I Was a Facebook Fact-Checker (2019)
buzzfeednews.com
buzzfeednews.com
- firstly, just because you can’t eradicate all fake news and misinformation does not make the effort “doomed” - you can’t get rid of all crime in the world either, or all disease, but you can still try and reduce it and it’s a worthy effort.
- the writer conflates a bunch of things - they wants the algorithms to be open, transparent and opt-in. While that would be nice, I don’t think that’s required to eliminate misinformation (nor will it solve it)
- what makes the article less credible is the author ends it with complaining about revenue sharing and saying Facebook must pay writers and publications. While potentially that could be nice it does not seem to have anything to do with the issue at hand and makes it read just like any other “let’s crap on Facebook” article
It might reduce it (see your first point).
It would be much more helpful to explain and teach by example the underlying cognitive mechanism that tricked you in the first place.
An example. Alot of climate change "sceptics" repeat over and over again, that CO2 is just ~2% of earth atmosphere (doesn't matter if the number is correct or not) and they conclude that therefore CO2's effect on the climate is neglectable. This is wrong reasoning, you got tricked by a small number, but it appears right for a superficial look. (Following this logic, cyanide would be harmless too.)
The over and over (and over and over and over) repetition of that aspect is like a chain reaction you can only effectively stop by halting the underlying process. Otherwise some else fake news get spread next iteration.
"Required" might be a strong word, but "would be nice" is too weak, IMO.
I suppose it depends on what you see as the goals and purpose of moderation. I've seen a lot of agreement about SM requiring moderation, while totally disagreeing about what, why and to what end.
If fairness s concern, nevermind freedom, openness is a factor. Also, while he doesn't state it... the author is concerned about moderating facebook, not just facebook users. That makes some sense, from where he's sat.
The job is to "prevent the spread of misinformation" on FB, especially headline stuff: political/electoral misinformation, covid related, etc.^ He sees this stuff repeatedly go viral... and that logically leads to the platform and recommendation algorithm itself. Applying moderation after may not (in his view) be enough to do the job. Opaqueness probably makes it as hard for moderation teams to understand the nuances as it does for us.
Maybe openness and free software adjacent principles are not the only way to approach an actual solution, but most of the best ideas are not unrelated. If you think of bullshit propaganda going viral as a bug, transparency is definitely relevant.
Opt-in & respecting user wishes generally, are also part of a lot of hypothetical solutions. One of the ways pre-FB social media dynamics was various "bad behaviour" rooms. HN works as an example. You can choose to view dead comments. Most choose not to. Still, that tiny piece of transparency and user choice moderates some of the moderation's weaknesses.
I don't think you can say anything definitively, but there are logical links between FB's control oriented model and the problems moderators are tasked with fixing.
^He cites Myanmar as an example. This is in a language & culture that FB don't speak. They really don't know what is going on, and I think it's likely they heard about FB's role in causing major violence against a minority from a third party. COmpare that to the level of
This is an interesting take. Doubly interesting to me because this is evidence for something that was always likely - Facebook fact checkers aren't particularly qualified in the areas they are fact checking.
This woman with a background in journalism would have been taking down posts by doctors and nurses for being COVID misionformation.
Furthermore, doctors and nurses can (and have) also spread COVID misinformation, it's not a bad thing to have journalists check them.
Many things cannot be easily audited, are about points of view, and fact checking enables censorship and bias towards one side or another. And last but not least, I am not a kid. When I go out at night I know there are good and bad places and people. I figure out what is good and bad and if I make a mistake, I learn.
I am an adult. I do not need babysitting. If there were far fewer people to control the rest of population, special mention to politicians here, who are kind of a mafia, then people would care way less about news. I know a bunch of people, just like me, that have taken the road to believe NOTHING from mainstream media because it is so obviously manipulated, even fact checking.
We have our own and different views and philosophies but we reached the conclusion that politics and trading to extract resources and create obedience from people are the main goals of people who want to control others.
I do not need fact checking.
Most people would agree an imperfect attempt with lots of transparency is better than no attempt at all.
The first thing you learn in journalism is that you can't be 100% objective, even if you're 100% truthful.
Let's say you're reporting on Obama's birth certificate, and you write 20 pages of a deep dive into where he was born and all the evidence for it. You are complete and thorough, and you end up concluding that he couldn't have been born in Kenya.
Just the fact that you wrote that much to debunk a nonsensical, evidence-free lie is actually going to make it seem like they had evidence!
You can also neglect to report on things. You can report on 150 burglaries a day and give people the impression that burglaries are skyrocketing when they aren't.
So of course there will be subjectivity that leaks out, but that's not something we should fear to the point of giving up.
Nobody wants or needs the media - even less social media - as proxy-judges of the information. Everbody just wants the information and judge it for themselves.
That is because most of the time, fact-control becomes opinion-control and we see exactly this happening just before our eyes.
Thus people on either side of the political spectrum will support the suppressing of the opinion of their political opposition.
At the moment this is happening with the evidently unholy coalition between the left and big tech as well as big media. Another unholy coalition is big pharma + governance and big media where anything slightly critical of certain vaccines will get insta-banned on Youtube for example.
