Hacker News Reader

It used to be much harder to knock someone you didn't like offline. You used to have to organize a DDOS attack. But now, all you have to do is organize a much easier mass-flagging attack. We need to update our threat models to account for how unreliable major platforms have become.

We can no longer rely on these companies. After big tech spent so much time investing in their own infrastructure, they are undermining themselves like this.

We need to just accept that a major liability of using a major platform like this is that your content may be removed and your account deleted, for unknown reasons.

Want to host your podcasts? Well don't rely solely on Apple Google or YouTube, they are not up for the job. Instead, you have to span it on multiple platforms and tell your audience where to find you if you get deleted.

This concept of de-platforming is becoming more and more prevailant.

The big techs cannot differentiate between legitimate de-platforming vs a Denial of Service de-platforming. I think while the big techs have a legitimate concern for de-platforming things that are deemed hate speech, this same power can be utilized by individuals to target a de-platforming for people whose opinions they don't like.

I think big techs should consider a public-court like system - where in order to deplatform, a jury of peers should be used to make judgements.

> The big techs cannot differentiate between legitimate de-platforming

I can try to give some guidelines, I'd love to see improvements (and also like to know if anyone points out I'm wrong :-)

Legitimate de-platforming is

a) when an individual or organization is removed from the platform based on a request from law enforcement followed by a review by the platforms staff to verify that the request is genuine and that the reasons stated are valid.

The account should still not be deleted for a certain number of days, and if requested by the user or organization, relevant details should be made available to them to take fight it in court against the relevant law enforcement agency.

b) when an individual or organization is removed from the platform based on an written policy, known by the users of the platform in advance and verified by a second team. Upon request an explanation of why the account was deplatformed should be made available.

> b) when an individual or organization is removed from the platform based on an written policy, known by the users of the platform in advance and verified by a second team. Upon request an explanation of why the account was deplatformed should be made available.

That alone would be great. They should have to justify exactly why they removed someone. And by making it explicit, they open a pathway for legal action.

But more importantly, these platforms need to be declared public utilities and forced to be transparent. This is not controversial and is usually welcomed by the public when it happens. Most public utilities that we enjoy today started out as being completely private.

The fire dept would let your house burn down if you were late with your bills. The telegraph company would ban you from their private network if you were a journalist that criticized them. The electric companies would make you agree to onerous terms to get electricity. And currently we have this mess with private medical insurance companies and tech platforms.

It is time for us to nationalize Facebook.

Edit: added quotes. Added part about not wanting any social network nationalized.

> But more importantly, these platforms need to be declared public utilities and forced to be transparent. This is not controversial and is usually welcomed by the public when it happens.

Even just a real threat that it might happen might, if it became visible enough, drive some introspection and push them in right direction.

> It is time for us to nationalize Facebook.

Assuming you are from US (but mostly valid anyway): Not sure if I'd want your president, now or in 6 years, to have more say over the biggest social media networks than they already have.

> It is time for us to nationalize Facebook.

Note that even accepting your previous paragraphs, this doesn't follow. Most public utilities are already not nationalized afaik; I'm not american, but none of my electricity, gas, water, phone or internet are provided by government-owned or -operated companies, and I'd be surprised if it was much different over there. They're regulated, but not nationalized.

Simple and fair. The above idea is what happens when you stick a reasonable, smart person without an agenda into a room and ask them to simply solve a problem without politics.

I was at Joyent when we deplatformed some group. The vast majority of the company was for this action, based upon what they interpreted as "their values." Their values were not shared in whole across the entire company. Some of us believed that speech, no matter how repugnant it is, should not be constricted.

This is the nature of the 1st amendment in the US. The government cannot restrain speech.

A private company was another matter. They can do whatever they want, within the boundaries of law. 1st amendment doesn't apply in this case.

The concern I had was that this would become "easier", and more frequent.

It has. This was the nature of my concern, that the slope would be more slippery.

Justifications for banning are getting "easier". Groups celebrate this deplatforming as "victories."

No one seems to worry much about the long term consequences of this. Well, Cloudfare did for a while, but then it stopped.

Right now, its hard for me to tell if bans are for ToS violations, or political wrongthink. I am troubled by this. ToS violations are generally fine, as long as the ToS aren't onerous. Most aren't. Wrongthink violations ... think of when the political winds shift, and when current rightthink becomes wrongthink.

Niemöller's poem[1] becomes an apt metaphor in this case.

[1] https://encyclopedia.ushmm.org/content/en/article/martin-nie...

Government can and does restain speach. It issues gag orders, classifies information, places and enforces injunctions, and facilitates private bounds on speech through patent, trademark, and copyright law, trade secrets, and recognition and enforcement of nondisclosure and nondisparagement contracts. Even forms of observation may be limited under espionage or treason laws. Limits are placed on incitement, inflammatory speech, pornography, indecency, and "fighting words".

