Live data from Hacker News

Facebook overrides fact-checks when climate science is “opinion”

arstechnica.com

231–240 of 259 posts

Re: Facebook overrides fact-checks when climate science is “opinion”

#231

Earlier quoted context omitted.

How does this argument apply to Google? When you search something e.g. is climate change real, Google ranks the results algorithmically. If Google doesn't do fact checking, are they liable for false statements?

It applies just as much because it is a fundamentally technically illiterate argument. It already ignores that the whole point of the algorithim is to prioritize. In this case just as much means "none" for both because it is asking for an outright contradiction to - provide prioritization without any bias. Or bias without bias. Nonsensical stupidity. The mobster with a cudgel making vague threats rhetoric is just ano…

You may be misunderstanding. In this case, I think people conceive of an "obvious" baseline to compare to - presenting posts of your friends in a chronological order.

There's no inherent contradiction when people say "this simple, transparent, obvious algorithm is better". Even if implementing it would in fact result in forgoing billions of dollars.

Re: Facebook overrides fact-checks when climate science is “opinion”

#232
post #43

Earlier quoted context omitted.

Facebook can’t have it both ways. They can’t tell Congress they are like the telephone company (an agnostic communication channel) and then turn around and censor whatever they feel like. Either take responsibility for the content or stay out of it.

The whole point of section 230 of the CDA was to say that they can. Before 230, there had been court rulings that said if a platform moderated user supplied content they were liable for that content. Section 230 was written to reverse those rulings. I'm not sure why, but a myth has taken hold that under 230 you have to be neutral to get the liability shield.

Maybe some people would actually like to repeal that though.

Re: Facebook overrides fact-checks when climate science is “opinion”

#233
post #87

Earlier quoted context omitted.

> A thing that does not occur very often, statistically, can get blown completely out of proportion. 9/11 only happened once, and killed far fewer Americans than American police have since then. Therefore terrorism isn't a problem. (This argument isn't false , but it's highly misleading!)

I don't want to go down this rabbit hole but a lot of people believe that terrorism isn't a large problem. The outsized response to it has been. Neither position is an absurd one to take.

Generally speaking, your attitude towards anything depends on your assessment of what will happen in the future if it is unchecked.

If you see a crack in a dam, and a rivulet of water coming through, you base your actions on whether you think it's going to become larger.

If a person is pretty certain that the dam is about to collapse, then talking about the water coming through right now seems palpably absurd, but reverse the assumptions and taking drastic action seems equally absurd.

There's no neutral way based on facts and independent of opinions to decide what's absurd, extreme, or hysterical. But at the same time, different views are different, and some are going to be severely wrong about anything controversial and important.

I don't know how you make a generic framework that tells you the right thing to do when people are talking about invading Iraq, and also tells you the right thing to do when people are talking about shutdowns for covid.

Re: Facebook overrides fact-checks when climate science is “opinion”

#234

Earlier quoted context omitted.

Thanks for finding the article. I would argue that this is clearly in the realm of interpretive history. The core statement "Lee opposed both secession and slavery" is too poorly defined to have an objective answer. It is a question of what a historical figure thought and felt comes with all the nuance of a real human being. Part of what I object to with the "fact checking" is that it often tries to adjudicate questi…

> The core statement "Lee opposed both secession and slavery" is too poorly defined to have an objective answer ... ? Doesn't the burden of factuality go the OTHER way though? It's an objective-sounding statement of what purports to be a fact. But it's belied by other obvious facts about the man's life. It makes him sound like a union abolitionist and not a slave-holding confederate, which is misleading. I suppose yo…

I think there mere fact that the two of us can read the same political article and then debate the interpretation of "opposition" in this context lends credence the idea it is a statement of interpretive history, and not fact. Similarly, it is open to the reader's interpretation. You read it as making him sound like a union abolitionist, I did not.

I totally agree that "Lee owned other human beings" and "Lee fought against the United States" are factual statements, and could be fact checked. So could statements that "Prior to the war, Lee wrote that slavery was evil" and "Both Lee and Lincoln would be considered a racist by modern standards, and had very similar opinions on slavery."

I agree that facts can lead readers to misinterpretations. Given how hard it is to verify actual facts, I don't think it should be Facebook's job protect readers from facts which are accurate, but could lead to undesirable world views.

In this instance, one group clearly wants to paint Lee as a moral man trying to navigate the political and economic realities of the antebellum south. Another group clearly wants to paint a picture of Lee as a symbol of southern racism and oppression. These are interpretations of history.

Re: Facebook overrides fact-checks when climate science is “opinion”

#235

Earlier quoted context omitted.

I agree there are edge cases. Are you saying the nipple issue is harder than moderating politically contested statements of fact regarding the nature of the world we live in?

I thought you were saying that “universal rules for nudity (subcontent: nipples)” are easier to write than “universal rules for fact-checking.“ If so, I disagree. They are either: a) equally hard, given the common understanding of what “facts” mean nowadays, or b) nudity is harder, given cultural contexts. Culture makes a lot of difference for what’s considered nudity (e.g. women breastfeeding goats, as illustrated i…

>I thought you were saying that “universal rules for nudity (subcontent: nipples)” are easier to write than “universal rules for fact-checking.“ If so, I disagree. They are either:

I think we still have a misunderstanding. I am saying that enforceable rules for nipples are easier to write, regardless of the cultural approval of those rules. A well written rule can be enforced by only looking at the content that is on the platform ( eg the image). That is not to say people might disagree (its Art!) or they might get it wrong (its a male nipple).

