I'm interested in knowing how you think a real "free speech" platform can actually work, however. We have message boards that do this and they are just toxic cesspools. The idea of an online "public square" that isn't that sounds impossible. How many public squares are reaching millions of people instantly?
Parler was more than happy to moderate speech, even political speech, as is their right as the platform creator. Their breathless claims about being free speech absolutists is absolute nonsense.
I'd like to know the answer to this, too. I wonder if the reason why they turn into toxic cesspools is precisely because the only people who use free speech platforms are the people who were kicked off the others.
If you accept that this is plausible, then is it feasible that the more reasonable folks that just want to talk about politics in a less divisive manner (or maybe not even politics at all!) might help bring down the temperatures if everyone was swimming in the same pool instead of just a few extreme viewpoints forced to move into the same swamp? (NIMBY!)
This would also be in keeping with that classically liberal axiom, "The remedy is more speech, not enforced silence." (Supreme Court Justice Louis Brandeis, the Democratic justice who is credited with creating the "Right to Privacy" in 1890: https://en.wikipedia.org/wiki/Louis_Brandeis)
You don't have to wonder, we've seen this time and time again with virtually every open community. Without tireless moderation, the swamp grows.
In other words, the majority of people who use free speech platforms have already answered your question: they've shown themselves unable to co-exist with (a much larger number of) reasonable folks and were kicked out.
Suppose that the dominant platforms (e.g. Facebook) are not free speech platforms. They boot off a hundred thousand people. 10% of them actually deserved it and are militant jackasses who ruin everything.
Now someone else creates a "free speech" platform. Everybody is allowed in. Well, 80% of the initial users are going to be a subset of the ones who got kicked off of the incumbent platform, and 10% of those are jackasses, so your platform is now 8% jackasses. That's a huge percentage and it's going to immediately turn into a dumpster fire because the jackasses will drive out ordinary people and become an even larger percentage. There are plenty of instances of this happening, e.g. Voat.
But suppose you go the other way. Somehow get a large number of ordinary users. Now the jackasses are only 0.5% of the users. Combine this with something like a voting system so that nothing is ever actually removed, but spam and fascism end up at the bottom of the feed where nobody sees them by accident.
That has the potential to work. The key is to somehow get enough users to dilute the jackasses before they take over, e.g. because the incumbents overreached and a large number of non-jackasses are moving in protest.
Moderation didn't come before toxicity it came in response to it. Therefore, moderation doesn't cause / focus toxicity.
If you want to address this, you need to look at education.
It turns out that people don't, as a general rule, actually enjoy using totally unmoderated fora; they tend to quickly fill with spam and awful stuff.
It worked great for multiple decades until "social media" applied algorithms that promoted controversy (i.e. anger-inducing hyperbole and conspiracy theories) to increase "engagement" and sell more ads.
That you can find an ASCII swastika or goatse on Slashdot which is instantly downvoted to -1 (but not actually removed from the site) was never a real problem. That Facebook put QAnon at the top of your mom's feed was a major problem.
But then we get calls for censorship as a response to problem created by bad moderation.
Notice that there is a difference between voting (where the community determines prioritization in a decentralized way but nothing is rendered unavailable) and censorship (where some fallible central authority is deciding what people are not allowed to know).
Turns out outraging people increases engagement so the “jackasses” get amplified.
Arguably FaceBook is too.
I’d love to see a decentralized free speech platform. I’m no blockchain evangelist, but it would seem like such technology could be used to build a real online “public square” as you say.
Exactly. Alt-left speech was not welcome on parler - only alt-right speech.
I will, however, point out, that much of the alt-right is terrified of the antifa boogieman... And that Parler, the platform of free speech was quite happy to ban anything that smelled of antifa... As well as much that could, be described as moderate leftism.
Its tolerance for speech only extends to a very narrow slice of the political spectrum.
What gets blocked on Parler is more appropriately called "not supportive of white nationalism." That way you are sure to capture the entire space of content, whether Alt-right or not, instead of narrowing the universe to Alt-right and things the alt-right is encultured to fear but don't appear in reality.
Patterns of negative conversation defined by fewer quality reply posts get negative points and vice versa for positive or patterns to be encouraged.
