A civil suit for $2 billion which alleges that Facebook's algorithm promoted popular posts with lots of engagement more than it promoted unpopular posts is frivolous and deserves to be tossed.
The article expressly states that violent content is not allowed and was actively policed. Facebook moderators are not fully omniscient and omnipresent. That is not grounds for a $2 billion lawsuit. That fact that people died is sad, but ultimately immaterial to the arguments here considered.
That's a good answer to this. They should be responsible for what the algorithm decides to inject into people's feeds, when they are acting as a publisher.
If they don't have enough resources to review the outputs from their algorithm, then to bad, just show posts from friends only.
There will always be false negatives. Is one false negative enough to drag them into court? So every website on earth will just have that one moment where they accidentally promote (or don't actively de-promote) something toxic, and they'll get to go to court over it. I wonder how much Twitter will have to charge now. $8/mo isn't going to cut it. More like $80,000/mo
> Abrham Amare, a researcher and son of chemistry professor Meareg Amare, said that his father was targeted by a series of threatening posts because he was ethnically Tigrayan. He reported the posts, some of them shared over 50,000 times, to Facebook because they contained his father's address and called for his death.
> Eight days after the elder Amare was murdered, Facebook finally removed the post, according to the researcher.
Doesn't sound like they "actively policed" the content sufficiently.
The discussion wasn't about perfection, it was about whether this was a "frivolous lawsuit", which it obviously is not. You probably knew this, metamann, brand new account which has only commented in this thread about how Facebook's inaction lead to death.
People commit crimes inside businesses all the time, and under most circumstances the owner of a business is not liable for damages resulting from a third-party committing a crime on their property. People also use various communication tools to advocate and plan crimes, and the operators of those tools are almost never liable.
What's different here is that Facebook makes decisions about what content to show to people. It uses an algorithm to do so, and it is aware that the algorithm has a tendency to promote violent content. While I'm broadly for platform immunity, that seems like a good situation to add an exception.
Hey, please don't break the site guidelines like this, regardless of how wrong someone is or you feel they are. I understand that some important issues involve strong feelings, but we can't have discussion degenerating to this level.
You've unfortunately been breaking the guidelines in other places to, e.g. https://news.ycombinator.com/item?id=33945042. We have to ban accounts that keep doing that, so please don't!
How that was handled. Had they seen human review and wrongly rejected? Was it thrown up the chain and a VP said “fuck then skinnies!”? Was it just lost in a message queue?
All in all, this reminds me a bit of the old joke “The mail man is my drug dealer, and he doesn’t even know it.”
Well, if I mail a letter to the Post Master telling him my neighbor buys fentanyl via the post and he doesn’t read it in time, I can’t blame him for not getting to it… barring dereliction.
So, no, blanket hating FB here is just smooth-brain outrage culture.
sharing the flight plan of an airplane, apparently.
the reality is, these people truly do understand how posts impact real world safety. they fully understand and are outraged if the posts are directed in anyway towards their group—they just don’t care, or maybe worse.
If they use an algorithm to promote posts to increase their bottom-line instead of some neutral and transparent way (such as reverse-chronological feed of what people explicitly followed) they effectively act as an editor and should lose whatever Section 230-equivalent protection they are currently using.
Facebook (and other "growth & engagement" platforms) currently want to have their cake and eat it too. They act as an editor, selectively promoting certain content that increases engagement and thus their bottom-line while not taking any of the liability. Newspapers can do this too (and to a certain extent, do) but are under strict regulation which limits the damage (yes, tabloids are a thing, but none of them would publish outright calls to violence and promote illegal activities, something very commonplace on Facebook because outrageous content generates more engagement).
No, Facebook’s role was considerably more active than that. The messages that were amplified, the complaints that were ignored and they allowed the organisation of a genocide.
The NYT article covers it quite well but there is no shortage of information on their abysmal performance in the atrocities.
Please explain it to me as if I was a 5 year old. I do not understand the similarities between a communication platform and an ad-supported social network that uses software to maximize user engagement.
From my naive standpoint I feel like when you manipulate people’s attention you should be somewhat liable for what happens as a result of that.
That’s too simplistic in my opinion. Social networks have publishing aspects and Facebook is responsible for what appears and gets amplified on their properties.
What happens in real life when a city gets plastered with posters containing calls for violence?
I think the more you give people reach the higher is your responsibility to moderate the content.
You're being unfairly downvoted because it's perceived as trendy to hate facebook and pretend they can magically police all human activity. But you're quite right. The courts will find this too and have done in the past multiple times. This is just dumb clickbait and people are biting.
In this case they should have demonstrated better security around their place. A person was singled out and PII was published. It was reported and they took down the information only 8 days after the person’s death.
They clearly need more moderators for what they do and the way they play with people’s attention.
These things always come down to people not being able to define "should".
Would it be a good thing if Facebook had more moderation? Probably.
Are they morally or legally liable because a bunch of third parties committed a crime off of the platform? No, not at all.
You might disagree, but then you have to explain why facebook in particular are liable. Why aren't the actual murderers? Or the people who posted the information? Why aren't the electricity suppliers who enabled people to get online? Why isn't the government and local police?
Would it be nice if facebook solved all social problems? Yes, it would be lovely. But it's not actually their job or problem and more than it is your job. Just because they're big and you're (understandably) angry doesn't make them guilty.
Exactly. People want a special carve-out for Facebook and don't want to apply the same logic to public places and text apps. The magical "algorithm" is supposedly at blame, simply because there is supposedly an "algorithm" somewhere inside Facebook that sinisterly pulls the strings, whatever string, in order to make money.
> Abrham Amare, a researcher and son of chemistry professor Meareg Amare, said that his father was targeted by a series of threatening posts because he was ethnically Tigrayan. He reported the posts, some of them shared over 50,000 times, to Facebook because they contained his father's address and called for his death. Eight days after the elder Amare was murdered, Facebook finally removed the post, according to the researcher.
Facebook caused the death of a man. If you think that Facebook did just because the posts were "popular". then Facebook caused the death of a man for money, and that makes it worse.
I have a friend and an ex-coworker who work at Meta. I've asked them about things like this in the past, their responses have boiled down to something like this:
If Facebook's algorithms reduced millions of micro-aggressions but simultaneously enabled the death of one person, then they're probably still in the green. It's like public policy-making. There are tradeoffs everywhere, and somebody will always draw the short straw.
