Discussion: Support the Anti-Meta Fedi Pact

@Rich Jensen Understood. Still, from me, a big -1 on a preemptive kneejerk defederating Meta, without a larger discussion. As members of the ActivityPub and Fediverse community, we worked really hard to be taken seriously, and specifically to provide an open protocol that all social media platforms should use. We should have specific criteria for defederating (even if they're specific to Meta), and only do that when they break those criteria (in which case, I'll be very enthusiastically pitching in on banning them and getting everyone to do that).

@Nathan Schneider "Would it have been better if Google had not created Gmail?" that's a clear "yes" from me.

Google has been on a years-long campaign to enact the "extend extinguish" parts of the choreography (most recent example I'm aware of is "AMP for email") which they couldn't even be considering if they'd been less successful with signing up users on the promise of compatibility. I've been running my own email servers for over a decade, and Gmail is the unparalleled #1 worst offended when it comes to exhausting non-delivery issues, requiring signing up for their non-standard, proprietary tools to have any hope of fixing, and with close to zero of their billions-of-dollars-large budget given to supporting email operators in using.

Gchat is a great example, where it used to be federated, and was turned into a walled garden immediately as soon as they thought they had enough of a captive userbase to do so.

@Ed Summers Hi Ed, do you know if a user blocking an entire server as in the docs you link to will prevent that server from accessing your data? Or does it simply block you from seeing any content or users hosted on that server? I don't see that clarified in the docs.

@Eduardo Mercovich So why is current policy to block/not-federate with instances based on such content not enough? As it stands, if an instance is allowing nazis and white supremacists to post freely, social.coop blocks/de-federates anyway, doesn't it?

@Chris Lott

Yes, totally so. I believe (not being a pro on this issue) that our general values allows us to not federate with Meta. But in this values there is no mention (that I know of) about privacy and selling personal information. We can add it. :)

Other issue to discuss in parallel is to analize the social/symbolic impact of adhering to the anti Meta FediPact...

This is a nice extension. Although I wrote earlier that I would prefer not blocking until we see, I'm also very comfortable with a pre-emptive block and a commitment to reconsider strategy when we know what they're actually doing.

I have a hard time understanding what people want to "wait and see" about, from a company with a track record like Facebook's? They are a known quantity.

I also don't understand the position that we should look for a broader rule, or a more permanent coalition, before acting on this case. We don't need an ironclad timeless principle of blocking to agree that Meta is not aligned with our values.

I really like what others have expressed in regards to principles; let’s find general policies on the basis with which meta can be federated with or degenerated from, rather than design something that targets Meta directly.

I’d also like to add a general caution around the use of defederation as a way of expressing disdain for behaviour in the past. If Meta does something we consider dangerous in regards to the fediverse, we can act. Why risk the upsides when downsides have yet to materialise. Let’s be practical about our actions, not political. Meta hasn’t done anything yet. Why would we not give them a chance to redeem themselves when there is no evidence of negative behaviour yet?

I have relatives on Instagram. Friends. Being able to connect with them in a way that doesn’t give meta anything meaningful seems a golden opportunity.

Meta will stress the fediverse’s ability to scale. We should wish for this! There are problems running at instagram’s scale that we won’t discover for years lest a large community joins.

@Bjorn Toft Madsen

Let’s be practical about our actions, not political. Meta hasn’t done anything yet. Why would we not give them a chance to redeem themselves when there is no evidence of negative behaviour yet?

They did an awful loof horrible things already, for many years. How many times someone must show bad faith and bad behavior until we start to take precautionary measures?

Anyway, the idea that it is coalescing (about positive values that we require to federate) is much better in every way, so in both cases, we cannot start federating with Facebook until they sign that pact (which they will never sign, or they loose the huge business of selling people's information). :)

I think this proposal needs work. Any first pre-emptive block in the fediverse's history (?) needs to try to limit future whataboutery and potential confusion. The AntiMeta Pact says Meta 'must be fought back against at every possible opportunity" but I think some here see blocking more as part of a 'wait and see' approach, which is different.

For me the question that lacks consensus on here is why pre-emptively block? If it's because of Meta's past behaviour; the fact they're an advertising giant; their demonstrated disregard for GDPR / right-to-be-forgotten; their history of embrace/extinguish such as with XMPP; their approach to disinfo/hate-speech, genocide-incitment-in-languages-their-mods-don't-speak, their 15,000 seemingly underpaid moderators etc – then I'd rather we adopted a generic policy against such actors, as I doubt Meta will be the last.

