[newacct583]

Editorialized title. Link is to the response Amazon filed yesterday to Parler's law suit.

That said, yeah. I don't know why people are screaming so hard about censorship when the simple truth is that this was an attempt to retroactively apply moderation to a community that had verifiably gotten out of control, at a moment where it was clear that many of these threats were real and not just rhetoric.

[gfodor]

There are two things that can be true at the same time:

- Amazon was within its legal and ethical rights to do this, and ought to win any court cases, etc.

- The aggregate situation of de-platforming of individuals and entire platforms from the Internet by a few unilateral decisions by corporations reveals a power structure we ought to not want (it has always been there, but now it's undeniably shown itself willing and able to exercise its power when needed) since it will inevitably be abused given incentives and the lack of checks and balances on it.

[JamesBarney]

> - The aggregate situation of de-platforming of individuals and entire platforms from the Internet by a few unilateral decisions by corporations reveals a power structure we ought to not want (it has always been there, but now it's undeniably shown itself willing and able to exercise its power when needed) since it will inevitably be abused given incentives and the lack of checks and balances on it.

When it's more than five separate organizations that all independently decide to deplatform is it still a unilateral decision? I also doubt these decisions were made unilaterally in the companies themselves.

In this specific circumstance do we need to regulate these businesses to keep them from removing content related to organizing a violent insurrection?

No one even removed this content until these groups literally stormed the capital chanting about hanging members of congress. I don't see the problem.

[gfodor]

I'll admit the use of 'unilateral' here is probably poor, I don't know the proper one. These companies exert monopoly power in the markets they are in. Decisions made by companies which are monopolies are "unilateral-ist" because once a small number of them make a decision together (perhaps colluding) then it leaves no recourse for the person or group they are deciding against.

Mentioned elsewhere, the process we should all wants is one where if a company wants to nuke an entire speech platform like this due to illegal speech, they can get a judge to affirm the legality of the content and the negligence of those who are hosting it. With that in hand, their actions would be immune from criticism. Without it, we find ourselves here, where we basically have to trust them to make the right decision and not abuse their power. It's not what we should want, for the same reason that we should have wanted anti-trust laws in place to reduce the power of monopolists, despite the fact that monopolists were operating entirely legally and often ethically.

[JamesBarney]

Banning content moderation unless you have a judge's order seems like it could be problematic. If this law existed could dang still ban people for people for being a-holes and not following site guidelines or would he need a court order?

[gfodor]

I'm referring specifically to the situation where an entire service is going to be shut down, and silence many, many people who presumably were not doing anything illegal.

At the risk of a stupid analogy, AWS have near-Death Star like capability at this point, so blowing up Alderaan in the name of killing a few people on it the local government seems unwilling to seems like it should have a few checks and balances.

[JamesBarney]

So just for hosting?

If someone buys a rents a bunch of vms and use it to DDOS Amazon do they need a judges order then?

[vel0city]

AWS has somewhere around 30% marketshare in a market with well more than dozens of players. I have a hard time seeing it as a monopoly.

[gfodor]

I'm referring to monopoly power - which doesn't mean a singular entity. A good thought exercise to gauge this in this situation specifically is if it's only incidental that they were on AWS. If they were on Azure or GCP I think we can assume the same result would have occured.

[vel0city]

Good thing there's still more than AWS, GCP, and Azure in the cloud hosting market. There are dozens if not hundreds of companies willing to rent you computing hardware on the internet. There are dozens if not hundreds of colocations willing to rent you space, power, and network connectivity.

[tshaddox]

Is there a feasible plan to obtain better power structures in the same time scales that these people are literally plotting to overthrow governments, murder political opponents, etc.?

[chmod600]

Twitter and FB got a lot of credit for the Arab Spring.

[noWeigh]

Seems like a difference between between overthrowing fascist religious minority in power and preventing fascist religious minority from rising to power?

Electoral college picks the President but two elections on and no popular vote wins for Trump.

If we’re talking about suppression of speech, seems like Trump was keen on doing that as well, on the backs of a political minority.

