So many HN replies amount to "we all agree this is a problem, but we can't fix the entire problem perfectly, and it has some hypothetical drawbacks, so we shouldn't even try."

(Never mind that as a result of inaction in the face of disinformation and hate speech our societies are rotting from the inside, and many, many real-world atrocities are being carried out as a direct result.)

This is, by the way, a fundamentally conservative viewpoint. Cf. gun violence, homelessness, living wage, etc. Just because something is a complex issue with imperfect solutions doesn't mean we have permission to do nothing.

Myanmar's language and culture are completely alien to people drafting Facebook policies, driving forces behind intercommunity violence include things like [likely at least partially true] news reports of other intercommunity violence and official government statements, and then there's nuances like Burmese people seemingly accepting the false claim the ethnically-cleansed Rohingya actually were Bangladeshi regardless of where they stand on other things, and the outpouring of support for Aung Sung Suu Kyi after Western criticism that might have been signals that they believed the conflict was the generals' doing rather than hers or might have been mass endorsement of the government's violence. I suspect my Myanmar-based Facebook friends' one or two allusions to burning villages and politicians are probably calls for peace and meditation, but honestly, I don't know.

That was largely a result of campaigning against giving rights to the Rohingya.

> the outpouring of support for Aung Sung Suu Kyi after Western criticism that might have been signals that they believed the conflict was the generals' doing rather than hers or might have been mass endorsement of the government's violence

Yeah, because Aung Sung Suu Kyi keeps denying, on live TV, that any problem exist other than the insurrectionists are responsible for everything thats happened thus far. The insurrectionists/terrorist according to her are composed of muslim Rohingya that are financed by foreign "Muslim" powers.

The matter of the fact is that most power is held by the military, NOT Aung Sung Suu Kyi. Thus, Aung Sung Suu Kyi stance on this issue is probably a result of the military's position. At any moment, the army can choose to remove her from power. Her position is that fragile.

The definition will be less than ideal, open to abuse, and problematic, but having it is better than not, just as having a definition of "adult", "self defense", and "libel" are better than not having them.

And why would we engage in a "complex negotiation" (that sounds to me like a euphemism for right/left extremists and massive unprecedented violence) to reach a "less than ideal" outcome over an issue that 250 years of history prove is not only unnecessary but in all likelihood extraordinarily dangerous??? Again, why was free speech the first one???? It f-ing works, that's why. It's the foundation for the best governance human beings have ever achieved. To have people now in 2021 just kinda shrug about its importance is mind-blowing to me.

In any event, the Internet changed things. Before, it was very difficult to light the entire country on fire.

With the Internet, and social networks in particular, that has changed. The reach is unprecedented; it is a difference of kind, not just magnitude. Anyone can reach very specific groups of people and incite hatred and violence through targeted propaganda campaigns. There's more than one reason rhetoric and mass violence has increased since the late '90s (on a national scale), but I believe this is one of the biggest contributors.

Facebook is influential yes, but they are still one private organization of many. Why do we need a consistent definition of hate speech between facebook, twitter, reddit etc?

That’s the “slippery slope” argument. If you define what’s allowed, people will ask for more, and others will push past it saying it’s not much different than previous.

And besides that, the line has been drawn many times by the Supreme Court. Hate speech is allowed by the First Amendment, but inciting violence may not be. There’s “tests” for these sorts of issues that lower courts are supposed to apply.

I like Indian curry.

Yes, Indian curry is the best curry.

Yes, Indian curry in Japan is not the same. They don’t know how to make it.

Yes, Indians are better at making curry. After all it was invented in India and we have the local spices.

Definitely, I would not buy curry inJapan unless it’s made by an Indian who knows how to make it.

I mean you can see this potentially going in an unwanted direction. But is it hate?

To me hate is when you have immediate violent outcomes in mind.

This is why I think it’s impossible to “monitor” and purge or ban violent tweets or what have you.

One, what is the intention of the speaker, two who is responsible for the audiences reaction? Do you take the most extremely negative interpretation? It cannot work.

And that’s not even taking slang into account where violent words don’t mean violence and normal words can take on other meanings.

"Let’s go and burn that police station down now"

If anyone I knew said that I would treat it as a dry joke.

Lots of people kid around about horrible things but don’t mean it. One day you find out that one of the kids was not joking and actually meant it. Even people who are friends of the perp have problems sussing this out and can only do it in retrospect.