Did facebook _cause_ it? Or would it have happened anyway?
Lynchings have happened frequently throughout human history, long before the invention of facebook. Racism and ethnic hatred have always been there.
To me this is like saying 'well the guy who killed him was wearing nike shoes, nike caused his death'. It's pretty clear to me that if he didn't have nike shoes, he would have done the same things wearing any other pair of shoes.
Facebook allowed them to disseminate information leading to his murder. You can theorize all day about whether you think it would have happened anyways, but that doesn't really add any facts to the case.
The concept of a controlled experiment is quite important to scientific discovery and obtaining new knowledge.
What you just said is pretty equivalent to saying: When the moon is full, there is more murder. Therefore full moons must be contributing to violence. No need to theorize what happened when there's not a full moon, that's not adding any facts to the case!
> This type of thing would not have happened in the US
That's because the real police would intervene, not because murderers are afraid of social media moderators.
That said, I do believe that if Ethiopian law enforcement can't (won't?) handle their murderous mobs, they should censor their social media until they can figure out what is going on. And foreign social media companies should cooperate with them.
Your saying that I’m missing an obvious point about the 50k shares, and I think your missing the even more obvious backdrop that this took place amidst a civil war!
Is this the only guy that died in this civil war? Did fb cause the civil war?
> His fucking address and call for murder was shared 50,000 times.
That still doesn't mean that the medium for sharing caused the murder or is responsible for it. This is one of cases where intent and/or explicit action is required for both moral and legal culpablity - simply doing nothing may be an asshole thing to do, but it doesn't make you an accomplice in a crime, and if you do try assisting by moderating but are bad at it and fail sometimes, that's definitely not being an accomplice, because there is no duty to be perfect at that; a failure in moderation policy is just that - a failure in moderation policy, not becoming co-guilty of the too-slowly-moderated thing unless that post was kept up by someone with the explicit intent to facilitate that murder. I'd say that every single one of the 50,000 sharers is each more culpable than Meta.
Like, we have serious precedent for propagandists considered liable for inciting murderous hatred like https://en.wikipedia.org/wiki/Ferdinand_Nahimana and https://en.wikipedia.org/wiki/Jean-Bosco_Barayagwiza , but the bar there is met because they actively and intentionally made the propaganda; on the other hand, all the many people who (for example) kept the radios running and/or could have somehow stopped the propagandists but didn't - well, they are not liable for the crime.
So a US based company, operating under US law, is responsible to use its US based platform and employees to decide how users in a totally different country use the internet?
If you're gonna support the British Empire, fine, be my guest, but be aware that this is pretty close to noblest oblige views of colonialim
When Facebook operates in Ethiopia, it is expected to follow Ethiopian law.
> If you're gonna support the British Empire [...] this is pretty close to noblest oblige views of colonialim
The British Empire has forced subjugated countries to recognise a concept of extraterritoriality for British merchants operating in those countries, where they would not be subjected to local laws. But the US has not made such an agreement with Ethiopia about Facebook.
Facebook is being sued for allowing its platform to be used to incite and facilitate murders in Ethiopia. If they want to do the same in another country where it is legal, Ethiopian law does not forbid them.
Similarly, EU laws that restrict the sale of firearms do not apply to Facebook Marketplace in the US, only to the EU. If they allow firearms to be sold on Facebook Marketplace in the EU, they will face prosecution, even though they are based in the US.
Facebook can decide to ignore local laws, in the Middle East or anywhere else, whether they apply locally or globally, but if they ignore laws in a jurisdiction where they have assets or revenue, then legal judgements against them can be enforced. That is what is happening here.
Ethiopia is currently involved in a civil war between two ethnic groups, with all of the base brutality that entails.
You're not addressing the point, which is what right do American Executives from California have to decide on the rights of Ethiopians to speak.
Additionally, even if such postings were illegal, this assumes that the government was not currently on one side of a civil war.
You're expecting a level of nuance and decision making about events that are so far out of a social media company's understanding and frankly, ability that I'm pretty sure you're just excited someone is suing Facebook. From FBs perspective, there aren't a lot of principled stances and even if they do take one, there will still be a million hackernews comments about how FB is responsible for bad thing.
The Rawandan genocide was driven forward in large part by disk jockeys on broadcast radio who used their platform to direct the mob toward their targets.
Violence like that described in the article often does not happen organically without mass media of some kind inspiring it, contrary to what you have asserted.
Ethnic violence requires people to identify someone as "other" than themselves, and requires people to take the massive leap towards using that otherness to justify murder. These ideas need to be organized by a rational and malevolent mind, and they need to be widely disseminated before they encounter people who are motivated to act upon them. The people who come up with these ideas are usually not unhinged enough to carry out these atrocious acts, and the people who commit these murders are usually not able to come up with these ideas on their own. It's the mass media that links them together.
Violent ideas disseminated via Facebook can and do have the same effect that violent ideas disseminated via radio and newspapers have had in the past. And it should be assessed through the same lens.
Wouldn't you say that the radio station owners in Rawanda shared responsibility with the disk jockeys they put on the air? Wouldn't you say that the newspaper publishers in the Jim Crow South shared responsibility with the writers that they published? Of course you would.
Facebook was told that this call for violence before the man was murdered, and they did not act.
Yes, there is a clear difference between Facebook and these other examples in that Facebook does not share the politics of the murderers. But Facebook provided the platform, Facebook made money when the call for violence was disseminated, and the man died because Facebook failed to act in a timely manner to remove the information.
Different perhaps in intent, but not in effect, compared to these earlier examples. If they are not held financially liable for this, will they have sufficient incentive to act differently in the future?
> Violence like that described in the article often does not happen organically without mass media of some kind inspiring it, contrary to what you have asserted.
Dude what?
So what if we go back even further?
Spanish Inquisition? Crusades? Genghis khan slaying anyone in his path? Colonialism?
What about the thousands of small and undocumented tribal wars fought before people even had written language?
It’s as if you are saying racism and tribalism aren’t in basic levels of the human psyche and that it’s really just 1 or 2 bad ppl with a megaphone. Dangerous stance imo
>> Violence like that described in the article often does not happen organically
> So what if we go back even further?