But, imagine none of these things existed. Would federating be ok? I'm pausing most when I picture a social network of 1 billion daily active users (3 billion across Meta) adding Activity Pub and joining a fediverse of 8 million (?) with ~1.5m daily active users. Even if Meta was faultless - joining would still make me nervous, on two levels:

• Technically - the potential firehose of 100-1000x more content and users. Is ActivityPub, social.coop and the wider fediverse up to it? Who covers the costs of the caching servers and human moderation around that increase? We grew ten-fold-ish at the end of last year and it seemed tough for all instance admins. I'd welcome some expert instance sys-admin / or ActivityPub spec-writers input on what Instagram being an instance actually means.

• Re-centralisation - If Threads was a success and rolled out across Meta, the Fediverse as it stands today would be a tiny (< 0.1%) fraction of an ecosystem that was mostly just one company. That removes the decentralisation at the heart of this project, that drew many of us here, and that company's decisions on moderation, tech, the way regulators treated them, the allowed behaviours of their users, etc – would dominate everything here.

So I support the idea of a block until Threads' launch, that's reviewed then and periodically after that. (and of course it might launch and never get more than a million actives). I also support adopting broader policies beyond the Mastodon Server Covenant about not federating with actors involved in non-consensual data aggregation, or who are shown to attack the commons and not give back. But not the proposal as it stands.

NB - since drafting this the proposal changed, I think this adds new problems. ie it now says "Meta, through Facebook, is infamous for condoning the spread of far right ideology and of dangerous misinformation" - some people may say the same of Mastodon + ActivityPub, the software + protocol that powers Truth.social and countless other extremist / disinfo platforms. ie there's sufficient arguments without doing a 'you're evil and we're the good guys'. I've seen disinfo and hate speech on our instance - thankfully removed, but who knows what would happen if we'd a million times more users.

@Nic "Meta, through Facebook, is infamous for condoning the spread of far right ideology and of dangerous misinformation" was present when I first posted. I only changed the title, and the text, "Therefore, I would like us to discuss whether we should support this move, and if so, how best to do so. As a starting point for discussion, I suggest the following", and those changes were intended to back off from framing this prematurely as a finished formal proposal.

@Brian Vaughan ah sorry - I was looking for what had changed and hadn't picked up on that. Thanks. FWIW, living in a country where Facebook helped make Brexit happened thru weaponised disinfo, I don't have a problem with the statement, so much as feel the fediverse isn't as strong in this area as we'd like to think, so it's easier not to build the argument on that.

Any first pre-emptive block in the fediverse's history

@Nic wasn't the (successful) isolategab campaign also preëmptive?

@Calix - thanks, I stand corrected. I thought that had happened afterwards. Have adjusted.

Facebook actively encouraged the genocide against the Rohingya people. Zuckerberg is a war criminal and needs to stand trial at the ICC. https://www.amnesty.org/en/latest/news/2022/09/myanmar-facebooks-systems-promoted-violence-against-rohingya-meta-owes-reparations-new-report/ No federation with Facebook is the way.

I agree Meta's actions are concerning. I would also support a collective response to the risks involved. Yet this proposal has no actual goals, and it's not clear to me what this initiative would set out to practically achieve, and whether blocking Meta will accomplish it. I would happily support an effort that starts with practical goals and then develops a collection of interventions that we think will work on the basis of those goals. But without goals, I can't even assess the proposal.

Here are examples of possible goals:

1. Goal: getting them to be good stewards and neighbors (for example)

   1. Ask and pressure Meta and other companies to contribute resources and effort to the Fediverse in ways they can't claw back (lots of companies contribute to open source in this way)

   2. Ask and pressure Meta and other companies to contribute resources and effort to trust/safety in ways that are not exploitative like their other efforts

   3. Ask and pressure Meta and other companies to bring more people into the fediverse in ways that help it flourish through network effects

2. Goal: keep them out (for example)

   1. Force the company to abandon their efforts

   2. Reduce and limit the percentage of the Fediverse that they constitute

Those are just four examples— and the tactics would be different for each of them

One source of hesitation for folks seems to be the desire to form a coalition of servers before signing something like this pact in order to act collectively in blocking Meta and other instances that would be potentially harmful to the longevity of the Fediverse.

To me, the Meta Fedi Pact is a collective action and its signatories are a de-facto coalition (a federation even!). If social.coop signed on we could take initiative in building this loosely affiliated group of servers into a more organised and powerful network. Organising around an action point like this isn't premature, it's the best opportunity we have to build momentum around inter-server governance.

I'm not sure what the next step should be. I think there's sufficient interest in taking a public position of some kind against the threat of Meta's entry into the Fediverse, but it seems like we need some more discussion so that we can formulate a proposal that most of us will be happy with.

I see that quite a few of the Concern votes, are actually in general agreement about the overall nature of Meta and it’s potential impact to the Fediverse…

But would prefer a more generalized abstraction via specific rules that wouldn’t just apply to Meta, or creating a larger inter-instance structure that could deal with other future threats.

