Not to hammer the point to hard, but de-concentrating power is the exact reason why it is better to have moderation decisions across multiple layers of the network stack rather than in level 0 (the government).
Forum messes up on moderation? Not a big deal.
Web host starts making bad decisions? Tons of options.
Clouldflare banning you? Tougher, but there are multiple CDN services, if Cloudflare becomes evil it's not necessarily the end of the world.
ISP banning you? Now we start getting pretty dangerous, there are fewer options available to services and if moderation decisions are made poorly, that can have effects across the entire network for everyone.
The government prosecuting you? This is level 0 of the network stack.
The way that we guard against concentration of power is by de-concentrating it. Cloudflare (and to be fair, other large Internet companies too) are arguing for the opposite of that. In the specific case of Kiwi Farms, maybe this example is so egregious that it does make sense to have some new laws. I kind of agree with that. But no good law will be enough on its own to get rid of Cloudflare's responsibility, the only law responsive enough and fast enough to do that would be one that violated free speech rights.
I'm not disagreeing with your entire argument, just that portion.
Multiple layers of moderation are only safer for free speech to the extent that none of them have too central of a role and there's some degree of visibility as to what is happening.
One of the things that has made social media so toxic to speech is that it has A) gathered so much of the "share" of being a conduit of speech at scale, and B) creates a false feeling of consensus by creating playing fields that are tilted in various ways without the tampering being obvious.
I suspect where I differ from the CEOs of companies like Cloudflare/Facebook, is that I think the solution to that isn't to get rid of moderation, but rather to enforce antitrust and break up their companies. :)
Facebook in particular has this problem; it's constantly asking the government to tell it what to do because it doesn't want to be in charge of speech, and yet it has no problem buying competitors and trying to take over markets. It makes me wonder how concerned about speech these companies actually are, since they had no problem growing their companies to this size and putting themselves into situations where their moderation decisions carry so much weight.
I'm not sure whether Cloudflare deserves antitrust action. But, yes, Facebook is concerning.
Of course, you've got to acknowledge the flip-side. Sometimes it should be left up to the government. If your internet service is a natural monopoly (perhaps augmented with protections of a franchise agreement from the government), it's especially problematic for them to be making moderation decisions.
All in all, it's hard.
This is a bad thing, according to you? Should Facebook instead make those decisions on its own?
The thing is, no matter what Facebook decides there will be criticism. If they make the decisions on their own, bad. If they ask for government to decide what speech is lawful and what isn’t, bad. If they don’t block fake news, bad. If they block fake news and realise months later it was real, bad.
And your genius solution is to break up the company. But any network, regardless of size will have this issue. The rise of Tiktok makes this very obvious. There’s tonnes of misinformation on Tiktok but it doesn’t get the same coverage because that’s not what aligns with the NYT’s priorities. Tiktok gets around the content moderation problem by simply saying and doing nothing, hoping no one notices.
So what’s your solution to TikTok? Break it up as well? Into what pieces?
The problem with people who come up with simplistic, unrealistic solutions to hard problems is that when the obvious flaws are pointed out in their thinking, they’ll double down.
The principle behind breaking up platforms is that individual moderation decisions do become harder the more people that they impact. It's also not just a free speech thing, this is the same reason why it's dangerous to have a browser monoculture. If I tell you that having one company in charge of the entire web makes their individual decisions about the web more impactful and more dangerous, that's something you understand, right?
Same deal for moderation.
Of course, see mlyle's other sibling comment -- sometimes we genuinely can't do anything about a natural monopoly and we just need to recognize what they are. But in instances where we can, decentralizing power decreases the overall risk of moderation mistakes for the entire network.
CloudFlare acts against speech all the time. They'll sell you a service to screen the speech of others and then pass it onto you or not, at their decision.
CloudFlare's own terms of use for their Email Forwarding product is very clear that they will squelch your speech as well, in many conditions that don't come anywhere approaching "organizing an international manhunt to intimidate a minority": https://www.cloudflare.com/supplemental-terms/#email-routing
They should stop talking about this like it's "pure speech" because it's not that at all, and even to the extent that it is, they already limit actual "pure speech" in many other scenarios not nearly as threatening as this.