> Spanish Inquisition? Crusades? Genghis khan slaying anyone in his path? Colonialism?
I did not say that mass media is always is the only driver of violence. What's important is that in the modern era, mass media has often driven violence, as it did in this case.
Your original comment made the assertion that Facebook did not cause this outcome, because the murder would have happened anyway, whether or not Facebook was involved.
My point was that there are many examples where mass media has inspired ethnic violence that wouldn't have happened anyway without the mass media's involvement—so it shouldn't be surprising that in this case these Facebook posts inspired violence that wouldn't have happened anyway.
Are you really asserting that the plaintiff's father would have been killed even if the Facebook post in question hadn't been made? Or if Facebook hadn't existed, somehow the same people would have been inspired to kill the same man?
>Or if Facebook hadn't existed, somehow the same people would have been inspired to kill the same man?
Yes absolutely I am directly saying this and imo your points support my argument. When there is no Facebook, ppl use the radio to incite violence. When there is no radio, ppl use newspapers to incite and organize violence. And when there is no newspapers, ppl find a way to get their hate killing done anyway. There’s thousands of years of examples.
I don’t know what definition you are using for the word “often” but it seems like it’s not the usual definition of being frequency based. This is because, again, thousands of years of human ethnic violence happened where humans didn’t not have the capability of broadcast media. If we just count up the number of data points, it doesn’t make sense to apply the word “often” here at all.
I hate meta as much as the next person, but I find this whole line of reasoning farsical and a distraction from the many actually bad things to worry about.
I think there's a huge difference between "inciting hatred" and "not deplatforming racist stuff quickly enough in one case." It's my understanding that meta actively tries to prevent hateful content in over 500 languages.
It's a cool experiment to try to deplatform racist tendencies, but I don't see how you're accountable for failing to catch 100% of cases. I imagine if facebook didn't exist at all (and people were gathering in churches or in the street to organize their racism) we'd see a lot lot less of it being caught and likely more deaths.
Yes, they can. Ordering by time is still a "feed" - a simple one, but it's a feed nonetheless. Many modern phones surface messages from frequent contacts up to the top, above more recent messages from low-frequency contacts. That's also a "feed". Furthermore, most of the calls to violence were made over Facebook messenger which is arguably a lot more like SMS and IRC.
I find the argument that Facebook's feed was significant to hold little water. People will coordinate with whatever communication mechanisms they have access to. In the past this was SMS, and even further into the past this was radio. Do we forget the Balkans, Rwanda, and plenty of other atrocities that predate widespread adoption of social media.
Problem is they are not only giving a platform to racism but also amplifying it because that is what drives engagement and more engagement is necessary to show more ads and that’s good for the bottom line and that makes it possible to pay $2,000,000,000. Facebook simply needs to vanish into thin air.
Facebook doesn't care what becomes amplified, the other racist users on the platform are amplifying hateful posts by engaging with them. That's not the same as intentionally promoting it. Facebook vanishing would make zero difference in social media being used to promote racisms, Parler demonstrated that people will seek alternatives when deplatformed.
Unless there is news about Facebook employees intentionally changing their algorithm to promote violence directly, none of these arguments feel genuine. We're past the point where we can compare the world with and without Faceboook, when a social media site becomes unavailable to a certain group they will go to an alternative, as seen with Parler.
Facebook mining and profiting from harvesting political user data shows a direct conflict of interest between keep data private and making money. When Facebook is used as a platform by hate groups, I can't see how they would meaningfully profit from it. Saying that encouraging or ignoring it would increase userbase is very short sighted, at the bare minimum it would caused the group being targeted to leave.
These situations feel more like the debate of whether ISPs are a public utility and the EARN IT act and section 230. Holding Facebook 100% accountable would set a very dangerous precedent. Facebook is not an entity and without showing hateful posts/comments were highlighted while alternatives were not, this feels like an argument against math.
Popular posts reflect the interests of the vocal population. Assigning morality to the math used in an algorithm is a failure to understand the purpose or function of it. "Inciting" would means the algorithm promoted a hateful post despite the underlying signals saying it shouldn't be promoted.
Social media's aims are unlike any other social space to come before.
Restaurants gather people together around delicious food and everyone is able to enjoy themselves at the table with a meal and friends. Friendly engagement.
Parks invite people to have recreation at various stages in life, whether with team sports, a playground, just running around the grass, or a picnic. Friendly engagement.
Public transit helps people congregate on vehicles for the sake of traveling to a common destination, and therefore disparate people are united by this common goal of arriving safe and sound.
I could go on about all the public spaces that promote some kind of friendly engagement and socializing, but social media such as Facebook is calibrated and designed to put us at one another's throats. It's not good if we see a post and let it go by, we must lash out and correct or attack. Dislikes are as valuable as Likes in the Facebook economy. Hordes of censors are required to tamp down the inevitable violations that occur when people go out of bounds. Social media thrives and grows from attacks, libel, gossip, and falsehoods.
Of course this is all only an extrapolation of print and broadcast media as it came before. Nothing has changed since Moses broke the tablets and Luther nailed up his theses. But social media is concentrated and unadulterated poison.
I have not experienced the toxic negativity you have associated with only modern social media. I am in a couple of online groups of the things you are describing as social spaces of before, there's a local hiking group, an unofficial puzzled pints group, multiple AI and Defcon related groups, TV show discussion groups, and occasionally concert groups for when an artist I enjoy is playing near me. If you're not using Facebook or Discord or some other social media for this, I would suggest you look into it. Social media is focused on self curation as much as algorithmic curation. If you have to approach your social media feed like a warzone, I would start by muting or unfollowing as many problematic people as possible. It'll be up to you to make sure your feed doesn't become an echo chamber, but at a minimum you won't feel such negative anger towards a literal piece of code.
Like and Dislikes are only as valuable as users allow them to be. A great counterexample to many of your arguments is the social media site we are discussing this topic on. The culture on Hacker News de-emphasizes Likes and Dislike. Comment karma is visible to the poster, but to a reader, a +100 karma comment is equal to a +4 karma comment with the only difference being what is seen first. I personally forget about that and end up reading comments to the end of threads I enjoy anyway.