Progressive politics will say we have to consider the background of the singer and the audience as mitigating circumstance so they might let it slide.

Objectively it would have to be banned if we use Twitter’s new rubric.

Reasonable people can disagree

> Sealioning (also spelled sea-lioning and sea lioning) is a type of trolling or harassment that consists of pursuing people with persistent requests for evidence or repeated questions, while maintaining a pretense of civility and sincerity. It may take the form of "incessant, bad-faith invitations to engage in debate".

One post is not it.

Also, if you read that Wikipedia article to the end you'll realize that I pointed you to the original use of the term.

(Indian curry is good too but most people have access to try it)

My point was that people can have discussions about who or what is better and it can diverge into other areas that could be considered hate by very sensitive people. Japan notoriously claimed for example that their snow is unique[1] and thus made it difficult to import European skis some time ago. Today this would be viewed and xenophobic or something when all it was was protectionism.

[1] kinda true. It’s ‘wetter’ than snow in some other places, but doesn’t mean Rosignol should not be sold in Japan.

As I understand it Facebook, if it did anything at all, did id months/years after the fact, essentially doing CYA not anything meaningful.

I really dislike this argument. A lot of democratic countries have defined hate speech. In the US, individual companies define it and moderate it as they see fit. In other countries, their legislators and courts define it. The US has defined lots of difficult terms to define already.

>And why ban speech that is actually legal (even if highly offensive) anyway?

I mean if we're talking about why companies should ban it, I'm sure they have a variety of reasons that range from bad PR and loss of revenue, to the founders/owners/employees don't want to build a product that's used for violence.

If we're talking about at the society level, because it threatens democracy, peace, individuals safety

>And if it's worthy of banning on a major platform, why not just make it explicitly illegal?

A lot of democratic countries already have. UK, Canada, France, Germany, New Zealand, Sweden, Switzerland (and yes, they managed to define it somehow): https://en.wikipedia.org/wiki/Hate_speech_laws_by_country

They have very poor subjective definitions that boil down to anything any group considers offensive which is a moving needle, and that make make things like satire and certain forms of comedy illegal and have a chilling effect on valid criticism. It's also compelled speech, and in some cases leaves violence as the only resort, as opposed to a conversation and de-conversion from extremist beliefs. We've seen abuse of it in several cases, if no one is offended, they'll create a group or pay someone to be offended. There's no burden of proof beyond someone's emotional state upon hearing the words. You can say something to two people of a group, one might think its funny and laugh, another might report you and call the cops.

Also, turns out you can yell fire in a theater, especially if you at least believe it to be true, and that's something courts can't determine, lots of people say things that they believe to be true, that turn out to be false. Likewise, if speech is dangerous but true, its still should be protected.

In either case, best this be settled in courts and legislation, not corporate meeting rooms that are echo chambers of opinion.

>If we're talking about at the society level, because it threatens democracy, peace, individuals safety

We already have laws against that, its why hate speech laws are usually redundant and likely to be abused and scope creep into silencing valid criticism of an individual or group of individuals who can then be offended and have you arrested, at the very least putting you through months and years of legal trouble before you're acquitted, and that's only if you can afford proper defense.

You people need to look at history. It boggles me how uneducated people are today on the context of this issue.

Even the person that did the Parlor leak is a Meiklejohnian absolutist.

I'm really confused by what your points is. You spend most of your post talking about how laws against hate speech are ineffectual because courts can't determine a persons emotions and beliefs, and then say it's best handled by laws and courts.

Yeah, Weimar Germany could teach you some lessons, apparently.

Some of the BLM counter-protests in the UK also became rather heated: https://www.politico.eu/article/johnson-condemns-racist-thug...

Of course the Belgium bombing was a political attack: https://www.nytimes.com/2020/11/27/world/europe/iran-disside...

Are you suggesting only the US has political violence?

Because of the first amendment, and how the courts have very consistently interpreted to allow hate speech.

>and how has the US managed to do so well for centuries with it being perfectly legal?

If by "do so well for centuries" you mean the US's economic output and world status over the centuries, I would argue that profiting off of Europe rebuilding itself after two world wars probably outweighed the detrimental effects of hate speech (among probably several dozen other reasons).

If you mean "how has the US done so well handling the negative effects of hate speech for centuries without making it illegal", I would argue that hate speech has contributed to some of the most shameful and barbaric social dynamics over the centuries, and the US is historically well behind other modern countries on this front.