As things evolve and how people respond changes these patterns can change.
With free speech, as defined now, low quality arguments hit the same bar as high quality ones.
Yes we should strive to emphasise high quality arguments (even though I fear this is a near impossible task programmatically) but this is not what free speech means.
Tldr: according to free speech the dummies and assholes deserve to be heard.
That we're even having such a high-quality discussion now really speaks volumes to the seemingly effortless way in which the HN team has developed into a great place to have such a conversation or debate. To be honest, I do not think that HN is always fair to everyone, but it does seem like the moderators do try to be.
With that said: Twitter and FB have wildly missed the mark.
Technological discussions mostly stay high quality but any politics(-adjacent) topic turns into an obvious struggle of moderation and quality.
The same goes for Twitter but the balance of topics is heavily skewed towards the latter.
That being said, I do agree with GP on this point:
> To be honest, I do not think that HN is always fair to everyone, but it does seem like the moderators do try to be
You have no way of knowing if this is a true statement. He never said whether he used the app. Maybe he's speaking from personal experience with the app. You should focus on debating the argument and not introduce assumption as fact.
Exactly. And the problem was the users did like it. Lin Wood called for the VP's execution just days before a real mob took the capitol chanting "Hang Mike Pence!". And it was hugely popular and shared everywhere. And that's a problem that needs to be fixed, not celebrated. Major thought leaders across the platform were fanning the flames, not engaging in moderation.
Maybe it's true that Parler had a scheme for moderation. But it was objectively a complete failure.
Let me give this a shot. There are two types of content that need to be removed: (1) spammy content that readers themselves don't want to see, and (2) illegal content that society doesn't want anybody to see. I think these two need to be addressed separately.
Illegal content includes copyright infringement, violations of NDAs, libel, slander, perverting the course of justice (violations of court orders), incitements to violence, consipracy to commit a crime, exposing troop movements, and in general anything that directly causes damages which could be assessed and recovered in a court of law. Racism, sexism, homophobia, "hate speech", advocacy of violence, falsehoods, trickery, promoting very dangerous ideas, lying about (or being incorrect about) vaccines, lying (or being incorrect) about who won an election, offensive speech, and jokes of all kinds are, at least in America, all forms of protected speech.
Illegal content must be taken down when ordered by a court to do so. AFAIK the platform isn't responsible to take anything down unless ordered by a court to do so, but I'm probably just ignorant -- PLEASE dont take legal advice from some random armchair pontificator on the internet like myself. Section 230(c)(1) protects the platform as they are not deemed to be the speaker.
Most people are terrible judges of whether something is illegal or not. For example, many congresspeople think Trump's Jan 6 speech was illegal (it wasn't) because it incited violence (it didn't). It didn't even advocate for violence. The idea that it was illegal is so far beyond Brandenberg to be laughable if it wasn't so serious... but back to the point... Attempts to proactively take down illegal content will invariably take down some legal content, probably lots of legal content, which is why a true "ideal" free speech platform would not attempt to do this... one must have humility and recognize one's near utter inability to determine what is and what is not illegal to any reasonable degree of accuracy during times of political strife like we find ourselves in today. The popular opinion right now is that all of us should enact justice upon each other according to our own personal interpretation of justice... which is just nuts when you think about it. Let's leave law enforcement up to the law enforcement professionals.
The first issue, the spammy content, would then be handled the same way we handle email spam. Plug into the spam filtering service of your choice, subscribe to rulesets, or write your own. I don't know of any platform that lets you plug in your own moderation, and so that's where I think they've all gone wrong - they either get flooded with spam or antisemites and become nearly useless to nearly everybody, or else some moderation team thinks it is their role to filter spam on behalf of the community and the community gets pissy that they filtered the wrong things. These two extremes are both wrong. Users can filter their own content if given the right tools. BTW, Section 230(d) requires providers notify their customers that parental control protections are commercially available... it puts the onus on the users to solve this for themselves... even way back when that law was written.
I'm going to add a third issue, flooding. That has a content neutral solution: throttling.
Of course you are correct, but that doesn't make me factually wrong. You are talking about publishing illegal content. But (c)(1) "No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider."