I think letting such hostility build for the concept of social media is as damaging as any political extremism the censorship is supposed to combat. It feels like the extremism isn't gone, just redirected. Social media and the internet has changed who people interact with, not just how. In the past extreme ideas existed, but were tempered because of others around them. Social media has allowed people to choose to only listen to people with the same perspective, letting extreme ideas go uncontrolled. You can argue that it's the fault of the technology, but the main responsibility falls on the individuals.
Social media doesn't make people hateful, it makes it easier for hateful people to convince other to think like them. Bullies have existed long before social media and continue to exist in every context.
> Assigning morality to the math used in an algorithm is a failure to understand the purpose or function of it.
No this is rhetorical judo; no one intends to ascribe moral agency to the algorithm itself. The function of an algorithm is what it does, and the moral culpability is carried by the people who designed it for, and applied it to, a given purpose causing that outcome.
We can decide that for example they could not have predicted this outcome and so are not liable, or that they should have and were neglectful, or something else. The full range of tools and consequences we use when assessing culpability in other damaging situations are available to us here. The fact that an algorithm was the tool of harm doesn't constrain us.
Whether or not they profit from it is also beside the point. It may make it more evil, it may change their civil liabilities to harmed parties, but it doesn't increase or decrease their responsibility for the outcomes themselves.
> We can decide that for example they could not have predicted this outcome and so are not liable, or that they should have and were neglectful, or something else.
We fundamentally disagree on whether or not what Facebook "should" have expected or "should" have done is worth considering. I don't think it's legally relevant. Profits can demonstrate incentive that otherwise wouldn't exist.
The existing legal precedent for how algorithms are considered under tort liability [1] are still unclear. It's clear to me that allowing this lawsuit to move forward would be directly contradicting the ruling of Force v. Facebook (2019) and would force a ruling on how AI algorithms should be held liable with respect to section 230.
From the perspective of the user, making a platform liable would mean that the platform would fundamentally shift what users are allowed to post and strongly weakens basic rights to freedom of expression. It's hard to depict how different and more censored the internet would be if section 230 continues to become weakened.
"Critically, the finding that AI algorithms can be found to be harmful allows for the tort liability process to come into effect. While this kind of legal approach has not yet been taken, Patel indicates that corporate liability is a possibility. If an AI algorithm causes harm, it must be prosecuted through tort liability."
"However, in the same timeframe as Patel’s ruling in 2019, ... a group of U.S. citizens who were victims of a Hamas terrorist attack, sued Facebook, arguing that Facebook implicitly assisted Hamas by keeping their violence-promoting content online through their AI algorithm’s oversight. [12] However, in accordance with 42 U.S.C. 230(c)(1), the rule that protects social media platforms from liability for content on their platforms, the court found that Facebook could not be found to have criminal or civil liability for content. [13] This case complicates the implications of Patel, as it found that Facebook was shielded from liability because the content itself was not created by the platform. "
This is another contortion. Holding an entity accountable for the results of deploying an algorithm is not the same as holding them liable for content posted by other people on their platform. The algorithm in this case governs how and to who that content is shown, but they are still two separate things. A court could definitely find that for example there's no liability for posting certain content in general, but there is for posting it like that.
Court isn't a context-free zone, they absolutely may consider what an entity did know or should have known about the future consequences of their actions. As in the different categories of negligence, which this is similar to.
Are you a lawyer why are you using their jargon so heavily? The lawyers will do that part, in the courts. Here, in this context, it's just another rhetorical flourish to distract attention away from your actual position. No one cares what a tort is just say your piece in normal words and stop trying to obscure it behind the facade of technical jargon.
My position is that Facebook should be protected under section 230. Section 230 says no online content hosting platform will be liable for what their users post as long as they attempt content moderation in good faith. My understanding of this lawsuit is, Facebook is being sued for 2 Billion USD because it was enabling hateful speech and targeted posts that lead to the death of man. Facebook is allegedly directly liable because of poor moderation.
The distinction you are making doesn't make sense. Are you making an analogy to the idea that you can use the word fire but you can't yell fire in a crowded place? Facebook doesn't choose individually what posts are highlighted, the engagement of other users does. The complete weights of what is highlighted are also unique per user. You're arguing that Facebook has a duty of care, but as it stands section 230 already addresses that. The duty of care is fulfilled with good faith moderation. The only way Facebook can be found liable is if section 230 is weakened and good faith moderation is redefined.
"Tort liability indicates that someone is held accountable for wrong actions (other than under contract." Tort liability is the phrase you would search if you wanted to learn about previous legal rulings related to this. I thought the Columbia Law Review article I linked was pretty approachable. It's not possible to talk about this in a meaningful way without understanding tort. I haven't tried to obscure or distract from my position intentionally.
I think Force v. Facebook was the proper ruling. There's nothing new courts have to decide, because section 230 already addresses this. Fundamentally, people are hateful and will engage and interact with hateful content. It doesn't make sense to blame an algorithm for not realizing a post is hateful and highlighting it. I made my position clear without technical jargon in the top comment. The DW article says the lawsuit wants to legally force Facebook to change its algorithm. The change would directly be related to moderation. This would be a slippery slope. Failing to understand the limitation of what an algorithm does is the only way you could expect a separation of the concept of posting content vs posting it "like that". I hope I was able to clarify my position.
> Facebook doesn't choose individually what posts are highlighted, the engagement of other users does.
They do, by having control of the algorithm. It didn't just spring forth, they made it. More relevant, they have the off switch. "Good faith moderation" means monitoring the actual results of applying it and making changes if the results are harmful.
> It doesn't make sense to blame an algorithm for not realizing a post is hateful and highlighting it.
I'm not blaming the algorithm, I'm blaming the people who implemented it, failed to understand the potential consequences of using it in this way, and then failed to notice or understand what was happening.
You can't just abdicate moral responsibility by putting "the algorithm" between your actions and their consequences. A for loop and a couple of if statements is an algorithm. Someone makes it, gives it input, applies the output to some task. That is the culpable action, not whatever happens inside the algorithm.
A machine designed to self-optimize to manufacture as many paperclips as possible with no regard to any other parameters is not capable of morality. A person who creates one without taking steps to ensure it does not eat the world to turn the raw materials into paperclips is capable of morality and is morally responsible for that outcome.
Facebook has created a machine to capture attention and increase engagement with little regard to any other parameters. Despite being shown evidence that its machine contributes to political extremism and violence, it has typically refused to make changes that would reduce those effects if those changes would also reduce engagement.