Maliciously categorizing someone based on skin color, race, income, social status and religion in an attempt to denounce hate speech.

Poetic

Because we've generally dealt with that by deplatforming that sort of speech socially and/or in the private sector.

The history of the civil rights movement is filled with boycotts and other sorts of social pressure campaigns.

If you were a minority, it really has not done well at all.

That violates freedom of speech

I'm sure this experiment of ripping up the constitution in our de facto new online lives that usurp what the government is capable of doing will go just fine...

I think a better idea is to force them to provide an API for interoperability, alternative clients and exporting the user data. Together with the federated social networks, it will change everything: https://the-federation.info.

It's a stupid majority.

Consider email as an example of a federated network: My name, email address, and text of my emails exist today on hundreds (thousands?) of different servers. If I decided I want to "delete" my name from the global email network, that's close to impossible.

The same would be true of a federated social network, but with a more intrusive and personal set of data. Users will not understand that they cannot delete their own mastodon toots (or whatever it calls them).

This is not how federation works. You can talk to any other node without having an account there.

Sure, your private messages may be delivered to your friends' nodes, but is this a big problem?

Toots are public by design. Everyone understands that and it has nothing to do with privacy.

It kinda depends on the content of those messages, doesn't it? In the context of this OP, if my messages are organising a genocide then yes it might be a pretty fricking big problem.

> Toots are public by design. Everyone understands that and it has nothing to do with privacy.

Says you. Most users will have been accustomed to centralized services where it is possible to delete a tweet and have it gone, unless somebody else has gone to special effort to archive it or screenshot or what-have-you.

It's one thing to be a neutral platform and allow people to say what they want (even if it sometimes means things that shouldn't be said), but at least the platform here is neutral and doesn't directly profit off the bad content (or at least doesn't profit off it any more than it profits off "good" content).

It's a totally different thing where the platform starts acting as a publisher and starts promoting the bad content (since divisive/outrageous content generates the most "engagement"). Here, the platform directly profits off bad content and should be held liable for it; you shouldn't be able to have your cake and eat it too.

They are free to run algos to auto reject, sort whatever order.

This would likely be too burdensome, and the socials would almost certainly have to start charging for access, which in of itself would probably remove much of the problem.

* Break them up, and keep them small enough so that no one site can do that much damage.

* Remove safe harbor provisions on algorithmic content feeds.

* Pass a transaction tax on ad auctions to discourage free-for-attention business models.

But with real-time auctions, though, there are whole swaths of people whom the quality ads have decided are a waste of time, and now the hyenas can circle in to prey on everything that is left.

There are some ad networks out there that try to avoid the real time auction and cross-site tracking aspects of these conventional ad networks. I don't know how successful they're being, but maybe they're on to something. Something that should be encourage by public policy.

> * Break them up... How do you ensure that all companies in a vertical will remain small? What would the legislation look like? The internet certainly enables more 'winner takes all' type markets, so not sure how you consistently keep social networks small when everyone will naturally gravitate to those that have the best UX + network effects (there are surely other variables, just listing 2 as an example)

> * Remove safe harbor provisions... What does this mean? How are algorithmic content feeds currently treated relative to non-algorithmic content feeds? How would you target the legislation to not have externalities?

> * Pass a transaction tax Never heard of this, what does this mean? Who is getting taxed? Advertisers? Publishers? Website visitors? How will your proposal change incentives? What's your take on how smaller news sites are dependent on the ad model and that ad models promote websites that are open to more people (as opposed to subscription businesses which incentivize optimizing for your paying customers and giving them exactly what they want)

2. I'm not going to explain the Safe Harbor provision to you. Look it up. We already have externalities - removing algorithmic feeds would take (e.g. Facebook) back to where it was in the late 2000s, where users were in control of the content they saw. The only negative externality is that Facebook and other companies would make less money of targeted ads.

3. You've never heard of a transaction tax? I find that hard to believe. Why don't you think about other times you've paid tax on a transaction, and then maybe you can come up with answers to the rest of your questions.

2. I'm familiar with safe harbor provisions, but am asking you for more specifics. Many sites/app personalize content (aka algorithmic content feeds) for the user, based off of what will drive engagement. I understand you dislike Facebook, but if you attack Facebook for personalization, how do you not hurt everybody else that is trying to cater experiences to what their users are likely to want?

