RACHEL MARTIN, HOST:
Twitter has tried a whole lot of things to curb misinformation. It has slapped warning labels on false claims about election fraud and the coronavirus pandemic. It's put links to credible news outlets debunking those claims. It's even banned users who break its rules. But those moves have only gone so far. So now Twitter's asking its users for help. NPR tech correspondent Shannon Bond has more.
SHANNON BOND, BYLINE: The pilot project Birdwatch is an experiment to find out if people who are on Twitter all day can do something the company can't.
KEITH COLEMAN: We know that not everyone trusts a single tech company or any singular institution to make the decisions about what context to add and when.
BOND: Keith Coleman is Twitter's head of product. He says Birdwatch aims to use the power of the social network's 192 million daily users to fact-check each other's tweets.
COLEMAN: The idea is we can allow the community together to come to a consensus on a case where that context really should be shown directly on the tweet.
BOND: The way Birdwatch works is you can write a note about a tweet saying, I think this is incorrect and here's why. Then other Birdwatch users rate your note. And if a lot of people say your notes are helpful, you build up a reputation. Your notes and ratings get more weight. So think crowdsourcing combined with consensus. Coleman says eventually you'll be able to see these notes right on tweets. The hope is Birdwatch users can flag and even correct misleading information more quickly than Twitter.
COLEMAN: The idea is that people would be able to come away from Twitter better informed.
BOND: Birdwatch is still in its early stages, with just a thousand participants on a special site. And looking through it right now, some people are using it to fact-check, like linking to research showing masks help prevent the spread of COVID-19 on a tweet claiming otherwise. But there's also a lot of opinion, the kind of partisan bickering you can find on almost any contentious Twitter thread. Madelyn Webb is a researcher at First Draft, a nonprofit fighting misinformation. She's been digging into how people are starting to use Birdwatch.
MADELYN WEBB: It's sort of just replicating what we see on Twitter - things go viral, everybody wants to talk about them, and then the rest of it sort of falls to the wayside.
BOND: So how do you keep Birdwatch from recreating the problems Twitter has, like becoming another channel for misinformation? Molly White as a longtime Wikipedia editor, so she knows a lot about the power and risks of crowdsourcing.
MOLLY WHITE: What if there's an accurate tweet and someone fact-checks that, they say something inaccurate? Then who fact-checks that and then who fact-checks that? You know, it's like you just keep going deeper, right?
BOND: She's also wary that people could use Birdwatch for harassment and abuse. Over at Twitter, Coleman says the company worries about that, too.
COLEMAN: But we believe that we can design it to work differently than Twitter in a way that can handle that.
BOND: He says it's about changing the incentives. Twitter's platform rewards users who gain a lot of followers and get a lot of interactions, replies and likes. You see their tweets more. Birdwatch users are encouraged to build up reputations as being helpful and credible. Reputable users get more prominence.
COLEMAN: Those dynamics around what gets elevated, you know, they're different from Twitter and we think that will create a very different result. We think that will result in genuinely helpful information being elevated.
BOND: But for a Birdwatch to succeed, it will have to overcome an uncomfortable truth about Twitter and, really, all social media. Tiffany Li is a law professor at Boston University who studies technology.
TIFFANY LI: People on Twitter do not agree on what truth is. They do not agree on what, you know, real news is, and that's a problem. If they don't agree on what the truth is, they're going to have different opinions on whether or not a tweet or a post is truthful.
BOND: The question is, can enough Twitter users agree on anything to make the Birdwatch experiment work?
Shannon Bond, NPR News.
(SOUNDBITE OF MARLEY CARROLL'S "FIREFLIES")
NPR transcripts are created on a rush deadline by an NPR contractor. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.