Big Tech Can’t Stop The Lies

IRA FLATOW: This is Science Friday. I’m Ira Flatow. Maybe you’ve seen a relative say on Facebook that President-elect Joe Biden did not win the election two weeks ago despite overwhelming evidence to the contrary, or a friend on Twitter claiming wearing a mask to protect their communities from COVID-19 actually interferes with their oxygen access, also despite no evidence. Or perhaps someone you know has passed on a YouTube video about a little conspiracy theory called QAnon, a completely baseless set of theories that have helped propel misinformation about the coronavirus, the election, and more.

Well, whatever the topic, misinformation, conspiracy theories and outright disinformation, they are thriving on our social media platforms in a way that’s completely different from what we saw even during the 2016 election cycle. And indeed, you might have noticed your social media have looked a bit different in the lead up to this year’s election. Twitter has been flagging certain claims about the election as disputed, more than 300,000 tweets during the lead up and immediate aftermath of the election, including some from President Trump.

And in the immediate aftermath, Facebook appended posts, making claims about the vote tallies with notes that votes are still being counted. And in a hearing Tuesday with the Senate Judiciary Committee and Twitter’s CEO Jack Dorsey, Facebook CEO Mark Zuckerberg defended his company’s fact checking as having been successful.

MARK ZUCKERBERG: We built sophisticated systems to protect against election interference that combine artificial intelligence, significant human review, and partnerships with the intelligence community, law enforcement, and other tech platforms. We’ve taken down more than 100 networks of bad actors. We’re trying to coordinate and interfere globally. We’ve established a network of independent fact checkers that covers more than 60 languages. Altogether– and I’m glad that from what we’ve seen so far– our systems performed well.

IRA FLATOW: So where does misinformation come from? And how does it spread so widely if these platforms are taking so many precautions? Here to talk about it is Davey Alba, technology reporter covering this information for the New York Times, and Dr. Joan Donovan, director of Harvard’s Shorenstein Center on media, politics, and public policy. Welcome to Science Friday.

JOAN DONOVAN: Thank you.

DAVEY ALBA: Thanks so much for having us. David help us understand we’ve talked about the kinds of misinformation people might see in the election lead up what kinds of claims have you been finding yourself debunking

DAVEY ALBA: A whole host of different claims, much of it around mail-in ballots because this election is so different from any election we’ve had in the past, so ballots being magically found or lost, vote tallies changing, that was a big one during the election week, widespread voter fraud, dead people voting, votes being stolen by people who were using folks’ maiden names and a lot more.

IRA FLATOW: What is new about the misinformation on the internet in 2020? Because didn’t we see some of this in 2016?

DAVEY ALBA: We did. In 2016 we saw a lot of content that really pushes people’s buttons. In the lead up to 2016 and the immediate aftermath it was all about foreign disinformation. So we heard, obviously, about Russia interfering in our elections and disinformation became a buzzword then. Now the content on social media platforms had various posts that contained false and misleading information look quite the same. But the platforms have gotten a lot better at detecting and removing foreign disinformation when they have technical signals on the back end that show that posts come from a foreign country, that kind of thing.

The problem is a lot of those posts actually do, they do prey on real emotions, and feelings of outrage, and the economic woes of real Americans. So that stuff, when it is coming from Americans in the US and spread conspiracy theories that help them understand their situations, unwittingly, in a lot of cases, that stuff is still up there and that the platforms are less willing to take down.

IRA FLATOW: How does it spread so far and so fast?

DAVEY ALBA: Well, there have been many, we should say, I call them influencers who have track records of spreading misinformation. And those folks have been able to amass audiences over the past few years and have become networked. A lot of these folks play off of each other’s posts, use similar hashtags, coalesce around a specific theme. And they spread it far and wide.

And there are a lot of communities that have formed around these influencers. There are different camps like people who are susceptible to anti-vax content. There’s QAnon, as you mentioned, the set of conspiracy theories about the deep state controlling politics and the global economy. There’s coronavirus truthers that masks don’t work and that the virus, itself, has long been planned. And all of these communities have collided in 2020 to make a very different environment compared to years past.

IRA FLATOW: And Joan, I just rattled off a few big groups of falsehoods people might be exposed to. And Davey just also mentioned some. Are they all related somehow?

JOAN DONOVAN: Yeah, so the way that I think about how people end up encountering misinformation is that any range of searches online, you’re going to get– you put in your keywords and the algorithms are going to tell you basically what’s fresh and relevant. And unfortunately, misinformation tends to be, as David was saying, densely networked and fresh and relevant, that is, it’s of the moment. And so algorithms tend to reward engagement on these very emotionally-charged topics that have all of the things that draw in humans, like they’re scandalous, there’s Hollywood intrigue.

There’s palace intrigue, in the sense that we don’t know who our next political ruler is going to be, right? And so people tend to enter into search, probably, very with a very open attitude as to what they’re going to find. But the way in which our communication system and our media ecosystem online is designed, unfortunately rewards the tactics of media manipulators and dis-informers.

IRA FLATOW: It’s almost as if, sometimes I look at it as if people are trying to live a reality TV show in real life.

JOAN DONOVAN: Yeah, I mean it does have this feeling of like The Truman Show, right where everything is constructed around you in a way that once you– or even The Matrix, right? Once you see through the matrix, you see everything differently. And that’s the way in which I think conspiracy theories provide us with a sense of mastery over the world. It feels a lot less dangerous to go out in public if you think coronavirus is a result of some kind of grand scheme rather than what it is, which is a virus that rather randomly affects any and all types of people.

