Facebook Shuts Down Researchers Looking into Misinformation | Hacker News Reader

Facebook Shuts Down Researchers Looking into Misinformation

nytimes.com

128 pointswglb36 comments

Open articleView on HN

AP: https://news.ycombinator.com/item?id=28112991

Daily Dot (first thread on HN, I think): https://news.ycombinator.com/item?id=28063637

Mozilla Statement: https://news.ycombinator.com/item?id=28068359

Just felt like there's a lot of discourse and context that could be helpful.

Thanks! Combined and macroexpanded, this is the list that I get. There may have been other threads as well:

Facebook shuts out NYU academics’ research on political ads - https://news.ycombinator.com/item?id=28112991 - Aug 2021 (34 comments)

U.S. FTC says Facebook misused privacy decree to shut down ad research - https://news.ycombinator.com/item?id=28082371 - Aug 2021 (112 comments)

Facebook’s claims about the Ad Observer are wrong - https://news.ycombinator.com/item?id=28068359 - Aug 2021 (105 comments)

Facebook bans researchers who were investigating Facebook ads - https://news.ycombinator.com/item?id=28063637 - Aug 2021 (77 comments)

AdObserver Blocked by Facebook - https://news.ycombinator.com/item?id=28059869 - Aug 2021 (46 comments)

Facebook has a problem with misinformation - it's just that their problem isn't the same as what people assume their problem is.

"Misinformation" tends to be spectacularly engaging content - which is quite profitable.

People upset about misinformation on Facebook tend to use Facebook less, which is less profitable.

The ideal solution for Facebook is that they can show misinformation to people they know will outrage-scroll as a result, while not showing misinformation to people who might be upset about Facebook not really having done much about it.

"Transparency" extensions and such violate this, badly. You might find out what sort of ads Facebook shows your aunt out one side of their site while talking about how they're combatting misinformation on the other.

They've solved the problem.

I wonder if Facebook has a big list of misinformation published by major news organizations that they aren't sure what to do with.

I often see the same people who talk about how dangerous sites like PatriotFreedomEagleNews are in terms of misinformation respond belligerently when shown evidence of similar misinformation coming from CNN or MSNBC. The list of reliably credible news sources is much shorter than most people believe.

They're reacting to the false equivalency you are drawing, not the evidence that CNN or MSNBC spread misinformation, but I doubt I'll be able to change your mind on the matter.

What makes the equivalency false?

I'm pretty sure that people upset about misinformation on Facebook are upset about misinformation being shown to other people, including people who are pleased (which is to say engaged) by misinformation.

Right. And even people who like their own brand of misinformation are not happy about other people enjoying other conflicting brands of misinformation. And this is true whether we are talking about objective misinformation or subjective "misinformation".

And so Facebook is fantastically incentivized to give each person not just what they want, but to leak as little about other people's objectionable/"objectionable" (in your eyes) experiences to you as possible.

For Facebook "cleaning up" means, optimal silo-ing.

Yeah, it's easy to imagine that as a pretty natural direction for Facebook. It's already barely "social" in the traditional sense of sharing stuff with your friends (at least for a lot of people, based on my impression). If it's way more engaging to separate each person from whatever vestige of a "real friend" group they still have, and funnel them down whatever rabbit hole they engage with the most (regardless of whether it's a misinformation rabbit hole, although I suspect many would be), then I would certainly expect Facebook to do so.

While I wouldn't be surprised at this at all given their track record and incentives, is there any evidence that this is what they're doing?

At some point, many years ago, it was reasonable to assume Facebook had something resembling honest intentions.

That has been proven wrong so many times, across such a wide range of topics, that at this point it's reasonable to assume actions by Facebook correspond to "What makes us more money," and any press release about the actions is "How can we spin this in such a way that the most people possible will swallow it and keep using Facebook?"

I lean toward your side here, but I still like to seek out confirmation for things I agree with.

Is there a name for this skewing?

Zucking it up?

I just put it in the category of "What's Good for Facebook is Good for Facebook."

I’d give the researchers the benefit of the doubt here. But I’ve been in situations where people were doing security research and in the heat of the moment vastly overstepped reasonable bounds. It’s possible the truth falls somewhere in the middle here.

I suspect they are not reasonable bounds. Facebook is telling the researchers to "trust our scrubbed data sets". This is absurd as they can provide, omit, and modify the data as they see fit. The FB provided data sets are just free advertising for FB, and useless for research. I would expect a good researcher to say "No, I'll collect my own data, thanks."

They were still using the same means of gathering that info as malware would. It's like wiretapping someone's conversation with you - not illegal in most states but they don't have to play nice, even if you were trying to 'research' their behavior.

