I wouldn't have written it if I hadn't seen that you frequently engage on these threads, so give the credit to yourself for actually building something for the community, instead of just talking about it in one-way blog posts.
>We've tried to model it after real life, for better or worse. If someone chooses to use the same identity for discussing both firearms and (say) cooking, then they may get shunned by some cookery folks. But most of the cook group won't care and ignore it. You might get unlucky and discover the head chef is anti-firearms and kicks you out, but frankly that sounds like a good reason to find a more broadminded cook group.
>You're right that a large server like Matrix.org could take an opinionated view and go and apply radical blanket bans for all sorts of stuff, and set a precedent that it's okay to throw your weight around
My biggest fear is both of these things in tandem. I agree that individually, they're solvable problems. For instance, I'm not actually worried about Matrix.org banning me by association, and in fact, welcome a reduction of child porn, spam, etc. And if there was a plethora of cooking groups, and getting banned by one still left me with others, it wouldn't be a huge deal.
But its the potential combination of clout and curation that makes me concerned. For instance, email is a "federated" standard, in that anyone can get a domain, and host their own email server. However, for all intents and purposes, it's not a federated standard because nearly everyone uses Google, and getting Google to accept email from my domain is nearly impossible for a variety of factors that I don't have control over. That means that effectively, it's no longer federated, unless you play by Google's rules.
I don't know how to solve this problem, and I feel bad complaining about this solution without proposing a better one. I see the predicament that you're in, as well. Governments obviously want to be able to peek into conversations, and will use any excuse they can get to do it. If you try to play the game by their rules, and come up with a way to preserve privacy while stopping the spread of things like child porn, by democratizing moderation of the entire ecosystem, then I see how that potentially solves the problem, or at least kills the child porn excuse. But I think the game is rigged, and this capitulation comes with costs to the platform, that, because of the rigged game, ultimately won't protect what you're trying to protect. Governments will come up with some other reason why they need access to private conversations, and instead of a single death knell, it'll instead be a drawn out one.
The only solution is to replace the game with something entirely different, the same way for instance, that cryptocurrency did with financial markets. If they had tried to play by the rules of the existing game, it'd have gotten nowhere, because the financial game is designed to specifically stop things like that.
>Finally, P2P Matrix will change the dynamic entirely on this - no more servers (by default) means that users will be entirely making their own choices on where to hang out and who to hang out with.
I'm very excited about this, because this is the type of outside of the box thinking that I think falls into the category of "whole other game where existing rules don't matter". I really hope you guys can get the UX good enough to pull it off with wide adoption.
> But it's the potential combination of clout and curation
Our plan for the Matrix.org server is, in an ideal world, to turn it off once we have portable accounts (and especially once we have P2P). Users can easily pick a set of other servers for home, or just use their devices. We have absolutely zero desire for any server to have clout or to end up as a Gmail-style centralisation point.
However, in an account-portable world, I suspect all we'll see is that communities (rather than servers) will emerge which have equivalent risk of disproportionate social influence. Then all we can do then is to arm the users with tools which allow them to visualise and curate that influence and make up their own minds, rather than accidentally getting trapped in someone else's filter bubble all over again.
> The only solution is to replace the game with something entirely different, the same way for instance, that cryptocurrency did with financial markets. If they had tried to play by the rules of the existing game, it'd have gotten nowhere, because the financial game is designed to specifically stop things like that.
From my perspective, introducing a morally relative reputation system as a core primitive in the protocol, is very much replacing the game with something entirely different. Imagine if SMTP had the concept of subjectively modelling spam built in from day 1. Or if the Web had had the concept of subjective search result quality.
Nobody has pulled this off before (as far as I know?) but we're having a go at it to see what happens. If it goes horribly wrong then worst case we just turn it off as a failed experiment.
>We've tried to model it after real life, for better or worse. If someone chooses to use the same identity for discussing both firearms and (say) cooking, then they may get shunned by some cookery folks. But most of the cook group won't care and ignore it. You might get unlucky and discover the head chef is anti-firearms and kicks you out, but frankly that sounds like a good reason to find a more broadminded cook group.
>You're right that a large server like Matrix.org could take an opinionated view and go and apply radical blanket bans for all sorts of stuff, and set a precedent that it's okay to throw your weight around
My biggest fear is both of these things in tandem. I agree that individually, they're solvable problems. For instance, I'm not actually worried about Matrix.org banning me by association, and in fact, welcome a reduction of child porn, spam, etc. And if there was a plethora of cooking groups, and getting banned by one still left me with others, it wouldn't be a huge deal.
But its the potential combination of clout and curation that makes me concerned. For instance, email is a "federated" standard, in that anyone can get a domain, and host their own email server. However, for all intents and purposes, it's not a federated standard because nearly everyone uses Google, and getting Google to accept email from my domain is nearly impossible for a variety of factors that I don't have control over. That means that effectively, it's no longer federated, unless you play by Google's rules.
I don't know how to solve this problem, and I feel bad complaining about this solution without proposing a better one. I see the predicament that you're in, as well. Governments obviously want to be able to peek into conversations, and will use any excuse they can get to do it. If you try to play the game by their rules, and come up with a way to preserve privacy while stopping the spread of things like child porn, by democratizing moderation of the entire ecosystem, then I see how that potentially solves the problem, or at least kills the child porn excuse. But I think the game is rigged, and this capitulation comes with costs to the platform, that, because of the rigged game, ultimately won't protect what you're trying to protect. Governments will come up with some other reason why they need access to private conversations, and instead of a single death knell, it'll instead be a drawn out one.
The only solution is to replace the game with something entirely different, the same way for instance, that cryptocurrency did with financial markets. If they had tried to play by the rules of the existing game, it'd have gotten nowhere, because the financial game is designed to specifically stop things like that.
>Finally, P2P Matrix will change the dynamic entirely on this - no more servers (by default) means that users will be entirely making their own choices on where to hang out and who to hang out with.
I'm very excited about this, because this is the type of outside of the box thinking that I think falls into the category of "whole other game where existing rules don't matter". I really hope you guys can get the UX good enough to pull it off with wide adoption.