It's fair to claim innocent intent early on, but someone who is aware their behavior has harmful impacts and continues that behavior has, at best negligent intent.
Yes. And this is an unprecedented industrial process, not least because attention is the input, and behavioral changes (well-tracked viewing and purchasing) are the output.
I like this take on social media as the industrialization of behavior modification. Framing it that way effectively communicates the hazards to the layperson who thinks its purpose is to help people stay in touch.
So a perfect free-speech forum where there is no algorithm would be okay? Or is the lack of algorithm also an algorithm? Is there any way to win this? Or is every website now doomed to an arbitrary level of required content moderation?
Large crowds in some kinds of shapes of open spaces randomly result in stampedes and people dying.
All shapes of spaces are not subject to this.
There are general designs that don't have this problem and there are people that have various practical guidelines to counteract the stampede problem.
Same with forums where people can say things.
For example a simple algorithm that does not suffer from as gross feedback effects, that facebook eliminated early on, was the timeline view of activity.
Timeline sorting did not use popularity or advertising promotion to sort items.
Would the service be usable like this, well yes, facebook was perfectly usable during the era it worked like this.
What is the hit to their bottom line if they changed their algorithm back to simple timeline sorting? Well we can guess it must be pretty big because their current algorithm seems to be pretty psychotic by comparison (in that it intentionally fans the flames of outrage with it's interaction seeking behavior) and they are willing to take all the negative PR for the sake of this interaction uplift.
Facebook engineers don't need to explicitly change their algorithm to promote violence. They do it by negligence; by letting the algo maximize for engagememt regardless of the outcomes. Facebook spends most of it's moderation efforts in just a few countries and does relatively little for the rest.
It is not reasonable to moderate to a certain level in a few locations, thereby implicitly acknowledging such moderation is important, while then letting other "gardens" in other parts of the world run wild and untended.
Once the moderation has begun and leaders in the company have acknowledged that it is necessary for the good of the platform and the communities in which it operates, it ceases to become reasonable to say "oh well, what we meant was, moderation is needed where the profile is highest or will cause the most political/societal blow-back if we don't".
Radio station broadcasting played a big part in initiating the Rwanda genocide
However innocuous 'algorithmically promoting popular posts' is, we should not be blind, legally or otherwise to the consequences
I suppose it is hard; letting a dance video go viral, is not dissimilar to enabling any other message go viral. We are eager for it; but uncheck virality typically kills the host
People keep saying "the algorithm did this" but where's the proof? Or do we count "showing the latest posts in chronological order" an algorithm now? Sheesh.
There's still no proof that the algorithm contributed to this. Maybe chronological order would result in MORE hate online. This is the argument any good lawyer would make, so you have to defeat that simple before you can move on.
There’s a difference between simple and simplistic. This 5-year old might be mislead into thinking that there’s no difference between a phone and a social network.
A phone call is low-bandwidth and ephemeral. The time it takes to disseminate information that way to a larger group is huge. Societal self-regulation has a chance to kick in.
On the other hand you have high-bandwidth posts that once published can be upvoted by a single click and spread like wildfire.
96 comments
[ 3.9 ms ] story [ 169 ms ] threadkudos to the promoters
It was raised with them, they didn’t act and someone died. They have no responsibility for spreading death threats?
If that content is ok, what content is not ok?
If they don't have enough resources to review the outputs from their algorithm, then to bad, just show posts from friends only.
> Eight days after the elder Amare was murdered, Facebook finally removed the post, according to the researcher.
Doesn't sound like they "actively policed" the content sufficiently.
If an ethnic conflict occurs in a relatively obscure region are they going to spin up new moderators on the AWS spot market?
These things aren’t perfect or fast.
See that's the problem with omniscience, you kind of need it to police EVERYTHING. Something will always make it through.
The same is true for venues, sports, shops, workplaces, etc... why not software companies?
People commit crimes inside businesses all the time, and under most circumstances the owner of a business is not liable for damages resulting from a third-party committing a crime on their property. People also use various communication tools to advocate and plan crimes, and the operators of those tools are almost never liable.
What's different here is that Facebook makes decisions about what content to show to people. It uses an algorithm to do so, and it is aware that the algorithm has a tendency to promote violent content. While I'm broadly for platform immunity, that seems like a good situation to add an exception.
You've unfortunately been breaking the guidelines in other places to, e.g. https://news.ycombinator.com/item?id=33945042. We have to ban accounts that keep doing that, so please don't!
If you wouldn't mind reviewing https://news.ycombinator.com/newsguidelines.html and taking the intended spirit of the site more to heart, we'd be grateful.
How that was handled. Had they seen human review and wrongly rejected? Was it thrown up the chain and a VP said “fuck then skinnies!”? Was it just lost in a message queue?
All in all, this reminds me a bit of the old joke “The mail man is my drug dealer, and he doesn’t even know it.”
Well, if I mail a letter to the Post Master telling him my neighbor buys fentanyl via the post and he doesn’t read it in time, I can’t blame him for not getting to it… barring dereliction.
So, no, blanket hating FB here is just smooth-brain outrage culture.
sharing the flight plan of an airplane, apparently.
the reality is, these people truly do understand how posts impact real world safety. they fully understand and are outraged if the posts are directed in anyway towards their group—they just don’t care, or maybe worse.
Facebook (and other "growth & engagement" platforms) currently want to have their cake and eat it too. They act as an editor, selectively promoting certain content that increases engagement and thus their bottom-line while not taking any of the liability. Newspapers can do this too (and to a certain extent, do) but are under strict regulation which limits the damage (yes, tabloids are a thing, but none of them would publish outright calls to violence and promote illegal activities, something very commonplace on Facebook because outrageous content generates more engagement).
https://www.nytimes.com/2018/10/15/technology/myanmar-facebo...
I understand that HN has a massive hate boner for Facebook, but don't let it cloud your judgment.
The NYT article covers it quite well but there is no shortage of information on their abysmal performance in the atrocities.
And Apple participated in that. Also the internet providers.
> the complaints that were ignored
I am sure Apple gets a lot of complaints about Facebook
> and they allowed the organisation of a genocide
Still fits Apple
From my naive standpoint I feel like when you manipulate people’s attention you should be somewhat liable for what happens as a result of that.
