Paul Cantrell
Re YouTube’s decision to re-allow disinformation about the 2020 election:
I’ve seen comments to the effect of “lies and propaganda aren’t free speech.” That’s exactly wrong: they are. We •want• them to be. And that’s the heart of the problem here.
Free speech is a contact, and Google isn’t holding up their end of it.
Huh?! Let me explain my thinking.
1/
Paul Cantrell
In the US constitution, “free speech” means that the government can’t criminalize political dissent. We •do not• want the government to have the power to decide what is and is not a lie, to decide what is and is not propaganda, and then to declare lies and propaganda illegal.
Because if the gov does have that power, I 100% guarantee that it’s not the Nazis they’ll use it it against first.
2/
Paul Cantrell
The First Amendment is one ingredient about why, for example, police in Atlanta have to go through all these legal gymnastics to strike back at activists they don’t like: https://unicornriot.ninja/2023/three-atlanta-activists-arrested-home-raided-over-bail-fund/ They’re spinning these trumped-up charges, trying to turn fundraising into fraud, and their charges just don’t have a legal leg to stand on.
But if the gov had the power to decide what’s “lies” and “propaganda,” and then make that illegal? Then the police could just go straight for the jugular.
3/
Paul Cantrell
Here’s the thing: we really do need lines around speech. Not everyone and everything deserves to be heard. Some speech should be unacceptable. Some speech should have consequences.
Dig deep enough with anyone who considers them a speech tolerance absolutist, bring them face to face enough real-life situations, and you’ll find that there is always, always a line for them somewhere.
There needs to be a line. But we don’t want the gov drawing it (because authoritarianism). What do we do then?
4/
Paul Cantrell
The answer is that the lines around acceptable speech can be, must be, socially negotiated.
The First Amendment says that we will keep the government’s power to regulate speech to an absolute minimum ••with the expectation that society will take on that responsibility.••
•That• is the social contract of free speech in this country.
5/
Paul Cantrell
If you know the 1A, you know that it doesn’t give anybody the right to be heard, or the right to be free from moderation, or the right to say anything in somebody else’s newspaper, or the right not to be kicked off of social media, or the right not to experience negative consequences for speech.
The 1A says that the government will not make those determinations ••with the expectation that society will••.
The 1A only works if society holds up its end.
6/
Paul Cantrell
Questions about where to draw the line around what speech is acceptable are really, really hard. And dangerous!
That’s why we want many news organizations, many web sites, many spaces that are varying degrees of small and large and public and private — and social forces acting on all of them — so that we can continuously hash out these questions.
7/
Paul Cantrell
This means any actor who controls a space where speech occurs needs to be ready to engage with this messy social negotiation of speech. That’s why, for example, every Masto instance needs to have a moderation plan in place from the get-go.
The larger the actor, the greater the responsibility.
Which brings us to YouTube.
8/
Paul Cantrell
We live now in a world where a handful of private entities — Facebook and Google the prime examples — have such massive domains of speech under their private control that their power over speech seems almost government-sized.
This creates an untenable situation. If they apply any sort of standards to speech, aren’t they almost like an oppressive government?
9/
Paul Cantrell
This line of reasoning is of course a fool’s trap. Every time one of these massive privately controlled speech spaces tries to adopt a 1A-shaped moderation policy, they become the Nazi Bar.
The 1A works only because the government doesn’t control the whole space in which speech occurs — and these private entities do. They cannot escape the messy responsibilities of moderation.
10/
Paul Cantrell
It’s paradoxical: when a private space adopts a speech policy that •sounds• like the First Amendment, they are in fact •undermining• the fundamental premise of the First Amendment.
When YouTube says they’ll allow disinformation about the 2020 election, they’re dropping their end of the 1A contract.
And their mistake gets worse.
11/
Paul Cantrell
YouTube has a recommender system. They aren’t just allowing Nazi shit to flow through their pipes. They have an armada of computers that (by design or by accident, doesn’t matter) are figuring out who’s most susceptible to Nazi shit, and then pumping the shit straight into their eyeballs.
That’s a whole new category of problem.
12/
Paul Cantrell
YouTube isn’t just pulling the plug on that essential process of the social negotiation of speech. They’re replacing it with machine negotiation.
Whose voice is amplified? elevated? ignored? shunted off into the void? These are essential questions at that heart of the social negotiation of speech boundaries — even more essential than what is technically allowed, IMO — and YouTube is letting machine learning decide it.
13/
Paul Cantrell
Machine learning is new. It is quirky and surprising. Its consequences when we stick it into a feedback loop involving humans are wild and unpredictable. We barely know how to •think• about this stuff, much less what to •do• about it.
I don’t know exactly what Google should do. I don’t have all the answers.
But I do know that throwing up their hands ain’t it.
The First Amendment requires Google to take up their responsibility, and deal with the shit they’re pumping through their pipes.
/end
Paul Cantrell
@kissane
Yeah. I truly do not know what exactly that means, but it’s very clear that our current systems were not designed to deal either with actors with this kind of consolidated global power, or with the cascading effects of computers being in the middle of everything.
Erin Kissane
@inthehands I keep thinking that the terms of the bargain here should be much, much clearer
Paul Cantrell
@donmelton with the r! Thanks, edited to fix
Don Melton
@inthehands Did you mean free speech is a social "contact" or "contract?"
Óscar Morales Vivó
@inthehands we can expect these giant conglomerates to design and apply their policies in a way that will maximize their goals (max money and rentism, min risk)
They won’t ban the nazis because enraged engagement drives clicks and because they don’t want to believe the nazis could take over and if they did it wouldn’t be bad for the bottom line.
Calls for the communist revolution or to dissolve their corporate charter may get a different treatment 🤷🏽♂️
Paul Cantrell
@MyLittleMetroid Monopoly(ish) power is certainly a factor here. If becoming the Nazi Bar threatened their user base, they’d have reason to avoid that. But it’s not really effectively possible to “leave” YouTube or Facebook. Delete the app, delete your account, but they’re still ubiquitous, still profiling you, still the conduit for information you need to function in society.
John Blair
@inthehands this is a great thread and captures the nuance of 1a.
Paul Cantrell
@jpallas @artt
Yes, 230-shaped protection for hosting but •not• recommending is my desired policy exactly.
230 is about liability, not moderation. I would not agree that recommending is what induces moderation responsibility. Mastodon instances, for example, also require moderation.
Joe Pallas
@artt @inthehands I’ve heard the suggestion that the line between hosting and recommending should be the line where the section 230 safe harbor ends, and if you recommend content you are a publisher with all the attendant liabilities. If google were contemplating being sued the way Fox News has been, maybe their responsibility to their shareholders would suddenly align with their social responsibility.
Arthur T
@inthehands Well said, though I think that the responsibility come from the recommendations, not the hosting. As soon as youtube starts recommending content instead of just being a searchable archive they are effectively exercising a form of editorial control, albeit an indirect one. I wouldn't expect my ISP to moderate content because they only give me what I ask for. My relationship with google is different, they tell me what content to consume so my expectations with them are different.
GNU social JP is a social network, courtesy of GNU social JP管理人. It runs on GNU social, version 2.0.2-dev, available under the GNU Affero General Public License.
All GNU social JP content and data are available under the Creative Commons Attribution 3.0 license.