IRA FLATOW: Is there something special about 2020, this year, this point in our history that makes us, as a society, more susceptible to misinformation?

JOAN DONOVAN: Well, I think we can’t talk about 2020 without understanding what the pandemic has brought us in terms of the way in which we have all been isolated. Most of us are getting online much more frequently and seeking answers where we’re all seeking answers about very similar things. And so that even the key words of coronavirus, COVID-19, we didn’t have those prior to 2020.

But if you look online, the hyper-concentrated attention on those key words means that media manipulators, grifters, hoaxers, scammers are going to flock to it in order to try to get a piece of that attention. And so a medical misinformation conspiracy theories and then also election misinformation all gets bound up in one giant, grotesque hairball. And it becomes really difficult to disentangle.

IRA FLATOW: Do we have any evidence, though, that people actually do anything with this information? Do conspiracy theories on social media actually affect our daily lives, Joan.

JOAN DONOVAN: Yes, so we have a new resource out at media manipulation where we are going to be adding case studies every week to this casebook. And what we try to focus on is a research team is this effect we call the wires to the weeds, that is how does what happens online show up in people’s lived experience, right? How do people in one area think that somehow Antifa and protesters are going to be bused or flown into their area and therefore they need to take up arms and go protect the public square?

That happened in Idaho, that was real. Ways in which people assess the coronavirus, you can see it very clearly in if they believe that masks are going to help them or not. This is an effect of encountering medical misinformation. And then being staunchly opposed to any kind of medical advice.

One thing in particular that I think we really need very swift action on in terms of the platform companies is the delivery of timely, local, relevant, and accurate information. And that is not how these companies designed their tools to work. But we need that now more than ever as people are trying to navigate the risks heading into the holidays here.

IRA FLATOW: I mentioned all the things that Facebook, Twitter and other social media say they’ve been trying. They say they’re really trying their best to slow the spread of misinformation. Davey, how do you react to that?

DAVEY ALBA: Well, I think, again, as I said earlier, the social media platforms are much better at taking down foreign disinformation. Now, you know, these social media platforms are more willing to label posts and call out when a claim is disputed. What we don’t know is how effective those posts are at actually stemming the spread of conspiracies that are, as Joan said, designed to catch fire and spread quickly among the population.

So these baby steps are good. But it took a lot of years of researchers and journalists kind of yelling about the problem and trying to demonstrate and document the consequences of misinformation for them to edge towards taking these measures.

IRA FLATOW: Do you think there’s going to be some legislation? We had those hearings this week with the social communities trying to defend their actions. Do you see anything coming out of this, Davey?

DAVEY ALBA: It’s hard to say. Obviously, we have a new administration stepping in. We don’t know how the Biden administration is going to bring down the regulatory hammer on these companies yet. But the flipside of stronger regulations is you don’t necessarily want your government to be exerting that much control over speech. So, you know, there is a balance to be struck there. And it will take all the different pillars of society kind of watching and scrutinizing this process that will come out of any government regulation.

And we should all be part of the effort to raise our collective media literacy, to look at this problem and think of solutions. It’s going to be a community-based effort, not just relying on the government or relying on the social media companies themselves.

IRA FLATOW: I’m Ira Flatow and this is Science Friday from WNYC Studios. In case you’re just joining us, we’re talking to New York Times reporter Davey Alba and misinformation researcher Joan Donovan about the spread of conspiracy theories online. Joan, do you agree that it’s going to be a community-based effort? And if so, what do online platforms need to do?

JOAN DONOVAN: Well, certainly we can’t leave it up to these companies to decide how they’re going to be regulated because, internally, they are at war. And Davey’s reporting as well as others is really pointing us to this reckoning within tech companies that even the workers themselves are leaking so many different documents and missed opportunities for platform companies to get a handle on this problem.

And so it’s really important that we understand that, internally, even the folks that are building these systems, maybe not in the executive suite, want change. And so it’s up to journalists, researchers, the people that work at these companies to keep the pressure up. I’ve been looking really at some of the older epidemiological literature around secondhand smoke, right? We had to invent this concept of second hand smoke because once we started to see that smoking wasn’t just an individual choice but was actually something that was harming many more people that were encountering people who were smoking, then the concept of secondhand smoke made the legislation possible.

And so what I’m trying to understand as a researcher right now is who pays the true costs of social media? The fact that an entire beat of journalism had to develop on disinformation over the past few years in order to counteract the negligence of these platform companies is really astounding. As well, public health professionals are asking me every day, how do I become an influencer? And that’s just the wrong model.

We actually need social media to deliver timely, local, relevant, and accurate information so that our doctors can do the work of doctoring. And the list goes on for professions that have been paying for social media and paying for the widespread disinformation campaigns that are happening. And if you think even Davey’s recent reporting on election officials, for instance, and the ways in which people have had to come together to counter the messaging of disinformation campaigns, we are facing an enormous problem.

And we really have to get beyond just demonstrating the harms and move into a regulatory schema that deals with misinformation at scale. Because scale is different. More is different than just someone being deluded. And so that’s what we really have to reckon with over the next four years.

IRA FLATOW: Well, it’s been a great conversation and a very sobering one and something that we will all be having in the future months in years as they go by. I want to thank both of you for taking time to talk with us, Davey Alba, technology reporter focusing on disinformation for the New York Times, Dr. Joan Donovan, Director at Harvard University’s Shorenstein Center on Media, Politics, and Public Policy. Thanks again for taking time to be with us today and happy holiday season to you.

DAVEY ALBA: Thank you.

JOAN DONOVAN: Thanks so much.