I appreciate and am in full agreement with those types of positions… so

Let us not let perfect be the enemy of good!

I think it’s also helpful to quote one of the ActivityPub designers

The point of a federated social protocol is actually NOT to be incredibly interconnected but to be very flexible in forming sub-networks. NOT to have a massive space but to have bespoke community-owned communities that can still be seen/discovered at-large if desired.

Wilkie

I'd support joining this pact. I think it's an important collective action against domination of the fediverse by one actor or by a small set of actors. The revelation that Eugen Rochko met with Facebook* under NDA is concerning and makes me even more inclined to support the pact. Giving certain people privileged access to information is a classic tactic to divide a community. Mastodon gGmbH's apparent acquiescence to Facebook creates all the more need for other servers to band together in standing up to Facebook.

*Also, I choose to call it Facebook instead of Meta because I think the rebrand is serving to obscure the fact that, actually, this company has a long track record, and we do not have to wait and see whether it will act in good faith or not. We know. The wait and see comments are, to me, evidence of an incredibly successful rebrand in that it seems to be making some of us forget the company's well-deserved bad reputation.

And if anyone thinks this specific app will be designed by an autonomous team within Facebook who are more benevolent than the parent company, consider that Instagram was long considered a rare success story of an acquisition that was allowed to develop autonomously based on its original values, and yet its co-founders quit a few years ago after the meddling from Facebook executives got to be too much for them, too (described at length in the book No Filter by Sarah Frier).

I missed the vote but I do not support this. I came to free/open source software (FOSS) from the social and economic justice movements. The participation of massive technology companies in the FOSS ecosystem is problematic in many ways and big companies like Meta and their actions must always be approached with skepticism. They do not share our values.

But social movements in the FOSS space have gained quite a lot from being able to build on the resources of big companies in areas of overlap. I think its hard to say that the open source movement has lost more than its gained by the participation of big tech firms—if for no other reason than they have so many resources and have done so much that even small areas of overlap can lead to net benefits.

We should be careful not to let Meta set the terms of their participation. We should create institutional arrangements that are resistant to cooptation. But banning/blocking them before they even start seems to be reactionary and quite possibly counterproductive.

This is exactly what I was going to say, but @Benjamin Mako Hill has put it much better than I could have. To add to this points, I want to encourage every supporting pre-emptive blocks to look at that from the POV of our relationships with people using fediverse servers owned by DataFarms, rather than our antipathy to their owners.

Our whole complaint about Meta et al has been that people are trapped in their DataFarms. Unable to connect with people on other services, and therefore unable to leave the server without losing all their connections there. If the rest of the fediverse blocks those servers, then it's us keeping those people trapped there. That seems like a terrible failure of solidarity.

@Benjamin Mako Hill

If FB is asking developers/maintainers/admins to sign an NDA before talking with them, could we create a Full Disclosure Agreement that their representatives have to sign before they talk with us?

Cory Doctorow uses this approach. :D

Dear all.

There are some points that are coalescing from this discussion and a tad more in https://social.coop/@dynamic/110579677650355358

Do we have a pad to draw a rough sketch among us that we can open for more open participation later? Anything like an etherpad may do, that is -by default- open to social coop members, and then can be expanded (in the same or other platform) for others to enrich...

Thanks a lot. :)

Late to the discussion. Observations:

1. This is a both-a-crisis-and-an-opportunity situation. Reasonable people can make reasonable arguments on both sides.

2. Our tools are extremely rudimentary. The situation is complex, but we have only one bit of choice: block or not block.

3. The impact of what we do here either way is minimal (<<1% of Fediverse user population). Unless we do it in conjunction with others, only we feel the impact, the rest of the Fediverse and Meta are largely unaffected.

4. Even if we did it in conjunction with others, the impact may be too minor to matter. According to https://fedidb.org/current-events/anti-meta-fedi-pact, the pledge so far only affects <400k users.

5. The tool we have is not even remotely up to the task. It is imperative to build up more tools. What those are we will have to learn, but I think:

   1. A set of values that a big part of the fediverse believes are important.

   2. A democractic process by which those values can be updated as needed. (Perhaps an instance with coop experience could have an outside impact on how to do this properly :-))

   3. A public pledge (with branding program) that allows server operators (and perhaps also users, and app developers) to publicly declare their adherence to these values.

   4. Pre-defined policies on how to interact with parties that do not conform to those values. This should involve graduated sanctions, not just black/white block/not.

   5. Some more software features from apps like Mastodon that allow more granular sanctions, like "not blocked but none of the posts from your server get shown in the federated timeline, or only shown behind warnings etc."