Is that the case?
https://blog.cloudflare.com/cloudflares-abuse-policies-and-a...
As far as they do mention activity, they do say they ban content related to activity that is, for example, "libelous". So they'll block you for publishing insults about someone, without any further malicious activity.
They also say that they ban content used as part of malware command and control, which seems to cover spamming, meaning that they should have no problem blocking spammers trying to use their "security protection" service.
Of course it turns out I don't even have to use the analogy with spam because CloudFlare's own post that you linked to clearly states they can remove access to content that is "... harmful, or violates the rights of others, including content that discloses sensitive personal information, incites or exploits violence against people or animals ...".
That's literally been KF's modus operandi for years now. Unless CF changed their terms very recently, that behavior of KF has always been proscribed. Yet CF saw fit in their discretion to make a conscious choice to continue aiding and abetting KF in its campaign of doxxing and incitement of violence, something far worse than libel or C2.
You're again missing the distinction between their hosting policy and their security product policy. This was the important distinction that I first pointed out to 2 comments ago, and that I posted this document which explains clearly 1 comment ago.
> Hosting products are subject to our Acceptable Hosting Policy. Under that policy, for these products, we may remove or disable access to content that we believe:
...
> has been determined by appropriate legal process to be defamatory or libelous.
...
> Our conclusion — informed by all of the many conversations we have had and the thoughtful discussion in the broader community — is that voluntarily terminating access to services that protect against cyberattack is not the correct approach.
The problem with your logic here is that you’re considering the voluntary filtering of messages by a party as being the same as stifling someone’s ability to say something. The filtered party can still say what they want but the intended recipient should always have the ability to ignore that if they so choose.
“CloudFlare's own terms of use for their Email Forwarding product is very clear that they will squelch your speech as well”
The difference between controlling what gets sent out by their email service is more a question of legal liability than free speech. They are not limiting anyone’s ability to give free speech within the confines of the law here.
To make a stronger argument maybe you need to create a stronger definition of free speech than what is defined by law to prove any violations on CF’s part.
In the case of KF, CF has only suspended them on what they could identify as undealt-with legal violations. This is fundamentally different from revoking services to silence unsavory takes.
I also imagine the doxxed information on the platform (KF) is removed after a time so attacking the whole platform at this point just seems like an effort to stifle a community with subjectively unpleasant ideologies.
I'm not supporting Kiwifarms having a platform.
I'm saying speech being effectively squelched by a small number of powerful parties is problematic. If the small party is the government, this is obviously problematic. If Facebook is a huge part of people's discourse, and subtly tilts the playing field in various ways, this is problematic, too.
I can insulate myself (mostly) from the effects of Facebook's curation. But there are still profound social costs.
I am also not sure you're conversing in good faith. You're tossing out pithy one-liners that demand greater effort to respond to than to say them. This was also my experience a long time ago when we used to discuss things on IRC (including, I believe, this exact topic).
https://news.ycombinator.com/item?id=32712819
Concentrated power in the path of speech or commerce is dangerous, whether it's governmental or private.
Many parties deciding independently whether to "entertain my nonsense" is good. One critical party in the path (governmental or commercial) is bad.
We're not talking about hosting KF. There's lots of hosts. And Cloudflare was not hosting KF, but instead providing DDoS protection services.
But there's approximately 2-3 services that can reject DDoS at high scale. Or maybe slightly more. This is right at the threshold of concern.
Here, I think the decision that was made was a good one, but at the same time a very small number of unaccountable parties making this kind of determination is worrying.
Either way, if you're saying something so reprehensible that no-one will allow you to use their platform to say it, maybe you should look at what you're saying.
> > > Many parties deciding independently whether to "entertain my nonsense" is good. One critical party in the path (governmental or commercial) is bad.
Have a nice rest of your day; I'm done.
Perhaps. I'm skeptical of concentrations of power wherever it is: government, Cloudflare, Facebook, etc. At least the former is theoretically accountable for choices.
Also, Cloudflare asserts their position is that they largely do not want to restrict speech beyond that government baseline and they won't act themselves against speech. Here they claim they are forced to (and they probably were).