[gfodor]

There isn't a "deadline" to fixing these problems, they just need to be fixed. Unfortunately, like most things, fear will primarily guide our actions not sober analysis of the long-term unintended consequences of them.

[_jal]

On your second point, there will always be power structures. You can't not have them, when considering realistic arrangements made of humans.

So now what?

Is the complaint really that the gatekeepers are different than they were 30 or 100 years ago?

Or that there is no realistic alternative that could have acted, citing some authority you find more agreeable, in an effective timeframe?

[gfodor]

I would have much preferred if this action occurred due to a legal order, not a few CEOs making unilateral decisions. If they were bound to go through a form of due process before nuking people off of the Internet based upon speech which is assumed to be illegal, my guess is it would be much less likely to be abused. It certainly seems like such a system would have worked just fine in this situation and could be made extremely efficient.

I've argued for years having a few companies mediate all consequential human communication was going to be a big problem, so this happening to wake people up of the dangers isn't a surprise to me but just another stage of the process of people realizing it.

[res0nat0r]

The check and balance is working just fine in this case. Parler kept allowing people to post death threats, they didn't remove it because that's the entire reason their platform exists, and AWS enforced their TOS and stopped hosting them. They're free to go elsewhere.

[tehjoker]

I'm not totally sure what to think, but aren't many violent acts planned on Facebook and Twitter? Wasn't the insurrection at the capitol? Yet, the big boys aren't paying the price, it's the competitor that suddenly got popular that wasn't able to be influenced by politicos.

[brandmeyer]

> it's the competitor that suddenly got popular.

You're omitting a key detail. This competitor was getting popular specifically because of their promise not to remove such violent content, while Facebook and Twitter were steadily ramping up efforts to remove it from their platforms.

[michaelmcdonald]

The difference: FB is alerted to those posts, they (eventually) take them down / close the communities. Parler was straight up telling AWS to piss off.

[tehjoker]

Yea this is valid, I'm just terrified of the tech companies having this kind of power to just censor people and shut down competitors at will. They were kind of justified this time, but given corona, all our communications are flowing through these guys. They're not neutral, nor accountable, and they are very powerful.

They also just demonstrated the ability to collaborate to destroy a company they didn't like without some kind of legal order forcing them to.

[jcranmer]

Twitter definitely made moves to crack down on this stuff (after all, that's why Parler thinks it was poised to gain a lot of traffic). Parler complains about the "Hang Mike Pence" hashtag, but omits the fact that Twitter in fact did kill the hashtag.

Parler's sin was basically to attempt no moderation whatsoever, and even after it was clear that several individuals intended to make good on their threats (see the invasion of the Capitol), did not seem to be willing to make any serious attempts to moderate content. See ¶6 of https://www.courtlistener.com/recap/gov.uscourts.wawd.294664... for some discussion.

(Note that the executive's name in that declaration is redacted because (s)he is legitimately concerned for violence should their name be made available to the public.)

[newacct583]

> but aren't many violent acts planned on Facebook and Twitter?

Shouldn't anyone who plans a violent act on Facebook or Twitter (or Reddit, or HN, or wherever) be banned?

Note that both of those sites are ALSO in the process of desperately cleaning up their violent sub-communities. There's a somewhat humorous running gag right now of conservative thought leaders suddenly complaining that their follower counts are going down.

And as far as Parler specifically: it didn't "suddently get popular" in a vacuum. Parler absolutely was functioning as a kind of "ban evasion mechanism for Twitter". People who got tossed from Twitter (generally, yes, for violent rhetoric) would find a home there, and bring there followers. There was very little organic growth, it was effectively all cannibalized from Twitter.

[jonstewart]

If my “editorialized” you mean “edited,” yes. But it’s not opinionated, it’s stated as fact in the filing with specific details backing it up.

[dqv]

It's editorialized. The non-editorialized title of this post would be "Parler, LLC v. Amazon Web Services, Inc."