And I missed DMCA takedowns.
With this statement you are implying one of two things:
* You are very unhappy with AWS for de-platforming them
* You think AWS should not have been allowed to do so
If it's the first one, well, you are free to treat that however you like. Stop using AWS yourself, especially if you now fear AWS will do this to you. Don't do business with other companies that use AWS -- you can even block all AWS IPs at your firewall. Tell everyone how bad they are, and encourage others to boycott them too. If AWS continues to de-platform customers that don't seem to deserve it, this will get even easier over time.
If it's the second one though, that raises all kinds of tricky issues. If you run anything that allows user content, and a user posts something that hurts you, are you now obligated to host it (leaving aside illegal content and the tricky jurisdiction issues that go with that)? What if other major customers or users start leaving because you're hosting said content -- are you just supposed to let your site/business effectively die? Are you also forbidden from raising the price to that harmful user to $1-trillion-per-month (since that effectively is a way to "de-platform" them), and if so, what is the maximum price you're allowed to charge and who figures that out?
Twitter and Partler are content platforms, AWS is not.
They are an infrastructure, not a content company and there is a material difference.
Ford, GM and Telsa collectively refusing to sell to XYZ customers for whatever reasons gets out of hand very quickly.
'You can just use another service' generally doesn't work as an argument in oligarchic situations. If AWS style services were as varied as grocery stores, maybe, but they are not.
Asking a customer to stick to the rules they agreed to and eventually ending the business relationship when the customer chooses not to is hardly a 'partisan act'. Parler could have just done what AWS asked (even temporarily, to keep themselves online while looking for alternatives). They didn't.
"The Islamic State of Iraq and ash-Sham (ISIS) continues to use social media as an essential element of its campaign to motivate support. On Twitter, ISIS’ unique ability to leverage unaffiliated sympathizers that simply retweet propaganda has been identified as a primary mechanism in their success in motivating both recruitment and “lone wolf” attacks." [2]
Apparently US authorities were happy to want Twitter to keep those accounts alive for some time as they could use them to track activity - the point remains it's wrong to the point of naive to believe that these decisions are not political.
Facebook was enabling genocidal activity [3].
It's incredibly hypocritical to look at the historical activities of ISIS on Twitter, Facebook and all of these platforms, and then to hear about the 'political violence' caused by Parler.
I don't like Parler, the hypocrisy is staggering.
[1] https://en.wikipedia.org/wiki/Use_of_social_media_by_the_Isl...
[2] https://journals.plos.org/plosone/article?id=10.1371/journal...
[3] https://www.nytimes.com/2018/10/15/technology/myanmar-facebo...
The difference is, there was actual moderation. Parler, a tiny (2 orders of magnitude tinier than Twitter) appears to have had no moderation infrastructure at all.
Even the data in your first reference seems to suggest that Twitter alone has probably moderated more accounts than Parler has actually had.
Broken record: I'm still surprised so many people here have so little exposure to the hosting business to believe that there's actually an inviolate right to due process in keeping your stuff hosted somewhere. Hosting providers have to kick users off all the time. It is a basic fact of life in hosting: people abuse stuff.
Parler wasn't running a phishing scam, spamming, or DDoS'ing people, but that's not my point; my point is: if you're a hosting provider, your contracts basically have to give you the right to boot people at your discretion. The idea that anyone thought AWS's might not? Baffling.
The following is an excerpt from https://www.courtlistener.com/docket/29095511/34/parler-llc-... the order denying the injunction.
"AWS has submitted to the Court multiple representative examples, reflecting content posted on Parler during this period, that AWS claims violated the terms of the AUP and the parties’ Agreement. ... Parler has not denied that these posts are abusive or that they violate the Acceptable Use Policy. "
You obviously did not see the material that the company allowed to be posted on their services in contravention to their hosting provider's terms of services.
Always abide by your host's AUP. Otherwise, you can get yeeted.
AWS AUP: "You may not use.. or instruct others to use, the Services.. for any.. offensive use"
"Offensive"? That word is so vague as to be meaningless.
This turns into a dangerously simple attack. Just post something you know will offend someone. Et viola: you just killed a company.