Freedom of speech, as with other legal and civil rights, is an ideal, not an absolute. It has evolved tremendously within the US, arriving at a form we'd recognise currently only in the 1950s and 60s, about 60 years ago. The famous "shouting 'fire' in a crowded theatre" objection was raised in a case (Schenck) in free speech was found not to apply to anti-draft advocates during WWI, barely over a century ago.

And the US Supreme Court's "marketplace of ideas" standard is similarly quite problematic, owing far more to free-market fundamentalist advocacy than some doctrine of prevailing truth.

Turns out it was always more a bit of pro-free-market-fundamentalism propaganda than a pro-truth notion.

See Jill Gordon, “John Stuart Mill and the ‘Marketplace of Ideas’"

https://philpapers.org/rec/GORJSM

Or Stanley Ingber, “The Marketplace of Ideas: A legitimizing myth”.

https://scholarship.law.duke.edu/cgi/viewcontent.cgi?article... (PDF).

What big tech is doing so clearly violates the principles of free speech that I can't see an argument for it. It goes well beyond people expressing unpopular opinions, to the point of interfering with markets and elections.

Twitter has been interfering in elections by selectively deplatforming candidates that they don't like. It is so obvious that the biggest threat to the integrity of American elections is big tech. I can't believe that there are some that think that Russia buying a few thousand dollars worth of Facebook ads is a bigger threat than Google.

Today, we have autonomous engagement driven feedback loops that are easily, trivially gamed.

A little money can make a huge difference.

There already are laws, regulations and courts that rule on the exceptions to free speech. This should never be in the hands of a private company to decide.

These 'platforms' perpetually dance around issues and present a different face dependent on who they talk to on what issue.

>We need to just accept that a major liability of using a major platform like this is that your content may be removed and your account deleted, for unknown reasons.

No we don't. You make it sound like it's a law of nature how these companies behave. It's not.

> Instead, you have to span [your content] on multiple platforms and tell your audience where to find you if you get deleted.

OK, so what's the safest holdout?

Bitchute seems to be popular.

Mu.[1]

The principle of "spread across multiple platforms" is fundamentally inconsistent with a "safest holdout".

You need to hedge your bets.

You need to realise that free services can be discontinued, paid contracts can be terminated, content or accounts can be flagged, copyright or patent claims may be asserted (more the former than the latter, generally, but expect patent claims on more capable systems), domains may be hijacked or squatted, and self-hosted systems may be DDoSed, require ongoing maintenance, moderation (if user / third-party content is permitted), etc.

POSSE -- Publish on your own site, syndicate elsewhere -- is probable the most robust option overall.

https://indieweb.org/POSSE

________________________________

Notes:

1. A Japanese word, often translated as "unask the question". https://www.awakin.org/read/view.php?tid=583

Yes, what I've been saying is that we're moving back to where we were 20 years ago. We're moving back to a hosting model where you can not completely rely on any hosting company to get the job done.

Big tech companies changed things a lot, and developed very sophisticated infrastructure that we could not compete with.

Nobody could compete with YouTube for awhile. But now, YouTube has made itself so unreliable, that people are forced to find new video hosting services, and people will put up with slower speeds for more reliability.

> We're moving back to a hosting model where you can not completely rely on any hosting company to get the job done.

That's become my mantra about anonymization: that you can't trust any one system/provider, so it's prudent to distribute trust. Basically, what Chaum said, so many years ago.

But that's harder with hosting. Sure, if you lease a domain, you can easily point to different IPs. But then, what happens if you lose the domain?

And for people who don't lease domains, how will your followers find your other sites? Once the one they were hitting is gone, I mean. Some will bookmark all of them, for sure. But not most, I bet. Search services would help, but I suspect that'd be iffy.

This is exactly how I've been treating GitHub lately; in addition to my own SSH-enabled server, I have both GitHub and Bitbucket configured as remotes (strongly considering adding sr.ht to the mix), and every push pushes to all remotes simultaneously (well, technically sequentially, I guess).

We also need to update our definition of freedom of speech. It's commonly repeated on the Internet that freedom of speech doesn't cover private platforms. That's true because in 1791, when the First Amendment was written, only the government could impact individual freedom of speech. But now large companies can do it very effectively because they own the platforms everyone is using. Sure, you can still express yourself by shouting on the street or self-publishing a book, but that isn't a at all in the spirit of the Constitution. In practice, individuals can be silenced.

The principle of common carriers, over which all (lawful) conduct is permitted, including postal, telegraph, and telephone services, as well as equal-time and community-access broadcast services, offer alternate models.