3. What's the transaction in your proposal? Who is getting taxed? There are multiple actors in the advertising system, from advertisers to ad networks to publishers to website visitors. Your glib response is a signal to me that perhaps you haven't had much in the weeds exposure to how advertising functions, and may be operating from more of a mainstream mood affiliation perspective. Which is fine, but you answered a question to someone's query for 'good ideas' for regulation.

Limit M&A by denying acquisitions that would consolidate markets, like we did for a long time in this country. You don’t have to hurt startups if you have a little bit of creativity. Create different tiers of review, etc.

You cut down on hot money by reinstating the firewall between commercial banking/insurance and investment banking. It’s not getting rid of venture capital, but it does restrict the amount of money that can be levered by them.

2. Im not worried about hurting businesses that focus on algorithmic content. If their algorithms are feeding right wing propaganda to impressionable people I want them to go out of business.

3. An ad auction is no different than any other financial contract that gets sold on an exchange. The buyer sells a contract for placement, which an advertiser buys in exchange for a promise to pay. You levy a tax on this transaction to make the overall cost of doing it more expensive and to offset the cost of externalities created. I’ve thought through it plenty, but since it doesn’t feel like you aren’t asking questions in good faith or doing the least bit of research, I’m not going to bother taking you particularly serious.

3. Thanks for answering the question this time. So advertisers are getting taxed, which will increase the cost of advertising for companies seeking to drive awareness for their products. Some potential effects of this:

* It raises the cost of advertising. This means that you'll need higher margins for the increased cost of advertising to make sense as an advertiser. Given the higher cost/margin requirement, you may see smaller advertisers drop-off, having the potential to reducing competition in different markets. This would be a good thing for big companies/incumbents.

* Depending on the size of the tax, it could put upward pressure on prices elsewhere, as businesses attempt to maintain their margins and account for the increased marketing costs.

* Tax collection will add a new operational cost for ad networks. Google and Facebook, given their size and profitability will be able to deal with this cost easier than most.

* If it reduces total ad revenue for publishers, you'll likely see a bunch of smaller companies go out of business that depend on ad revenue. Smaller and local news/content sites will be hurt, leaving fewer sources, and further entrenching the major news companies. Ad revenue models by their nature promote strategies to appeal to the widest audience possible, in order to maximize impressions/engagement. The alternative model in news (just to use as an example) is subscription, which doesn't have that same incentive. The incentive is to maximize paying customers. In some niche areas you may likely see higher quality (which we already see today). But some of the downsides are: less privileged people won't have access to the same information that they may have previously, increased polarization & bubbles as publishers cater to exactly what their paying customers want to hear (and thus continue paying for).

They're already entrenched. What exactly are you worried about? You are just playing the devils advocate without actually having a point.

> It raises the cost of advertising.

Huge assumption that nothing better takes the place of algorithmic CPC-based advertising, or that a marketplace based entirely on ad-driven consumption is even desirable. It's not even worth discussing the rest of the speculation, because I reject the premise. The well-being of these ad-networks is no where near the top of my priorities list.

2. Sounds like a great way to encourage site moderators to bring down the ban hammer hard and fast.

Clearly, there is flipside. Private Facebook groups to share hate speech or otherwise echo awful thoughts. Legions of bots controlled by political organizations or nation-states trying to divide and conquer a population. It is a losing battle to moderate these platforms.

Facebook has no inherent right to exist as it does, or rather, make a profit as it does - and I wonder if extreme measures, such as a partial repeal of section 230, should be considered. Like all laws, small-p-progressive measures should be taken. Small and medium websites should not necessarily be held liable for content posted by others. But as a network grows to have hundreds of millions of users, perhaps they should?

I'm not sure either.

I feel that if we collectively understood social graph theory better we could propose tweaks to make these networks more "fair" and by fair I dont mean with a respect towards a point of view, but rather less likely to create emergent, undesirable social phenomena like ISIS, genocide, white nationalist rebellion.

But I have never seen any papers that explore these topics, on the other hand my interests have been elsewhere.

https://polandin.com/51388314/justice-minister-announces-onl...

  Under its provisions, social media services will not be allowed to remove content or block accounts if the content on them does not break Polish law. In the event of removal or blockage, a complaint can be sent to the platform, which will have 24 hours to consider it. Within 48 hours of the decision, the user will be able to file a petition to the court for the return of access. The court will consider complaints within seven days of receipt and the entire process is to be electronic.

https://www.amnesty.org/en/countries/europe-and-central-asia...