May I suggest that you only support this behaviour as it currently happens to align with your views?
This seems to be a fundamental misunderstanding of the basics of journalism. The primary job of journalists is to filter through the events and facts of the day and tell you what is important. Judgement of information is inherent to the job. The top story on the front page of the newspaper is chosen by editors not by random draw. They put it there because they think it is the most important bit of news in the entire paper. There is no way to make that kind of qualitative judgements on the relative importance of different pieces of news without inserting potential bias.
People may say they want this, but manifestly do not. There is too much information.
To illustrate: many people I know cared a great deal about whether the country was lied to about WMDs when we went to Iraq. As far as I know, zero of them have read the report by the Senate Intelligence Committee[0] but all were very glad to have the report summarized for them by the news media.
That act of summarizing, of picking out which information is and is not important, is itself an act of journalism. It is the curation of the set of shared facts, shared reality, in the public discourse.
[0]https://www.intelligence.senate.gov/sites/default/files/publ...
Exactly, I'm totally with you. That is the essence of journalism I would even say. But that is where the job of the journalist ends, no sprinkling in of their own presuppositions.
It's impossible for presuppositions not to seep into journalism, especially in a capitalist system where clickbait, outrage news etc. sells better than plain retelling of facts.
Government run news organizations with no profit motive have a very sorry record.
I'd like to be more specific than "a huge proportion" but it's hard because I've mostly read the more widely shared and thus sensationalist stories, which will have much bigger distortions on average.
(Social) media fact-checking also often becomes a form of silent censorship.
First, fact-checkers don't just publish fact-checks. They also work behind the scenes at the "high end" newspapers that you mention.
Second, websites like Snopes originally appeared to fact-check things like chain emails, memes, and other rumors. Mainstream news sites now do this to some extent, but there's still not a 100% overlap.
Also, sometimes there just isn't a story to a fact-check. The reader just wants to see a check of all the facts in a political speech or a debate. A series of 100 bullet points is not a compelling article, but it might be a perfectly fine fact-check.
Why? What's the difference? Analysis of a political speech or debate is definitely the kind of thing I expect from newspapers or equivalent, and it's not like changing the name makes it easier to read.
Not sure I would trust high end newspapers either, in Germany we had a small scandal with Claas Relotius who received over a dozen awards for outstanding journalism. Most of his articles where more fiction than fact. The newspapers publishing his articles outright ignored any complaints by his victims and even the papers claiming to have "world class" fact checkers didn't bother to review the output of high profile authors. It seems that to journalists anything is fine as long as it captures the attention of their target audience.
This is what Facebook fact checkers are??
They have hearings that go on public record and can be appealed up to a federal level?
Judges interpret the law, not current events. This comparison seems pretty meaningless.
The most that fact-checkers do when they are egregiously and utterly incorrect is a correction to an eye-rolling "needs context" label.
The standards of objectivity of fact-checkers (who are all left-leaning partisans) are ridiculously slanted. One set of strict standards for my opposition, another set of lax, wink-wink standards for myself - modern day "fact-checking". Any criticism of the same that becomes popular results in being banned.
To anyone for whom facts that don't agree with their worldview are signs of a conspiracy, fact checking is pointless anyway. All fact checks that show they're wrong are just "proof" of the conspiracy.
There's little need for a conspiracy here, just the Red Tribe distrust of institutions making them less likely to support fact checkers. And fact checkers get their money either from donations or from ads/clicks, so the incentive structure will push them left as that's the only way to survive in a competitive environment.
The limitations of objective fact-checking means we need to understand where we can/should apply it, and its efficacy are limited.
Judges or a legal system are a good example. These are also truth finding institutions. They are very expensive institutions, and produce a very limited stream of narrow truths. That's because truth finding is, as you say, difficult.
Law is an extremely limited system. We (in theory) apply it as little as possible, when other solutions have failed. The system is set up to verify or falsify claims as narrowly as possible to find this solution. Law has specifically defined truth categories like "reasonable doubt," and other rules about how to lean in cases of ambiguity.
The evidence for "person X killed person Y" may be sufficient to acquit person Z of a murder, but not to convict person X.
Fact checking can be used as a tool that newspaper editors use to maintain quality. It doesn't do the same thing in a different context. Fact checking at a newspaper, in wikipedia or for social media is a totally different thing with different limitations and applicability.
> Let's say you're reporting on Obama's birth certificate, and you write 20 pages of a deep dive into where he was born and all the evidence for it. You are complete and thorough, and you end up concluding that he couldn't have been born in Kenya.
Works perfectly fine for personal fact checking. You can do it like this. If it's being done in court, an editors's office, a wikipedia discussion page or FB's version of mechanical turk for moderators.... you will need a different method.
I don't see how that conclusion is drawn at all. The legal system is run by highly trained individuals in as public a way as possible, bounded by its own set of standard laws and procedures, to very specific sets of issues, often times throwing out cases because they are not "resolvable" by the narrow capabilities of the system.
To compare a random anonymous person behind closed doors spitting out a "fact check" on their laptop to judges is an insult to judges and the legal system.
I can't help but feel that people who have their bias confirmed by the title rarely bother to read past it + the big red "False!"