   6. Oh, and we definitely need a democratically governed test suite for interop, with a similar public branding regime, to make "accidental" incompatibilities just a tad more expensive in the court of public opinion.

   On the actual question: I'm fine either way, as I don't think it matters for this instance. IMHO we should focus our energies elsewhere.

I've summarized my thoughts on how this will go in this thread, which I won't repeat here in detail. I appreciate people's engagement. It seems to me there are two things at stake, and both are about profit:

1) Facebook is an entity promoting right-wing content

This is undeniable, and will happen on their fedi instance too. Their owners have a material interest in promoting this type of content, because the left threatens their profit margins.

2) Facebook is a $500B+ corporation looking to expand

Embrace, extend, extinguish is real, and something that Facebook has already done several times. They will absolutely do this with the Fediverse. Rumors exist that they plan to pay admins of large instances to federate with them.

A FediblockMeta is a minority of accounts but it's likely the most active accounts. So yes, I do think it's worth it, for the infrastructure alone.

It seems to me we can take the lead here by:

1) Helping create an Assembly across the #Fedipact, with representatives of instances.

2) Help smaller instances convert their infrastructure to a financially sustainable model (coops)

3) Create a set of standards for defederation, prior to #P92

4) Create a way for those standards to be shared

Agreed that we have momentum now. We can use it to react to Meta's takeover, or we can use it to proactively build something better.

I very much agree with what @Miloš has just written, and in fact I'd pulled up Loomio just now intending to post a link to that same thread they've just posted, along with what I thought was a good summary of the case against Meta/Facebook, https://privacy.thenexus.today/should-the-fediverse-welcome-surveillance-capitalism/

Perhaps we could brainstorm about how we could organize such an Assembly, and who we would invite specifically to start with. Such invitations would be a starting point, of course, because we'd want this to be open and public.

Responding to two themes I'm seeing here:

"the pact has minimal impact because it's a small % of fediverse users": There's no reason to believe the pact has reached full saturation. After all, we are still discussing whether to sign on or not. So why assume others are not also evaluating it right now, and may sign on soon? Our decision to sign on would encourage others to. (And as another member pointed out, those users disproportionately represent highly active users who would be open to organizing collectively with us in other ways.)

"instead of blocking, we should organize some kind of more nuanced/sophisticated/generalized collective response": In theory, great, but will that happen? On a timeline where it's still relevant? There is a risk of letting the perfect be the enemy of the good. This pact is a form of actually existing collective organizing for an open fediverse.

To respond to the two themes:

1 - is signing onto the pact now good? i think, yes - but only if it is used as a temporary measure to organize something more. we can also always reverse the decision, the damage seems minimal to me.

2 - is social.coop a good place to start in organizing a fediverse assembly, jumping off from the fedipact? our democratic process and our know-how in the solidarity economy would be immensely useful to other instances. this assembly could then collectively wield much greater power, and come to decisions democratically as well.

I know it is highly rough (being very kind), but please think of this draft as only a starting point to ignite a concrete text and port to other document editing platform/service so we can collectively edit in a short term.

Please also forgive my limits with this language and my very limited knowledge of any technical aspect (which I suggest not to include on purpose, since tech changes fast and we are trying to have something more stable along time)...

This spark-like proposal is:

Positive fediverse (PosiFedi) agreement/ treaty/ covenant/ pact/ accord/ charter/ deal/ entente/ league/ clan/ confederation/ whateverisgoodandfunny.

We value diversity, respect, mutual care, equivalence and justice, and understand that they make us and our society more humane, resilient and regenerative, and our lives more enjoyable.

These values are human but extends to other, non human forms of life that have and deserve the same right to live and are as necessary as humans to keep our biosphere able to sustain life on our shared planet, an emergent property that is right now in great peril.

Based on these values and beliefs is that those who sign this agreement agree, in more concrete Fediverse terms, to:

- never tolerate discrimination of any kind, aggression in any form, misinformation generation and/or spread, denial of historical and present events and threats; we will instead foster civil, informed and respectful dialog and open participation.

- never share private information about other people (without their explicit, conscious and consensual consent?); we will instead take all the reasonable measures and actions to the utmost of our forces to assure the privacy and safety of the people on our instances.

- do our best to moderate in our instances all these and other forms of negative behaviors, to the maximum extent of our forces.

- share information about those instances or people who doesn't share these values and/or break these rules.

We will freely federate with those who agree and comply with these terms. In case of need, will support each other by sharing information and actions.

We will not federate with those who don't agree with these values. Also, failure to comply with the desired behaviors will trigger escalating measures, starting from throttling until total defederation, and their information and acts will be stored and shared among us.

WDYT?