It's a lot cheaper and easier (and less likely to involve swat teams) than hiring some sketchy DDoS outfit.
AWS arguments:
https://twitter.com/KYWeise/status/1349200096345296897?s=20
> Parler itself has admitted it has a backlog of 26,000 reports of content that violates its (minimal) community standards that it had not yet reviewed
https://twitter.com/KYWeise/status/1349203942614335488?s=20
> When I said Amazon cited many vile and violent examples of posts that it flagged for Parler, I meant it. https://pbs.twimg.com/media/ErlVjN7U0AEr1gt?format=jpg&name=... https://pbs.twimg.com/media/ErlVjPBVEAEs9sg?format=jpg&name=...
https://twitter.com/KYWeise/status/1349202898459115520?s=20
> Such a requirement also poses a risk to Amazon itself, with posts calling for others to "burn down Amazon delivery trucks" until they "reverse course."
Amazon shouldn't be making these decisions, and certainly not without giving the other party ample time to make other arrangements. This should have been litigated in court prior to taking deplatforming actions.
Imagine if Amazon was a Chinese company and Parler hosted information about the plight of the Uyghur people.
Community websites have the right to enforce rules. I'm not so sure common carriers do, and AWS is a common carrier. A good rule of thumb for American hosting companies is if the FBI won't seize the website for the content hosted therein, the website shouldn't be removed.
For context, I'm a liberal and dislike the content on Parler. Defending free speech is more important than zapping the republicans. Imagine if the tables were flipped and talking about abortion was what conservatives wanted to ban.
Amazon is still working with Parler to export their data off of AWS.
https://www.buzzfeednews.com/article/johnpaczkowski/amazon-p...
> We will ensure that all of your data is preserved for you to migrate to your own servers, and will work with you as best as we can to help your migration
> not without giving the other party ample time to make other arrangements
Going entirely offline like that can easily destroy a web business. Or at a minimum seriously harm them.
Not to say I care much for Parler's technical nor administrative approach to providing free speech.
Parler had ample time. They ignored warnings, they did not address their TOS violations. Then their contract was terminated, and AWS is helping them migrate after the fact. Parler displayed an ignorance of the law and how web hosting works.
Web companies that repeatedly violate the TOS of their hosts deserve to be at the mercy of their failover plans. That responsibility is borne by the web company, not the host.
Their position is usually to not rely on non free software
What happens when AWS and Azure design their own silicon, outstrip other hosting providers, and then become the only hosts on the internet?
Are they common carriers then?
I'm old enough that racking my own boxes and hosting everything myself is both natural and easy (and cheaper and faster).
But working with plenty of young developers, most don't have any exposure or interest in setting up anything other than AWS services tied together.
So like it or not (I don't) we're rushing towards a world where AWS (maaybe GCP, mostly not) is the only way to build something.
It's good to consider the implications of this.
By volume, GoDaddy is actually a bigger host than AWS by a factor of 2.
The HN crowd vastly overestimates how big AWS is compared to less sexy options.
0: https://www.zdnet.com/article/godaddy-q3-revenue-beats-expec....
1: https://www.cnbc.com/2021/02/02/aws-earnings-q4-2020.html#Bo....
In a very real way, a founder today has more options for hosting than they do for catered lunch.
Except that's not how it works in the slightest. You have moderation in place to work to prevent that. You are doing your due diligence. You are making an honest effort to solve this problem. Mistakes will happen. Nothing is perfect, but you are making an honest attempt, and it's reasonably effective.
Parler was not doing anything close to that.
Your comment operates under the belief that these things are binary. Simply post one thing someone can find offensive, and "I win!"
But most people, most successful companies, are smarter than that and operate in reality.
>This turns into a dangerously simple attack. Just post something you know will offend someone. Et viola: you just killed a company.
What's the probability at any given time that someone on HN is offended? YCombinator seems to use Amazon's Route 53 and would therefore be at risk.