I'm finding the "private platform, no free-speech rights" argument increasingly thin on platforms aggreggating hundreds of millions or billions of eyes and ears, and increasingly inextricably bound to social, commercial, and governmental institutions and activities. At the same time, the wild west alternative has also proved untenable, with strong civil-rights groups including EFF and ACLU increasingly acknowledging and accommodating this reality.

There's a middle way.

That’s exactly the spirit of the constitution. Back then few people could afford a printing press, so your ability to reach the masses could easily be frustrated by the refusal of private entities that did.

But the founding fathers never thought it necessary to regulate printing presses to ensure the access of minority opinions. The First Amendment was explicitly written to prevent that kind of social intervention.

I think they’d tell you, set up your own server and your own service, it’s not hard.

Yeah, I think there are two broad relationships between a person and a company: personal and statistical.

At my corner grocery store, it's personal. I know all the staff, we say hello, we look out for one another. My doctor knows me as a person, and we talk as equals. You see the same thing between companies of similar scales. If my company is a significant customer of yours, we can talk as peers when we need to.

Google Youtube and Twitter, on the other hand, are mainly selling eyeballs to advertisers. Not only do they not care about any given individual, they couldn't if they wanted to. The ratio between users and workers is 10,000:1 or worse. At that scale they know they're going to screw over people; at best, at best, they make sure the errors aren't too biased in any one direction. And of course they will work hard to clean up PR problems, as those are expensive enough to justify actual attention.

I think the underlying issue is that Google is under a lot of pressure from the EU and other governing bodies to police its YouTube content, Google being Google is trying to do this with algorithms, unfortunately when the algorithm gets something wrong there is seldom much recourse beyond knowing someone on the inside or being able to raise a big enough stink on Twitter.

But this is nothing new, so I think you have to go into things planning for the day Google demonetizes your videos or suspends your account for no reason. Make sure you have backups in GCP, S3, or some other non-Google facility that your kid won’t unplug while playing. Learn how to serve video yourself and have that as a backup - when promoting your video link to a resource you control so you can direct people to backup sources when needed.

Thanks to Apple there are only two, more or less interchangeable, standards for playing video on modern browsers, the only difference is that one standard (HLS) is controlled by Apple and the only option on iOS, and the other standard (DASH) is controlled by Google and works on everything except iOS. Most CDNs have support for HLS which means they dynamically recode video, help to generate manifests that direct users to lowest latency locations, and help to enforce entitlements (though as far as I can tell without proprietary DRM extensions all that is good for is largely keeping honest people honest which you may not care about when making video blogs or other public consumption content)

GCP is Google cloud platform isn't it? perhaps you mean AWS

GCP being a cloud service, it’s not susceptible to the same risks as YouTube. There’s no way for someone unscrupulous to “report your account” and the algorithms to be worried about are not scanning content. Instead, watchdog security systems may determine that your account had been compromised by bitcoin miners or something and lock things up until you get on the phone.

It's not just pressure from the EU, a lot of it is also coming from the USG [0], that's why Google, and pretty much any other major US-based social media, have outsourced a lot of content moderation to small armies of shadow-moderators in the Philippines [1]

[0] https://www.npr.org/2018/09/05/644607908/facebook-twitter-he...

[1] https://www.mediapost.com/publications/article/327887/pbs-th...

It's kind of like having all your stuff stored on the computer of someone who randomly breaks all contact with people for no reason.

The cloud is just someone else's computer after all.

Happened to me a few years ago. Google account got terminated with no explanation. I used it for personal e-mail, watching videos on Youtube and to store personal files on Google Drive. I never posted anything anywhere public, including commenting on Youtube. Since then I've been self-hosting everything I can.

Twitter isn't arbitrary or unknown, though. You can easily predict with near certain accuracy who will get permanently banned for a controversial tweet and who won't.

YouTube is more hit or miss.

Not sure why you are getting downvoted over this. It is pretty close to what I observe.

Express "wrongthink" and you will get banned.

That should terrify everyone. As you don't know what rightthink and wrongthink are. They can, and do, change with time.

Maybe have multiple independent google IDs for activities you want to do? So if you do YouTube videos and android apps, don’t do them on the same google ID, so an accidental account suspension in one doesn’t affect the other.

That typically doesn't work. Google antifraud detects and groups these "related" accounts, so they potentially could suspend them all at once.

I don't know what is do if I lost access to my Gmail account. I have tons of accounts on other services that only know me by that email address.

How do you backup Gmail?

You can download everything at once using Google Takeout. Or you can set up an email client with IMAP to download your email.

That seems to be for one-in-a-while use case only. The parent poster mentioned that they backs up their google data. So I am assuming some kind of regular backup. I find that hard to do with gmail. And as I understand it will soon be impossible due to their updated API/authentication?

I use imapsync, http://imapsync.lamiral.info/