I think the actual escape hatch would be enabling users to filter content collaboratively in a way so that there is no entity in the system could be considered a commercial publisher. The platform providers only remove illegal content and the "not a publisher under the law" users do the community-interest moderation rather than corporate-interest moderation.

Of course some could still choose to play under publisher rules and allow user submissions, but that wold probably be much smaller scale if they want to be on the safe side and and pre-check things individually.

> like spam and porn.

I assume (without having checked) that filtering spam would still be allowed even under platform rules.

Sure, but that’s orthogonal. Let’s say for the sake of argument that all those things came true — business models changed, financial interests separated, etc — but from the users’ point of view, everything works exactly as it does today. Success?

> I think the actual escape hatch would be enabling users to filter content collaboratively

This exists today — shared block lists, etc — and it’s insufficient. There’s a reason that when Twitter announced that they would suspend the accounts of people hoping for Donald Trump to die of COVID, the response from people who often receive death threats was “you mean you could have been doing this the whole time?!”

> I assume (without having checked) that filtering spam would still be allowed even under platform rules.

What is spam? “I made $X from my couch”? Repeated links to my blog? Trolling conservative communities with liberal posts?

The problem with carving out exceptions to your ban on moderation is that it moves your complaint from “all censorship is bad” to “censorship of speech I don’t like is bad”.

That is a logically inconsistent outcome so I would expect nasal demons. But yes, it's probably a success because it means the social networks have less of an interest to and a legal reason not to influence content (maximizing engagement, removing content that harms their brand, conforming to the advertisers overton window, etc.).

> This exists today — shared block lists, etc — and it’s insufficient.

I agree that it's insufficient, because it's not first-party tooling that's designed to scale as well as the rest of the network. You can't propagate them in realtime, delegate moderation to trusted peers and so on.

> The problem with carving out exceptions to your ban on moderation is that it moves your complaint from “all censorship is bad” to “censorship of speech I don’t like is bad”.

That smells like a bogus argument to me, akin to complaining that net neutrality will make DoS prevention impossible. Aspects of spam are things being unsolicited and in bulk. But how much you need to do about it in the first place also depends on communication platform is structured in the first place. Do you see any unsolicited content without prompting for it or do there have to be some network edges between you and the spammer? If the latter, sever them, if enough people do that then the tree will fall in the forest and make no sound.

If your position is "never remove anything ever", you circumvent this. But when your position becomes, "actually, you can delete only spam, but nothing else" then you're back in the exact same situation we are currently in - some things are removed by judgement call and not everyone is going to agree on every judgement. It completely removes the very core of the argument - that judgement calls should be unnecessary or not allowed.

If you're suggesting crowd sourcing moderation in a democratic fashion, I would contend this both isn't effective and also that by providing systems that facilitate this, companies would be implicitly affecting the "censorship" of their platform. After all, someone has to decide what the voting algorithm is, what the chance of being a spammer before you are blocked is, etc.

I think that what is being suggested is that there be multiple, user-selectable, opt-in providers of crowd-sourced moderation services that are orthogonal to the unfiltered communication platform.

Something like the 3rd-party spam filtering services for blog comments or email.

I imagine, though, that a 3rd-party "view" of the Twitterverse would also be a Twitter client, have it's own web presence, and basically be a white-label version of Twitter. Twitter itself would be largely relegated to being a backend pipe that the different sites have in common for interoperability.

I can't imagine how these federated Twitter clones would themselves be able to provide a curated experience without running into the same liability problems.

The alternative is that these optional collaborative moderation services are somehow integrated into Twitter itself, which, while not impossible, poses some serious issues of combinatorial complexity, scalability, and sustainability.

Yes, we have federated services now, but the largest and most popular services are non-federated walled gardens.