Domain Name: YCOMBINATOR.COM
Registry Domain ID: 147225527_DOMAIN_COM-VRSN
Registrar WHOIS Server: whois.gandi.net
Registrar URL: http://www.gandi.net
Updated Date: 2020-02-15T20:11:05Z
Creation Date: 2005-03-20T23:51:07Z
Registry Expiry Date: 2021-03-20T22:51:07Z
Registrar: Gandi SAS
Registrar IANA ID: 81
Registrar Abuse Contact Email: [email protected]
Registrar Abuse Contact Phone: +33.170377661
Domain Status: clientTransferProhibited https://icann.org/epp#clientTransferProhibited
Name Server: NS-1411.AWSDNS-48.ORG
Name Server: NS-1914.AWSDNS-47.CO.UK
Name Server: NS-225.AWSDNS-28.COM
Name Server: NS-556.AWSDNS-05.NETThe way you're wording this sounds very much like you haven't actually seen the kind of stuff that was going on there. It's not simply a matter of people being offended.
Regardless, this is besides the point: all a competitor has to do is post some garbage and point to it, and boom, you're dead. That's not ok.
In this case, they pointed to it weeks prior and it wasn't removed.
That is not what happened however. That is quite manipulative framing spread by Parler.
What happened was that Amazon itself identified those few dozen posts and asked Parler to remove them. Parler did not done that and claimed it is difficult for them. Amazon is not Parlers competitor. Amazon gave Parler plenty of time to remove those posts. The issue was that Parler was unwilling to do so, because they whole thing was to be safe place of exactly that sort of comments.
Twitter does not have too good moderation, but they are not refusing to take down inciting posts on principle. They did regularly took down accounts and tweets in the past. There is difference between not doing it perfectly and refusing to even try.
However, I don't think Amazon should interfere in customers' businesses and harm the relationship between customers and third parties (the users). What if Amazon had said, "Netflix, we find the Cuties movie offensive. We have shut you off and will delete all of your data within 24 hours."
That is exactly what is in Amazons court response. You either did not read it or are lying.
https://i.redd.it/om90nwadqca61.png is a Parler post with a racial slur and 25k upvotes. Twitter has its issues but an equivalent comment isn't going to survive two days.
The court documents from their case against Amazon are available at https://www.courtlistener.com/docket/29095511/parler-llc-v-a.... Amazon's declarations contain other examples of content that wouldn't survive on Twitter.
> Regardless, this is besides the point: all a competitor has to do is post some garbage and point to it, and boom, you're dead. That's not ok.
According to https://www.courtlistener.com/docket/29095511/13/parler-llc-... it took several months of moderation problems before AWS dropped Parler as a client. If Parler had done more active moderation and tried harder to keep Amazon happy they'd still be hosted on AWS.
The truth had come out, hadn't it? Even before the 2020 election, people had been mocking the network for arbitrary post removal and the requirement to show official IDs for at least some features.
Services that chose not to host Parler were exercising their freedom of association; they were _communicating_ that they did not wish to support Parler.
You have free speech, but zero freedom from consequence.
That's about the most Orwellian meme that is making the rounds and needs to be examined critically. Freedom to do a thing implies protection from consequence, or you're not free.
Also there were articles on HN about how the Capitol riots were organized on Facebook.
This is a wild take on freedom. Do you seriously believe that free speech means that you should be able to say anything and be protected from any and all consequences?
It's so vague, these one-liners, that we're talking past one another. Surely you don't think you should be able to doxx or punch someone in the face that has extremist views?
How could you possibly believe this in good faith? Consequences can also be, for example, people expressing their distaste in your speech and choosing to boycott your products if you're a business.
You said protection from consequences, not protection from criminal violence and harassment. If you meant protection from criminal violence and harassment, you should have said protection from criminal violence and harassment.
How about the real world consequences people could potentially be referring to: harassment, doxxing, job losses and violence. I've met plenty of 'punch a Nazi in the face' people to know that many people believe that as a consequence.
If signal gets deplatformed get your pitchforks. Till then I'm not worried. Also regarding the slippery slope, let it slide. If this is how we start regulating cloud services as utilities then so be it.
This is just false. Did you have an account there?
Parler should not have been de-platformed. That was a clearly partisan act, and the truth would have come out shortly about the company's total lack of care for its users. Say what you will about Gab, but they never pulled any of those stunts. Looking forward to whatever the new solution is.