Every social network is a communication platform.
What happens in real life when a city gets plastered with posters containing calls for violence?
I think the more you give people reach the higher is your responsibility to moderate the content.
You got an adult explanation above. You asked for ELI5.
They clearly need more moderators for what they do and the way they play with people’s attention.
Would it be a good thing if Facebook had more moderation? Probably.
Are they morally or legally liable because a bunch of third parties committed a crime off of the platform? No, not at all.
You might disagree, but then you have to explain why facebook in particular are liable. Why aren't the actual murderers? Or the people who posted the information? Why aren't the electricity suppliers who enabled people to get online? Why isn't the government and local police?
Would it be nice if facebook solved all social problems? Yes, it would be lovely. But it's not actually their job or problem and more than it is your job. Just because they're big and you're (understandably) angry doesn't make them guilty.
Facebook caused the death of a man. If you think that Facebook did just because the posts were "popular". then Facebook caused the death of a man for money, and that makes it worse.
If Facebook's algorithms reduced millions of micro-aggressions but simultaneously enabled the death of one person, then they're probably still in the green. It's like public policy-making. There are tradeoffs everywhere, and somebody will always draw the short straw.
There's a reason they're called microaggressions.
I think the parent has a point. Not sure I agree, but there are many ways to argue for or against.
One is impossible; the other should have bounced off immediately.
Lynchings have happened frequently throughout human history, long before the invention of facebook. Racism and ethnic hatred have always been there.
To me this is like saying 'well the guy who killed him was wearing nike shoes, nike caused his death'. It's pretty clear to me that if he didn't have nike shoes, he would have done the same things wearing any other pair of shoes.
Hair splitting.
"Guns don't kill people. Bullets do!"
What you just said is pretty equivalent to saying: When the moon is full, there is more murder. Therefore full moons must be contributing to violence. No need to theorize what happened when there's not a full moon, that's not adding any facts to the case!
And here you are, parroting the idea of controlled experiment and wanting to discuss scientific discovery.
This is someone's life we are talking about. Clearly it is a failure in moderation policy. This type of thing would not have happened in the US.
That's because the real police would intervene, not because murderers are afraid of social media moderators.
That said, I do believe that if Ethiopian law enforcement can't (won't?) handle their murderous mobs, they should censor their social media until they can figure out what is going on. And foreign social media companies should cooperate with them.
But the initiative has to come from Ethiopia.
Is this the only guy that died in this civil war? Did fb cause the civil war?
That still doesn't mean that the medium for sharing caused the murder or is responsible for it. This is one of cases where intent and/or explicit action is required for both moral and legal culpablity - simply doing nothing may be an asshole thing to do, but it doesn't make you an accomplice in a crime, and if you do try assisting by moderating but are bad at it and fail sometimes, that's definitely not being an accomplice, because there is no duty to be perfect at that; a failure in moderation policy is just that - a failure in moderation policy, not becoming co-guilty of the too-slowly-moderated thing unless that post was kept up by someone with the explicit intent to facilitate that murder. I'd say that every single one of the 50,000 sharers is each more culpable than Meta.
Like, we have serious precedent for propagandists considered liable for inciting murderous hatred like https://en.wikipedia.org/wiki/Ferdinand_Nahimana and https://en.wikipedia.org/wiki/Jean-Bosco_Barayagwiza , but the bar there is met because they actively and intentionally made the propaganda; on the other hand, all the many people who (for example) kept the radios running and/or could have somehow stopped the propagandists but didn't - well, they are not liable for the crime.
If you're gonna support the British Empire, fine, be my guest, but be aware that this is pretty close to noblest oblige views of colonialim
When Facebook operates in Ethiopia, it is expected to follow Ethiopian law.
> If you're gonna support the British Empire [...] this is pretty close to noblest oblige views of colonialim
The British Empire has forced subjugated countries to recognise a concept of extraterritoriality for British merchants operating in those countries, where they would not be subjected to local laws. But the US has not made such an agreement with Ethiopia about Facebook.
https://en.wikipedia.org/wiki/Extraterritoriality
Facebook can, and should, tell those countries to fuck off and build their own Great Firewall if they want to.
Similarly, EU laws that restrict the sale of firearms do not apply to Facebook Marketplace in the US, only to the EU. If they allow firearms to be sold on Facebook Marketplace in the EU, they will face prosecution, even though they are based in the US.
Facebook can decide to ignore local laws, in the Middle East or anywhere else, whether they apply locally or globally, but if they ignore laws in a jurisdiction where they have assets or revenue, then legal judgements against them can be enforced. That is what is happening here.
You're not addressing the point, which is what right do American Executives from California have to decide on the rights of Ethiopians to speak.
Additionally, even if such postings were illegal, this assumes that the government was not currently on one side of a civil war.
You're expecting a level of nuance and decision making about events that are so far out of a social media company's understanding and frankly, ability that I'm pretty sure you're just excited someone is suing Facebook. From FBs perspective, there aren't a lot of principled stances and even if they do take one, there will still be a million hackernews comments about how FB is responsible for bad thing.
I know I'd feel _totally_ OK with that if I worked at Meta. /s
https://en.wikipedia.org/wiki/Radio_T%C3%A9l%C3%A9vision_Lib...
Even lynchings in the Jim Crow South were often driven by the mass media of the time, newspapers.
https://www.nbcuacademy.com/catalog/newspaper-archives-print...
Violence like that described in the article often does not happen organically without mass media of some kind inspiring it, contrary to what you have asserted.
Ethnic violence requires people to identify someone as "other" than themselves, and requires people to take the massive leap towards using that otherness to justify murder. These ideas need to be organized by a rational and malevolent mind, and they need to be widely disseminated before they encounter people who are motivated to act upon them. The people who come up with these ideas are usually not unhinged enough to carry out these atrocious acts, and the people who commit these murders are usually not able to come up with these ideas on their own. It's the mass media that links them together.
Violent ideas disseminated via Facebook can and do have the same effect that violent ideas disseminated via radio and newspapers have had in the past. And it should be assessed through the same lens.
Wouldn't you say that the radio station owners in Rawanda shared responsibility with the disk jockeys they put on the air? Wouldn't you say that the newspaper publishers in the Jim Crow South shared responsibility with the writers that they published? Of course you would.
