Facebook versus The BMJ: when fact checking goes wrong • Hacker News Board

Re: Facebook versus The BMJ: when fact checking goes wrong

#201

Post by kryogen1c » Thu, Jan 20, 2022, 3:17 PM UTC

SilasX wrote: post #178

jerf wrote: post #171

Earlier quoted context omitted.

"The vast majority of fact checking exists to call out stuff that's demonstrably false," Prove it, and prove that the underpaid non-specialist moderator who labelled it as such was justified even if they were right. Facebook is not and can not be an arbiter of truth in the abstract, and in the concrete the specific people doing the work certainly are not. Where are Facebook getting these experts who know so much abou…

>Prove it, and prove that the underpaid non-specialist moderator who labelled it as such was justified even if they were right. >Facebook is not and can not be an arbiter of truth in the abstract, I'm not sure you're replying to the right comment, because I wasn't claiming that the BMJ story was demonstrably false[1], and I wasn't defending any action that labeled it as such. The action was to label it as "needs cont…

> I wasn't claiming that the BMJ story was demonstrably false[1], and I wasn't defending any action that labeled it as such. The action was to label it as "needs context", a category which specifically exists not to claim that the story is false.

which stories on facebook dont need context?

this is not snark and the question is not trivial. nothing - NOTHING - means the same in isolation as contextually. facebook isnt applying this label equally. theyre selectively moderating and claiming impartiality.

Re: Facebook versus The BMJ: when fact checking goes wrong

#202

Post by aww_dang » Thu, Jan 20, 2022, 3:20 PM UTC

rdtsc wrote: post #64

> Facebook’s ‘independent fact-checker’ doesn’t like the wording of the article by the BMJ. And if I don’t delete my post, they are threatening to make my posts less visible. Obviously, I will not delete my post . . . If it seems like I’ve disappeared for a while, you’ll know why.” Just treat them as ideological censors, not unlike those employed by authoritarian regimes. You want to publish an article criticizing th…

>No, they won’t send you to a labor camp or come after your family.

https://www.bbc.com/news/world-australia-54007824

"The arrest of a woman in Australia for promoting an anti-lockdown protest online has drawn criticism, after video of the incident went viral.

Footage shows officers handcuffing pregnant woman Zoe-Lee Buhler, 28, in her home in Victoria on Wednesday in front of her partner and children."

>And, yes it’s a private company and if anyone doesn’t like it they can leave, but the mechanics of censorship are exactly the same.

"MS. PSAKI: Sure. Well, I would say first, it shouldn’t come as any surprise that we’re in regular touch with social media platforms — just like we’re in regular touch with all of you and your media outlets — about areas where we have concern, information that might be useful, information that may or may not be interesting to your viewers.

You all make decisions, just like the social media platforms make decisions, even though they’re a private-sector company and different, but just as an example.

So we are ma- — regularly making sure social media platforms are aware of the latest narratives dangerous to public health that we and many other Americans seeing — are seeing across all of social and traditional media. And we work to engage with them to better understand the enforcement of social media platform policies."

https://www.whitehouse.gov/briefing-room/press-briefings/202...

https://en.wikipedia.org/wiki/Boiling_frog#As_metaphor

Re: Facebook versus The BMJ: when fact checking goes wrong

#203

Post by treis » Thu, Jan 20, 2022, 3:21 PM UTC

CWuestefeld wrote: post #185

treis wrote: post #157

Earlier quoted context omitted.

Having read the original article and BMJ's follow up I think Facebook is right here. The original article does lack context and is not a measured response to the problems found. Especially so considering it was published in November 2021 after hundreds of millions of doses were given and the vaccine was obviously safe and effective. They should have known better than to give fodder to the anti-vax movement.

You seem to be making a normative statement, that journalism should be making editorial decisions in order to nudge toward the outcome they believe is correct (or that some societal watchdog believes is correct?) rather than what's normally considered the goal of detached and dispassionate reporting.