It isn't like malware at all. They tell you exactly what data they collect, you need to opt in, and their code is opensource https://adobserver.org/

I have the right to do anything I want with the data facebook sends to my computer - including sending it to NYU researchers.

You can certainly assert that right and probably defend it successfully legally but it’s not obvious that its compatible with their terms of use.

https://m.facebook.com/terms.php

I’m no lawyer and have no idea if those terms are enforceable or even legal. Practically I think Facebook is just punching itself in the face here, but it is their platform and they do have some say in how it’s used.

The point of my original post is that I’ve had experience where researchers try to use the ends to justify their means and then cry foul when they are told that this means can’t be allowed to continue. (Eg in the infosec context rather than breaking into an account they own through some vulnerability, set up automation to break into thousands of those owned by customers and download all of their data).

Again I’m giving the researchers the benefit of the doubt, but i think the phenomenon still applies and in the general case is worth paying attention to when this kind of headline comes out.

> It’s possible the truth falls somewhere in the middle here.

Only if you believe Facebook has authority to decide what users do with the data Facebook shows them on their own machines. I find this suggestion repellent, the same as if I bought a copy of the New York Times, and afterwards the newspaper tried to forbid me from sharing which ads I saw on its front-page.

Authority comes from the law but Facebook obviously have some rights in that context. Copyright in the very least.

I thought we just heard for 6 months that private companies can ban whoever they want…

Did you read somewhere that Facebook didn't have the right to do this?

There's a difference between legality and ethics.

What exactly were the ethics behind removing any reference to wuhan institute of virology being the source of covid?

Ethics to you does not mean ethics to me.

Okay great, let's discuss the ethics of twitter banning the sitting president of the united states. This is ridiculous. Those of concerned by the ridiculous excusal of this behavior as being okay because they're 'private companies' made this argument many months ago only to be told we're in favor of overriding human rights.

At the time he was banned he was actively attempting a coup. Many of his followers still believe he will be reinstated as POTUS and he has done little to disabuse them of that notion.

Considering the riot in the capitol building started while he was giving a speech during which he asked those listening to him to make their voices 'peacefully heard'.

Many other American politicians were behind similar riots and get away with it. For example, the riots surrounding the nomination and confirmation of Justice Kavanaugh where the Supreme Court building was stormed.

Actually Twitter suspended trump two days later. Now that we have more evidence of trump trying to have the DOJ to lie in order to keep him POTUS it is hard to argue that Twitter was in the wrong. Their actions were certainly ethical as the coup attempt was still in progress.

Many of his followers still believe he will be reinstated as POTUS, allowing him the bullhorn of his Twitter account would be unethical as it would make it easier to galvanize these folks into action again.

> Many of his followers still believe he will be reinstated as POTUS, allowing him the bullhorn of his Twitter account would be unethical as it would make it easier to galvanize these folks into action again.

That's quite ridiculous. Twitter was more than happy to back open rebellion in other countries, such as those in the Middle East. It comes down to a matter of what is Twitter (and Facebook). Are they just platforms, or are they in the business of publihing certain opinions.

It seems Facebook is doing what Amazon does with its workers: Prevent the people from talking to each other, to get a real idea of what the platform is doing to them.

This tool the researchers developed was an opt-in program for people to submit to the researchers what they were served by Facebook. Facebook doesn't want the people, the government, or its competitors realizing the ways in which they serve content to people, or how tolerant they are of lies and hate on their platform if it makes them money.

Much easier than shutting down misinformation. In other news, of you want to reduce the number of new Covid cases, all you have to do is stop counting!

You are the product, not the primary user for Facebook. With this comes the spiral to create more engagement to be able to "sell" you more (selling your attention in the form of ads).

With this in mind it should come at no surprise that there is a possibility that Facebook promotes content based on its engagement, and not on its correctness nor its contribution to society.

I am interested if these researchers will post their learnings so far. Their message resonates with me personally due to the spiral I have seen among friends and family into Facebook content that is intentionally misleading (on all fronts, not just political or about vaccines).

misinformation: noun; information that causes individuals or groups to question the official narrative

And non-governmental creators and spreaders of misinformation love calling out the government for the same as an apparently effective cover for their own lunacy.

False official narratives are only one slice of the mucky pie.

Who has more power... the government or some tiktok influencer?

Research into "misinformation" is often used to justify politically-motivated censorship. If this research was intended, by its funders and/or the researchers themselves, to further politically-motivated censorship, then I have no sympathy for the researchers.

Reply on news.ycombinator.com