Facebook was told that this call for violence before the man was murdered, and they did not act.
Yes, there is a clear difference between Facebook and these other examples in that Facebook does not share the politics of the murderers. But Facebook provided the platform, Facebook made money when the call for violence was disseminated, and the man died because Facebook failed to act in a timely manner to remove the information.
Different perhaps in intent, but not in effect, compared to these earlier examples. If they are not held financially liable for this, will they have sufficient incentive to act differently in the future?
Dude what?
So what if we go back even further?
Spanish Inquisition? Crusades? Genghis khan slaying anyone in his path? Colonialism?
What about the thousands of small and undocumented tribal wars fought before people even had written language?
It’s as if you are saying racism and tribalism aren’t in basic levels of the human psyche and that it’s really just 1 or 2 bad ppl with a megaphone. Dangerous stance imo
> So what if we go back even further?
> Spanish Inquisition? Crusades? Genghis khan slaying anyone in his path? Colonialism?
I did not say that mass media is always is the only driver of violence. What's important is that in the modern era, mass media has often driven violence, as it did in this case.
Your original comment made the assertion that Facebook did not cause this outcome, because the murder would have happened anyway, whether or not Facebook was involved.
My point was that there are many examples where mass media has inspired ethnic violence that wouldn't have happened anyway without the mass media's involvement—so it shouldn't be surprising that in this case these Facebook posts inspired violence that wouldn't have happened anyway.
Are you really asserting that the plaintiff's father would have been killed even if the Facebook post in question hadn't been made? Or if Facebook hadn't existed, somehow the same people would have been inspired to kill the same man?
Yes absolutely I am directly saying this and imo your points support my argument. When there is no Facebook, ppl use the radio to incite violence. When there is no radio, ppl use newspapers to incite and organize violence. And when there is no newspapers, ppl find a way to get their hate killing done anyway. There’s thousands of years of examples.
I don’t know what definition you are using for the word “often” but it seems like it’s not the usual definition of being frequency based. This is because, again, thousands of years of human ethnic violence happened where humans didn’t not have the capability of broadcast media. If we just count up the number of data points, it doesn’t make sense to apply the word “often” here at all.
I think there's a huge difference between "inciting hatred" and "not deplatforming racist stuff quickly enough in one case." It's my understanding that meta actively tries to prevent hateful content in over 500 languages.
It's a cool experiment to try to deplatform racist tendencies, but I don't see how you're accountable for failing to catch 100% of cases. I imagine if facebook didn't exist at all (and people were gathering in churches or in the street to organize their racism) we'd see a lot lot less of it being caught and likely more deaths.
https://www.amnesty.org/en/latest/news/2022/09/myanmar-faceb...
I find the argument that Facebook's feed was significant to hold little water. People will coordinate with whatever communication mechanisms they have access to. In the past this was SMS, and even further into the past this was radio. Do we forget the Balkans, Rwanda, and plenty of other atrocities that predate widespread adoption of social media.
Facebook mining and profiting from harvesting political user data shows a direct conflict of interest between keep data private and making money. When Facebook is used as a platform by hate groups, I can't see how they would meaningfully profit from it. Saying that encouraging or ignoring it would increase userbase is very short sighted, at the bare minimum it would caused the group being targeted to leave.
These situations feel more like the debate of whether ISPs are a public utility and the EARN IT act and section 230. Holding Facebook 100% accountable would set a very dangerous precedent. Facebook is not an entity and without showing hateful posts/comments were highlighted while alternatives were not, this feels like an argument against math.
Popular posts reflect the interests of the vocal population. Assigning morality to the math used in an algorithm is a failure to understand the purpose or function of it. "Inciting" would means the algorithm promoted a hateful post despite the underlying signals saying it shouldn't be promoted.
https://www.culawreview.org/journal/extending-corporate-tort...
Restaurants gather people together around delicious food and everyone is able to enjoy themselves at the table with a meal and friends. Friendly engagement.
Parks invite people to have recreation at various stages in life, whether with team sports, a playground, just running around the grass, or a picnic. Friendly engagement.
Public transit helps people congregate on vehicles for the sake of traveling to a common destination, and therefore disparate people are united by this common goal of arriving safe and sound.
I could go on about all the public spaces that promote some kind of friendly engagement and socializing, but social media such as Facebook is calibrated and designed to put us at one another's throats. It's not good if we see a post and let it go by, we must lash out and correct or attack. Dislikes are as valuable as Likes in the Facebook economy. Hordes of censors are required to tamp down the inevitable violations that occur when people go out of bounds. Social media thrives and grows from attacks, libel, gossip, and falsehoods.
Of course this is all only an extrapolation of print and broadcast media as it came before. Nothing has changed since Moses broke the tablets and Luther nailed up his theses. But social media is concentrated and unadulterated poison.
Like and Dislikes are only as valuable as users allow them to be. A great counterexample to many of your arguments is the social media site we are discussing this topic on. The culture on Hacker News de-emphasizes Likes and Dislike. Comment karma is visible to the poster, but to a reader, a +100 karma comment is equal to a +4 karma comment with the only difference being what is seen first. I personally forget about that and end up reading comments to the end of threads I enjoy anyway.
I think letting such hostility build for the concept of social media is as damaging as any political extremism the censorship is supposed to combat. It feels like the extremism isn't gone, just redirected. Social media and the internet has changed who people interact with, not just how. In the past extreme ideas existed, but were tempered because of others around them. Social media has allowed people to choose to only listen to people with the same perspective, letting extreme ideas go uncontrolled. You can argue that it's the fault of the technology, but the main responsibility falls on the individuals.
Social media doesn't make people hateful, it makes it easier for hateful people to convince other to think like them. Bullies have existed long before social media and continue to exist in every context.
No this is rhetorical judo; no one intends to ascribe moral agency to the algorithm itself. The function of an algorithm is what it does, and the moral culpability is carried by the people who designed it for, and applied it to, a given purpose causing that outcome.
We can decide that for example they could not have predicted this outcome and so are not liable, or that they should have and were neglectful, or something else. The full range of tools and consequences we use when assessing culpability in other damaging situations are available to us here. The fact that an algorithm was the tool of harm doesn't constrain us.
Whether or not they profit from it is also beside the point. It may make it more evil, it may change their civil liabilities to harmed parties, but it doesn't increase or decrease their responsibility for the outcomes themselves.