Did you read the original BMJ article? It was not detached and dispassionate reporting. That's the problem. It was alarmist and blew the issue way out of proportion because it lacked appropriate context.

The only mitigating information they gave was that the company only conducted trials on 1,000 out of 44,000 enrolled patients. But that was buried 10 paragraphs in.

There was no neutral party evaluation of the seriousness of the errors.

There was no neutral party evaluation of the implication of the errors on the soundness of the study.

There was no mention of the obvious effectiveness and safety of the vaccine.

In short, it did lack context and it should have been obvious to the BMJ that it would be inflammatory fodder for the anti-vax movement. It should have been nixed and rewritten to frame the errors in a more, as you say, detached and dispassionate way.

Re: Facebook versus The BMJ: when fact checking goes wrong

#204

Post by NoPie » Thu, Jan 20, 2022, 3:22 PM UTC

Tabular-Iceberg wrote: post #125

Avamander wrote: post #36

Earlier quoted context omitted.

Plenty of quacks follow the principle of "do as I say, not as I do." It's a very common behaviour amongst populist politicians to talk shit about vaccines and then being fully vaccinated.

I didn’t read the article, but isn’t it plausible that whatever they did allegedly did wrong wasn’t significant enough to warrant recommending against taking the vaccine?

They didn't recommend anything against the vaccine. It is just overzealous fact-checkers who (similarly to antivax) couldn't admit that they were confused about the article and mislabelled it as false.

Re: Facebook versus The BMJ: when fact checking goes wrong

#205

Post by aiilns » Thu, Jan 20, 2022, 3:27 PM UTC

SilasX wrote: post #95

wolframhempel wrote: post #67

Earlier quoted context omitted.

It's true, there are fundamental problems with the concept of "fact-checking". Often the "fact" is just the majority consent, as e.g., demonstrated by the banning of the Wuhan lab leak theory prior to it becoming mainstream. But more broadly, there is a stark imbalance between the resources of the fact-checker (a few minutes or hours to decide) and the author who (e.g. in this case) conducted a month-long, in-depth i…

Be careful not to jump the gun here. The vast majority of fact checking exists to call out stuff that's demonstrably false, and it's to their credit that social media platforms link a refutation. But when you have a huge number of stories, there are going to be edge cases, and BMJ got snared by one here. As I understand it: BMJ found legit problems with one vaccine's clinical trial. But this kind of thing gets seized…

How is _BMJ_ an "edge case"?

And why should Facebook and its partners decide on what the conclusions to be drawn from this story should be?

I'm sorry, but to me your attitude seems arrogant and condescending: the masses cannot tell truth from lie and cannot be trusted to reach the correct conclusions. And I suppose that, if asked, you would advocate for education with emphasis put on critical thinking.

But you can't be both for critical thinking and for this paternalistic attitude where people get their cues from the truth-arbiter du jour.

Re: Facebook versus The BMJ: when fact checking goes wrong

#206

Post by ModernMech » Thu, Jan 20, 2022, 3:30 PM UTC

willcipriano wrote: post #161

mc32 wrote: post #123

Earlier quoted context omitted.

I’m not sure about the prevalence but one major peeve of mine is when someone states, X is ten meters high. Fact check: Wrong, X is 9.6 meters high. It takes away credibility.

One of my favorites, they have since been called out on it and had to fix it: This statement "We did not even have a federal income tax in this country until 1913." was rated "mostly true"[0] While this one "the U.S. federal income tax rate was 0 percent until 1913." was rated "half true"[1] What more evidence do you need that this whole thing is sort of arbitrary? [0] https://www.politifact.com/factchecks/2015/aug/2…

I mean, the little cartoon meter is pretty arbitrary. What’s the difference between half true and mostly true? Those two phrases seem mostly equivalent to me.

But the more important thing is the analyses I think, they are quite similar in content. And raise the same caveat about the civil war taxes. These fact checks were 3 years apart, so it may be completely different groups of people writing these fact checks. Yet, they both arrive at similar conclusions with slightly different language. I’d say that’s an okay result.

