our moderation guidelines" suggests that they don't realize just how isolated they already are.
@timbray @mekkaokereke in a “hilarious” twist what if the only people that federate with them is Gab and truth social or whatever and it ends up just being an increase in threat surface of fascist misinfo.

@timbray @mekkaokereke these days, my default assumption is that either:
- it doesn't occur to leadership that a predictable failure will happen, or
- leadership simply asserts that the predictable failure won't happen.

(And then the predictable failure happens.)

@timbray @mekkaokereke

For those mods who are wait-and-see about Threads federation, is there a way to ask the Threads team? “Hey, we notice this sort of content on Instagram, and that’d be grounds for us blocking or defederating a server that hosted it unchecked in the fediverse. Do you know that? Are you concerned about it? Do you have a workable plan to manage it on Threads?”

This isn’t enough for the mods concerned about embrace-extend-extinguish, but it would still be interesting.

@alix @mekkaokereke Exactly. The meta leadership I think has no idea of the instant defederation that will happen once people there start broadcasting the anti-trans poison that is currently accepted on Threads. Maybe they won't care? But it'd be nice if they knew.
@mekkaokereke exactly, if any of the corpo instances become worthwhile to federate with, i'll hear about it
@mekkaokereke it looks like the Threads moderation is much stricter about exactly these accounts than Instagram is, so there is that...
@mekkaokereke I predicted even before the launch that most of those claiming to be “on the fence, but ready to defederate if P92 violates the rules” will in fact *not* defederate when P92 violates the rule. Maybe I should take bets on this.