We fundamentally disagree on whether or not what Facebook "should" have expected or "should" have done is worth considering. I don't think it's legally relevant. Profits can demonstrate incentive that otherwise wouldn't exist.
The existing legal precedent for how algorithms are considered under tort liability [1] are still unclear. It's clear to me that allowing this lawsuit to move forward would be directly contradicting the ruling of Force v. Facebook (2019) and would force a ruling on how AI algorithms should be held liable with respect to section 230.
From the perspective of the user, making a platform liable would mean that the platform would fundamentally shift what users are allowed to post and strongly weakens basic rights to freedom of expression. It's hard to depict how different and more censored the internet would be if section 230 continues to become weakened.
"Critically, the finding that AI algorithms can be found to be harmful allows for the tort liability process to come into effect. While this kind of legal approach has not yet been taken, Patel indicates that corporate liability is a possibility. If an AI algorithm causes harm, it must be prosecuted through tort liability."
"However, in the same timeframe as Patel’s ruling in 2019, ... a group of U.S. citizens who were victims of a Hamas terrorist attack, sued Facebook, arguing that Facebook implicitly assisted Hamas by keeping their violence-promoting content online through their AI algorithm’s oversight. [12] However, in accordance with 42 U.S.C. 230(c)(1), the rule that protects social media platforms from liability for content on their platforms, the court found that Facebook could not be found to have criminal or civil liability for content. [13] This case complicates the implications of Patel, as it found that Facebook was shielded from liability because the content itself was not created by the platform. "
[1] https://www.culawreview.org/journal/extending-corporate-tort...
Court isn't a context-free zone, they absolutely may consider what an entity did know or should have known about the future consequences of their actions. As in the different categories of negligence, which this is similar to.
Are you a lawyer why are you using their jargon so heavily? The lawyers will do that part, in the courts. Here, in this context, it's just another rhetorical flourish to distract attention away from your actual position. No one cares what a tort is just say your piece in normal words and stop trying to obscure it behind the facade of technical jargon.
The distinction you are making doesn't make sense. Are you making an analogy to the idea that you can use the word fire but you can't yell fire in a crowded place? Facebook doesn't choose individually what posts are highlighted, the engagement of other users does. The complete weights of what is highlighted are also unique per user. You're arguing that Facebook has a duty of care, but as it stands section 230 already addresses that. The duty of care is fulfilled with good faith moderation. The only way Facebook can be found liable is if section 230 is weakened and good faith moderation is redefined.
"Tort liability indicates that someone is held accountable for wrong actions (other than under contract." Tort liability is the phrase you would search if you wanted to learn about previous legal rulings related to this. I thought the Columbia Law Review article I linked was pretty approachable. It's not possible to talk about this in a meaningful way without understanding tort. I haven't tried to obscure or distract from my position intentionally.
I think Force v. Facebook was the proper ruling. There's nothing new courts have to decide, because section 230 already addresses this. Fundamentally, people are hateful and will engage and interact with hateful content. It doesn't make sense to blame an algorithm for not realizing a post is hateful and highlighting it. I made my position clear without technical jargon in the top comment. The DW article says the lawsuit wants to legally force Facebook to change its algorithm. The change would directly be related to moderation. This would be a slippery slope. Failing to understand the limitation of what an algorithm does is the only way you could expect a separation of the concept of posting content vs posting it "like that". I hope I was able to clarify my position.
https://www.nytimes.com/2021/03/24/technology/section-230-he...
They do, by having control of the algorithm. It didn't just spring forth, they made it. More relevant, they have the off switch. "Good faith moderation" means monitoring the actual results of applying it and making changes if the results are harmful.
> It doesn't make sense to blame an algorithm for not realizing a post is hateful and highlighting it.
I'm not blaming the algorithm, I'm blaming the people who implemented it, failed to understand the potential consequences of using it in this way, and then failed to notice or understand what was happening.
You can't just abdicate moral responsibility by putting "the algorithm" between your actions and their consequences. A for loop and a couple of if statements is an algorithm. Someone makes it, gives it input, applies the output to some task. That is the culpable action, not whatever happens inside the algorithm.
Facebook has created a machine to capture attention and increase engagement with little regard to any other parameters. Despite being shown evidence that its machine contributes to political extremism and violence, it has typically refused to make changes that would reduce those effects if those changes would also reduce engagement.
It's fair to claim innocent intent early on, but someone who is aware their behavior has harmful impacts and continues that behavior has, at best negligent intent.
All shapes of spaces are not subject to this.
There are general designs that don't have this problem and there are people that have various practical guidelines to counteract the stampede problem.
Same with forums where people can say things.
For example a simple algorithm that does not suffer from as gross feedback effects, that facebook eliminated early on, was the timeline view of activity.
Timeline sorting did not use popularity or advertising promotion to sort items.
Would the service be usable like this, well yes, facebook was perfectly usable during the era it worked like this.
What is the hit to their bottom line if they changed their algorithm back to simple timeline sorting? Well we can guess it must be pretty big because their current algorithm seems to be pretty psychotic by comparison (in that it intentionally fans the flames of outrage with it's interaction seeking behavior) and they are willing to take all the negative PR for the sake of this interaction uplift.
https://www.businessinsider.com/facebook-ranks-the-world-int...
It is not reasonable to moderate to a certain level in a few locations, thereby implicitly acknowledging such moderation is important, while then letting other "gardens" in other parts of the world run wild and untended.
https://www.nytimes.com/2018/10/15/technology/myanmar-facebo...
Once the moderation has begun and leaders in the company have acknowledged that it is necessary for the good of the platform and the communities in which it operates, it ceases to become reasonable to say "oh well, what we meant was, moderation is needed where the profile is highest or will cause the most political/societal blow-back if we don't".
However innocuous 'algorithmically promoting popular posts' is, we should not be blind, legally or otherwise to the consequences
I suppose it is hard; letting a dance video go viral, is not dissimilar to enabling any other message go viral. We are eager for it; but uncheck virality typically kills the host
A phone call is low-bandwidth and ephemeral. The time it takes to disseminate information that way to a larger group is huge. Societal self-regulation has a chance to kick in.
On the other hand you have high-bandwidth posts that once published can be upvoted by a single click and spread like wildfire.