Three months ago I posted about the Atlantic Council’s interest in the controlling the fediverse: https://lemmy.ml/post/6641106
I think these projects are the continuation of the successful American “intelligence community” censorship of corporate social media platforms. They even tried to formalize the system two years ago as the Disinformation Governance Board.
This Tech Won’t Save Us podcast episode makes a very important point: any movement that does not have a structure and some form of leadership can easily be taken over by anyone willing and able to fill that kind of power vacuum.
Fediverse currently does not have a structure nor a form of leadership other than perhaps “whatever Mastodon is doing”. That’s problematic. I hope that we recognize this and do something to fix it, before that power vacuum gets filled by… someone we might not like.
I do see that the researchers involved in the OP link are Erin Kissane and Darius Kazemi. That’s fantastic. They are truly fedi old guard, deeply engaged, very knowledgeable, and generally wonderful human beings.
I can certainly tell you that Lemmy wont blindly follow what Mastodon is doing. They arent doing a good job for the Fediverse, for example they make zero effort to improve compatibility with other projects. Instead others are left to reverse engineer their federation logic.
I can certainly tell you that Lemmy wont blindly follow what Mastodon is doing.
Good to hear.
They arent doing a good job for the Fediverse, for example they make zero effort to improve compatibility with other projects. Instead others are left to reverse engineer their federation logic.
Yeah. Plus, the sheer size of
mastodon.socialand the monoculture of Mastodon-based instances is just unhealthy. I wrote about it at length.
Agreed, Kissane and Kazemi are community forces to be reckoned with in this context. Their research into governance has been funded for at least the first half of 2024.
Wonderful!
I think Mastodon is already compromised
I think throwing around vague but scary-sounding terms like “compromised” is a very bad idea.
Yeah, good luck with that. Controlled given instances might be achievable, but the protocol will eventually route around the damage.
I think that is engineer’s disease thinking.
Running instances costs money & labor, and I predict that many instance admins will be happy to accept various forms of assistance, and the CIA/FBI/etc-backed organizations will be happy to help. The assistance might be just money, which comes with strings attached such as simply having the ear of the admins. The assistance might be “moderation tools,” to save labor efforts.
I think the instances that choose to federate with corporate social media (which are already captured by feds) will probably be the easiest to gain control of. If you’ve been following the corpo federation discourse, you might have noticed that the instances with the largest user bases tend to be the most interested in federating with the corpo social media.
so you agree the only instances that can be compromised are ones that choooose to do so.
i think most of what you wrote here is wrong, and i have about as much to back it up as you do.
The ones that don’t choose to may eventually find themselves defederated from the ones that do. We’ll get walled off from the large instances that “play ball”.
uh huh. k. cool story. let me know if you ever get anything to actually back up that claim.
let me rephrase… your conjecture is unfounded, and hyperbolic.
RemindMe! 4 years "federation is going great"see; proof you have nothing but hopes and dreams. just like me!
ha ‘’‘you just wait, in 4 years i may be right!’ hilarious
rude ._.
Good. It’s better to have more instances, than larger instances (see: Reddit, Threads).
Do they also control email? As in all email servers that are “federated” with one another? Conspiracy thinking is dangerous. Maybe I work for “the feds”? Maybe you do and you’ve already been subverted against your own conscious wishes and this is a honeypot you’re posting?
At some point you have to trust something or someone or you end up in the trust-equivalent of solipsism.
I recommend this old paper from the early days of compsci – Ken Thompson: https://fermatslibrary.com/s/reflections-on-trusting-trust
The moral is obvious. You can’t trust code that you did not totally create yourself. (Especially code from companies that employ people like me.)
Take this and abstract. Eventually you’ll either need to be a wholly offline hermit, or accept that there is risk of subversion at every level and that risk must be tolerated in order to use the tech.
Your rebuttal to engineer’s disease is to quote Ken Thompson
From Michael Parenti’s Dirty Truths:
Those who suffer from conspiracy phobia are fond of saying: “Do you actually think there’s a group of people sitting around in a room plotting things?” For some reason that image is assumed to be so patently absurd as to invite only disclaimers. But where else would people of power get together – on park benches or carousels? Indeed, they meet in rooms: corporate boardrooms, Pentagon command rooms, at the Bohemian Grove, in the choice dining rooms at the best restaurants, resorts, hotels, and estates, in the many conference rooms at the White House, the NSA, the CIA, or wherever. And, yes, they consciously plot – though they call it “planning” and “strategizing” – and they do so in great secrecy, often resisting all efforts at public disclosure. No one confabulates and plans more than political and corporate elites and their hired specialists. To make the world safe for those who own it, politically active elements of the owning class have created a national security state that expends billions of dollars and enlists the efforts of vast numbers of people.
As for email, you say that as if Edward Snowden revealed nothing, or that the US doesn’t have a history of injecting backdoors into encryption standards.
There’s a difference between ‘governance’ and ‘control’. And I really doubt Erin Kissane of all people is involved in efforts to control the fediverse.
Not sure how you get from Fediverse people researching what server admin/moderation structures work well and which ones don’t to CIA censorship.
Someone’s researching governance in a federated environment and that’s scary because…?
You all can downvote me but if you’re not interested in living in the land of facts, demonstrating any attempt at backdooring Fediverse products beyond “bad money in a think tank”, then there’s literally not even the smoke to indicate a fire.
Can you share anything that indicates that understanding governance is even tangentially related to backdooring these products or the teams behind them? Is the best response really “wait four years and see”?
That hardly means they’re investigating how to take over instances… These egghead think tanks will do research on anything that matches a couple buzzwords and this is one of them.