When it comes to statements of fact, this is not the case. The enforcers must draw conclusion based on off platform research, a presumed intent of the post, and/or a presumed interpretation.

Re: Facebook overrides fact-checks when climate science is “opinion”

#236

According to the piece, the fact checkers claimed an opinion piece was false because all of the facts in it were cherry picked: "The researchers found that the post by the CO2 Coalition was based on cherry-picked information to mislead readers into thinking climate science models are wrong about global warming." This is not a logical argument. Facts are correct or incorrect, whether they're cherrypicked or not.

They claimed it was misleading, no?

Re: Facebook overrides fact-checks when climate science is “opinion”

#237
post #218

Earlier quoted context omitted.

The best mechanism against bad ideas is to allow trustworthy people who can speak clearly and honestly about those ideas to do so. The first amendment, in addition to freedom of speech, was supposed to enshrine the idea of a free press so that the two could keep the government in check. The press was supposed to allow honest trustworthy people to become known and to inform the public about issues of the day, and allo…

First of all, talking about "the press" as though it's some monolithic entity is just wrong. "The Press" is made up of thousands and thousands of uncoordinated organizations and people. There are still honest and trustworthy members of "the press" out there and available. Just as each individual member of "the press" has a responsibility and duty to serve that role well, we also have a shared responsibility to hold t…

The question of who is exactly the point, the American ideal is all people get a voice. The more our voices are squashed by a powerful special interest, like the collective media and big tech, the more deranged our society will become. Everyone goes insane when left to themselves and group dynamics are no different. Whatever group of people you revere as 'good' the most, need to go through trials of their character by opposing voices or your favorite group is just getting more and more deranged over time. This is precisely why freedom of speech is so important, because wielding the power of 'who' can talk is far too powerful and far too easily abused.

Re: Facebook overrides fact-checks when climate science is “opinion”

#238

Earlier quoted context omitted.

> The core statement "Lee opposed both secession and slavery" is too poorly defined to have an objective answer ... ? Doesn't the burden of factuality go the OTHER way though? It's an objective-sounding statement of what purports to be a fact. But it's belied by other obvious facts about the man's life. It makes him sound like a union abolitionist and not a slave-holding confederate, which is misleading. I suppose yo…

I think there mere fact that the two of us can read the same political article and then debate the interpretation of "opposition" in this context lends credence the idea it is a statement of interpretive history, and not fact. Similarly, it is open to the reader's interpretation. You read it as making him sound like a union abolitionist, I did not. I totally agree that "Lee owned other human beings" and "Lee fought a…

> You read it as making him sound like a union abolitionist, I did not.

Of course not. Because you know the context and have made your own decision to excuse his slave holding and treason when deciding whether he should be celebrated with giant statues. Which is your call to make, I guess.

But the point is that not everyone knows that context. And for those who don't, this advertisement, which is designed to affect public policy, is prima facie misleading because it presents two facts that are (at best) subjects of controversy.

You'd never allow this about other stuff you care about. Imagine a Facebook story with "The constitution gives you no right to own an assault rifle". Equally true, by your standards. Not really acceptable though, right?

> In this instance, one group clearly wants to paint Lee as a moral man trying to navigate the political and economic realities of the antebellum south. Another group clearly wants to paint a picture of Lee as a symbol of southern racism and oppression. These are interpretations of history.

And Facebook wasn't labeling considered interpretations. It was labeling something that looked like a fact. I don't think it's so heavy a burden for those who want to write a defense of Lee not to engage in the kind of spin in that story, is it?

Re: Facebook overrides fact-checks when climate science is “opinion”

#239

Earlier quoted context omitted.

It applies just as much because it is a fundamentally technically illiterate argument. It already ignores that the whole point of the algorithim is to prioritize. In this case just as much means "none" for both because it is asking for an outright contradiction to - provide prioritization without any bias. Or bias without bias. Nonsensical stupidity. The mobster with a cudgel making vague threats rhetoric is just ano…

I disagree. The "algorithm" is fundamentally an optimization problem and companies get to choose what the fitness function for each algorithm is. Facebook chooses "engagement". This is not an accident nor some fundamental property of math. It's a product decision made by a human. One that can be changed. And there is a fundamental difference between Google's algorithm and Facebook's: control. Google requires user inp…

There is an implicit search on Facebook though, namely the personal information you’ve entered into your profile and your relationships.

Re: Facebook overrides fact-checks when climate science is “opinion”

#240

Earlier quoted context omitted.

Let us be honest here. The people that have the job to sort out this stuff might also not be the brightest stars in the night sky. If I had vested interests where manipulation of social media nets me advantages, I would be very content with the situation.

In this case, the low-level moderators were apparently overruled by a self-describe “conservative” higher-up.

Yes, and I think whoever is responsible has a deeper understanding. People focusing on sanctioning people not believing in climate change are a larger hindrance to find support for climate policies than complete deniers.

I believe in climate change, but I am very skeptical why we need to censor people for speaking their mind. So if they had free reign, it would have been a net negative for everyone.

Post reply on HN