[dgellow]

So, I'm not arguing one way or another, I just want to try out the following for the sake of discussion as I haven't seen it elsewhere: if we consider Amazon to be an infrastructure provider, and that by doing this they have a duty of neutrality, instead of shutting down Parler directly the approach would be to send a complain to a relevant federal authority (not sure which one that would be, maybe the FBI? please correct me) and let them do their investigation. If they find something considered bad enough to take the platform down, they would ask Amazon to take it offline. An argument for this is that it gives elected or at least government-related people some overview over the process instead of giving full power to private companies (which will backfire against the good people if normalized).

Please tell me why that would be bad, I'm interested to know.

[vel0city]

I expect common carriers to have a duty to neutrality and defer to at least some government agency with oversight before deciding to pull the plug.

There is one rail line to carry grain in and out of a community. That rail line is a common carrier. It shouldn't be able to arbitrarily decide whose cargo to carry or not.

There is realistically only one ISP at my home. That should be a common carrier. They shouldn't be able to arbitrarily choose which IP addresses I can connect to or not, or what ports I can choose to talk on. I realistically have no choice on who to choose for that, so there's no market for alternative providers for me.

AWS is not a common carrier. If you get dropped from AWS there's still Google Cloud, Azure, Oracle Cloud, IBM/Softlayer, Hertzner, Digital Ocean, Linode, Hostgator, Dreamhost, and so many other smaller cloud/VPS providers out there. And that's assuming you're for some reason entirely unable to roll your own hardware and move your app into a colo, of which there are literally hundreds of providers in the US alone. There is lots of competition in the cloud hosting industry, and if you widen it to the hosting industry in general its extremely wide.

[newacct583]

> instead of shutting down Parler directly the approach would be to send a complain to a relevant federal authority (not sure which one that would be, maybe the FBI? please correct me) and let them do their investigation.

It has been suggested by many people that in fact the causality goes the other way. The reason for the near-simultaneous action against Parler from Twilio, Google, Apple and Amazon seems likely to have been a strong suggestion from Federal law enforcement about an imminent threat.

There's no evidence for that per se, but certainly the threat seems imminent.

[dgellow]

I agree with your last sentence. If that's what actually happened, the situation is completely different and the general debate is really missing an important point.

[NoOneNew]

According to this, both Twitter and Facebook use aws: https://www.contino.io/insights/whos-using-aws

There have been a lot of livestreams and general calls of hate and violence on those social media sites from all political spectrums. Where are their bans? That's why theres an uproar.

New York Times, A Genocide Incited on Facebook, With Posts From Myanmar’s Military

https://www.nytimes.com/2018/10/15/technology/myanmar-facebo...

Pretty sure that can be considered violence. I think I feel that way due to the genocide part. Where's Facebook's ban? Oh wait, that's right, big tech is it's own good ol'boys club.

[Maken]

Facebook and Twitter have evident moderation (if maybe insufficient) working on removing or flagging such calls to violence. The argument of this case is not about whether there were user messages calling for execution of public figures, it's about whether Parler did anything about it.

[gfodor]

The argument is Facebook and Twitter are doing a better job of moderating, and Parler isn't or is outright not moderating their content.

This may well be true. But the point is now we're arguing a very subjective standard, and the determiners of those standards aren't a judge or jury, but a few CEOs. The standard of "when is it justified to refuse to host an entire platform based upon the nature of how they moderate their worst content" is an incredibly complex question, with no objective answer, and how these company's answer it has a profound effect on freedom of speech, since deplatforming a site like Parler doesn't just censor the speech of those who ought to be censored (by any standard), but everyone else on the platform.

It's fairly obvious where this will lead, but I doubt we will correct it before it causes immense harm (not necessarily directly, but due to the blowback that will come from censorship without due process and tit-for-tat escalation.)

[fxleach]

Twitter and Facebook both use their own data centers, with Twitter only recently (mid December 2020) announcing that they will be migrating to use some AWS. They also try to manage their content moderation as best they can and have policies and practices in place to remove content that violates their standards. Parler relies solely on AWS and has 0 content moderation. There are stark differences.