Re: Facebook versus The BMJ: when fact checking goes wrong

#207

Post by FredPret » Thu, Jan 20, 2022, 3:32 PM UTC

The very existence of fact-checking demonstrates a deplorable intellectual arrogance.

Do these people meditate deeple and then receive divine inspiration as to what is really true? Do they have transcendent superbrains that work better than everybody else’s? Do they have access to Secret Wikipedia that contains only true and current facts?

The truth of everything we do will be visible only centuries from now, after all strong political feelings we have shifts somewhere else. (Today we look at Ancient Rome and do not care about the Reds and the Greens.)

Also, most human statements cannot be evaluated for truth, including this one. We communicate mostly feelings, sentiments, our sense of something, and the odd number or two.

Re: Facebook versus The BMJ: when fact checking goes wrong

#208

Post by honkdaddy » Thu, Jan 20, 2022, 3:33 PM UTC

krapp wrote: post #145

evancoop wrote: post #135

On some level, this is the strongest argument in favor of Facebook's original position - that they oughtn't be responsible for distinguishing fact from fiction. Ask any entity to discern truth at scale and there will be errors and vociferous debates. So, what's preferable? Unfettered venom and lies mixed with the best attempts at rationality or the discourse moderated by unseen arbiters and algorithms? Ideally, there…

False dichotomy. We can and do already have both, on multiple forums, and even on the big social media platforms. That's the case despite there being a few large entities, just as small restaurants continue to exist despite the existence of MacDonald's. The premise that "centralization" of the internet through FAANG gatekeepers forces people to adopt either one or another model as universal to either accept or reject…

>nor will it turn entirely into 4chan.

Sadly, it seems that instead, the internet is turning entirely into Facebook :(

Re: Facebook versus The BMJ: when fact checking goes wrong

#209

Post by tagoregrtst » Thu, Jan 20, 2022, 3:33 PM UTC

gadders wrote: post #121

tagoregrtst wrote: post #117

Earlier quoted context omitted.

“ and no idea if true or not” Well I think traditional media have acknowledged that the laptop is real. Of course that doesn't mean there was anything salacious on it and the excerpts arent in fact fake

The laptop is definitely real and his. I think there are salacious items on it (but I'm not sure if all the items I've seen claimed were on his laptop really were). I'm not sure if the stolen diary is really his sister's. FBI raiding Project Veritas might suggest it is?e Unfortunately what is lacking is some large news organisation with lots of resources tracking down and corroborating all the information. But that w…

The diary is real, that much has been eluded to. But, again, are the excerpts that have trickled out real?

For what its worth, I believe [1] PV didn't publish anything from the diary deeming its contents irrelevant to the public.

But the stuff on laptop? I think its probably genuine. Too much corroborating material. Too many sweet board positions. Too much success in the art world (btw, the art is not bad. Just… unoriginal.)

[1] I dont follow PV so I dont know what they do or not do as a matter of fact.

Re: Facebook versus The BMJ: when fact checking goes wrong

#210

Post by ModernMech » Thu, Jan 20, 2022, 3:37 PM UTC

wang_li wrote: post #181

SilasX wrote: post #167

Earlier quoted context omitted.

Cool, but what does that have to do with the present case, where they (claim they) just add context? That's not "squelching" the information.

When fact checkers “add context” to a statement it exposes that the fact checking isn’t about correctness or truth, but rather about narrative control. Which means it’s biased and one begins to wonder what context they’re leaving out of all the statements they rate as true. In the end calling it fact checking is bullshit, it’s just propaganda. This is also demonstrated by the fact that fact checkers almost always hav…

> Truth is binary.

My name is Steve and I’m 35 years old. This statement is half true.

The speed of light is 3x10^8 m/s. This is false, but really not far from the truth.

I think given these examples, there is definitely room for a spectrum rather than a binary. Otherwise everything is false because language cannot 100% accurately represent reality, and therefore you’ll always be able to find an angle to claim something is false.