jadero

I think that “all” of anything is going to be a dumpster fire no matter what your standards are. Sturgeon said that 90% of everything is crud. What he didn’t say was that everyone has their own opinion on which 10% is worth anything.

Or you could put on your hip waders and follow Kipling into the waters looking for the 20% that makes it worthwhile to wade through the other 80%.

(More at https://en.m.wikipedia.org/wiki/Sturgeon's_law, including a link to the actual text of what Kipling said.)

I agree with you. Everything I’ve read makes the assumption that we can do nothing in the face of “embrace, extend, extinguish.” Anyone who has ever played a multiplayer game of any kind knows that a new strategy can be devastating, but only the first time.

But now we know about that strategy and it has an inherent weakness. “Extend” is only a problem if we as developers, admins, and users accept extensions uncritically. If “extend” is on the critical path to “extinguish,” then we can interrupt the process by not accepting or not becoming dependent on extensions that put the Fediverse at risk, no matter who proposes or implements them.

In my opinion, the worst that can happen is that we ultimately find it necessary to defederate from Meta. If that splits social graphs, well, for anyone currently using a Meta property, that is where we are now.

Regarding https://ploum.net/2023-06-23-how-to-kill-decentralised-networks.html

I think it’s important to note one major difference between those historical events and today. There were no platforms with the reach of Meta at the time, today we have, well Meta (and Twitter, etc)

What that means is that it is longer possible to really capture our personal social graph. If we are using Facebook, Instagram, Twitter, or whatever right now, then we’ve already partitioned our own social graph for personal reasons.

If you are not on those platforms, then there is no social graph to capture except as members choose to disconnect from us.

My reading of https://blog.joinmastodon.org/2023/07/what-to-know-about-threads/ suggests that the writer is not terribly concerned about the presence of Meta.

Anyone who recognizes my username or who digs through what I’ve written knows that I’ve gone back and forth on this.

I completely understand the various motivations for defederation and in my heart of hearts would love to just flip them the bird. I also have concerns over the ultimate outcome. But my concern is less over whether accepting Meta is ultimately detrimental, showing that the Fediverse cannot withstand “attack”, than over whether rejecting them prevents the Fediverse from bringing them to heel.

I appreciate the push to be more analytical. I thought that I was being analytical, but this challenge to my thinking has made me realize that I’ve not been thorough.

Edit: speaking of thorough, I missed including the founding principles of privacy-centric and ad-free. That would seem to disqualify Meta right out of the gate, just as most people seem to prefer.

Ethically, it would seem that admins are limited by that, although I suppose technically, they should wait and see what an instance does before making the decision to defederate.

Anyway, here’s what I came up with should anyone decide to depart from the founding principles.

What our root concerns? I could be wrong but I think they are advertising, accuracy, and civility. Some people are concerned about sheer volume of new entrants, but I don’t know enough about fediverse architecture and tooling to address that concern.

Accuracy and civility are both moderation problems that the fediverse has to deal with whether Meta joins or not. Yes, there may be issues of scale and volume, but if the fediverse is to go mainstream, those problems need to be solved anyway.

There are forms of advertising that I think we could all live with in moderate amounts, but surveillance ads and malware have pretty much destroyed any tolerance for ads.

Okay, so how would ads get distributed in the fediverse?

1. Any instance could push ads of any type, in any volume to anyone registered on that instance. Any instance could collect data from anyone registered on that instance for serving surveillance ads on that instance or for direct sale to data brokers.
2. Any instance could push ads to anyone who subscribed to communities hosted on that instance, but would be severely limited in data collection without data sharing agreements between instances.
3. Anyone accessing an instance via web page could be served ads, including surveillance ads, whether logged in or not.
4. Anyone using a fediverse client could be served ads by that client, including surveillance ads.

1 through 4 are all instance-specific and can be avoided by any user who wishes to. There will always be nooks and crannies inside the fediverse that are free of ads and surveillance. The very existence of the fediverse is evidence that there is a critical mass of people to keep those spaces alive.

5, the client, has nothing to do with any instance or, in a sense, the fediverse. Clients can always do what a client wants. It’s up to individuals to choose wisely.

So what are the potential benefits? If Meta comes it at Meta scale, then anyone currently using both Meta properties and the fediverse might be able to keep the value of both while using only a single platform. If people using a Meta property follow Meta into the fediverse, that can increase the variety of communities and make very niche communities viable.

My take is that the fediverse can cope with an entity like meta because nobody is forced to join their instance or interact with anything or anyone hosted on Meta.

There are a couple of caveats:

1. As previously noted, I don’t know enough about fediverse architecture and tooling to assess the risk of a very large influx of people over a very short period of time.
2. If “embrace, extend, extinguish” is truly a concern, then I think being vigilant with respect to “extend” will prevent “extinguish.”

You’ve officially changed my mind.

Up until now, I’ve been harping on the concept of “controlling interest” in which a single entity is large enough to control the direction taken. But I hadn’t considered that the new direction might be one that limits the potential for a negative result.

Personally, I think that a sufficiently large instance does represent a major risk. But now I think it’s a risk we have to take. If this federation experiment fails, then what is learned can be used in the next experiment.

Now to track down and add a note to all those comments I made…

I don’t know about great, but thanks! I’m working on the underlying attitude that leads me to make statements like that. I hope seeing it for what it is and acknowledging it is progress over pushing back. 😀

Yes, even I cringe when reading what I wrote.

Yes, not my finest hour…

Edit: this comment changed my mind. In a nutshell, if we can’t keep a large instance controlled by “the enemy” from destroying what we’ve got, then we just have to do better next time.

My instinct and ideology tells me that no, we should not make that decision before it whatever Meta does becomes reality.

Fortunately, I have laboured long and hard to develop a scientific frame of mind, where decisions are based not on instinct and ideology, but on the evidence of reality, even in the face of incomplete information.

TLDR: we should not federate with Meta and that should be the decision taken by the majority of instances. Not because it’s Meta, but because of the power that comes with a having a “controlling interest.”

This is a network of independent actors. Other such networks are publicly traded companies, condo developments, and most cryptocurrencies. Whenever you have such a structure, you must be vigilant against the 51% problem, because whoever controls the majority, controls all.

This is why publicly traded companies do stock buybacks when they have extra cash and the company holds less than 50% of outstanding stock.

This is why condo developers (especially in condo conversions) like to hold the majority of units as rental units. That way they dominate the vote, and do so in their favour.

This is why cryptocurrencies must be vigilant against “51% attacks.”

Also, the reality is that it is frequently unnecessary to actually hold 51% to gain control. In a system of widely dispersed small holdings, it may be possible to have de facto control with much smaller percentages. If the typical holding is less than 1%, one actor holding 30% might gain effective control with a minimum of effort. Think of them being a “super swing voter” whose interests must be satisfied in order to get things done.

If you’ve read this far, then you can probably guess that my real point is that we must be vigilant against any instance becoming too large. Not because they are untrustworthy or bad citizens, but because we have little recourse should they decide to exercise the power they have. In other words, Meta is just a specific case of the general problem.