I’ve grown to realise that this is not only intentional because it is harder to disprove, but also people tend to be more emotional than statistical so having an emotional argument tends to resonate and get shared on social media with far greater prevalence than statistical memes.
With the scientific studies, we see here on HN all the time how studies, viewed from various angles, can be contentious and not easy to decide with certainty about.
And surprisingly, quotes and "what I saw" situations often do get remembered differently for two people observing the same event. There's a lot of room here for some seemingly simple facts to be very difficult to know the actual fact of.
Another particularly goood piece is this from The Guardian in 2018:
'They don't care': Facebook factchecking in disarray as journalists push to cut ties
Current and former Facebook factcheckers told the Guardian that the tech platform’s collaboration with outside reporters has produced minimal results and that they’ve lost trust in Facebook, which has repeatedly refused to release meaningful data about the impacts of their work. Some said Facebook’s hiring of a PR firm that used an antisemitic narrative to discredit critics – fueling the same kind of propaganda factcheckers regularly debunk – should be a deal-breaker.
https://www.theguardian.com/technology/2018/dec/13/they-dont...
More are appearing on her feed, see: https://nitter.kavin.rocks/brooklynmarie
In other words, I daresay that people of less education and less experience reading (or where the article is written in a language which is not the reader's first language) would not realize it is satire.
This has happened many times where certain political groups have promoted Onion stories that appear to favor their policies - but which are clearly satire to some of us.
So even if the fact checkers could stay on top of the flood of misinformation, I expect there could be a minimal formula for getting content marked as Satire while still delivering false information to gullible/ignorant readers.
A problem too great (or late) to solve is the relative ignorance of many of the readers. These are also the people who vote in politicians who work directly against their needs. When a person can stand in a burning building and say, "We're sure glad we got our guy leading the government, because those other guys would have ruined everything we care about." No amount of fact checking can solve this mental problem.
Good luck trying to teach a >20 year old this.
When you teach children that the scientific method is useless, logic is not necessary, and therefore Creationism is fine, repairing that "damage" once Children are older is pretty much impossible.
You have "habituated" them to the lie, just in the exact same way that "child soldiers" and "terrorist" get habituated. Part of the habituation process is making them personally invested on the lie, e.g., for a child soldier or a terrorist that's their first "kill". From that moment on, either they are right, or they are a horrible human being.
Same here, it might start with Creationism, then shaming someone for wanting an abortion, etc. One of these is going to be their first "kill" moment, their point of no return, and then the only way to "heal" is for them to discover that they are horrible human beings. Pretty much impossible to heal in practice.
In the exact same way that this is not the child soldiers fault, it is not the fault of children that have been educated their way to end up raiding the capitol. I'd go as far that it is not the fault of these children to even then lobby as adults for the process to continue, for more creationism, for more extremism, etc.
Maybe at some point someone in the USA was at fault for allowing this to happen. But at this point I think that a big part of the population particularly in central america are just victims of the system. They were raised into what they are now, and lots of psychological hard work would be required to "fix" any single one of them.
Simplifying this issue to "this people are just dumb, and facebook should send them a link to a page that teaches them how to read" would be as effective as trying to address terrorism by repeating on TV that killing people is bad.
What's needed of who? I agree that epistemological hygiene is good, but how is FB supposed to tackle this. Even the education system may have trouble, and even if they don't it will take a generation.
Problems need to be framed in ways that don't avoid dealing with them.
Yes, it will take a generation for people to figure out how to navigate a world like this, but ultimately its the only option that doesn't end up with a centralised arbiter of truth restricting communication (where a dystopia will inevitably follow in a few decades).
Quite often it is not the issue that people don't know they might be wrong, but that they don't care whether they are. In their worldview small (or sometimes even big) lies are justified in a "war" that conveys "a bigger truth". In their eyes, people who stick with the truth no matter what are naive, stupid or both.
As with the deprogramming of Nazis like my grandfather after WWII, the hardest ones to get back to the truth are ot the ones that couldn't tell, but the ones that could tell, knew it was wrong, knew what was going on and went for it anyways.
Of course that one has by now been "fact-checked" off YouTube.
So even if I wanted to, I couldn't refer you to what was a very neat presentation of some of the issues with fact-checkers with clear evidence showing where these in fact present misinformation and demonstrate clear political bias.
And that is the problem with fact-checkers. (In this case, the "fact-checker" called YouTube that decided it didn't like the facts presented in said video).
A quick search of YouTube shows a very large number of videos using that (somewhat clichéd) title, so as far as I can tell that video may still be available. I call a fact-check on whether who fact-checks the fact-checkers was in fact fact-checked off YouTube.
https://www.wnycstudios.org/podcasts/radiolab/articles/faceb...
You will say, well, the fact-checkers tend to get political. Nevertheless, they're still fundamentally different than censors.
Right, identifying information as “false” is censoring it on Facebook because it suppresses it.
Obviously there is a stance going around that 'calling things wrong' is being done for political/ideological purposes, but that doesn't make fact-checking censorship _in principle_.
You can never hire enough "fact-checkers" thank God, it's democracy in action.