英语访谈节目:推特机器人如何再用冷战方法散布政治分歧

John Yang: All the attention on the Nunes memo goes back two weeks, even more. And there are questions about whether some of that attention, at least on social media, has been coming from real users, or whether it's being ginned up artificially. William Brangham has more.

William Brangham: In the days leading up to the release of that controversial Nunes memo, there was a hashtag on Twitter called #releasethememo that became hugely popular on Twitter.It was used by conservatives and allies of the president who hoped the memo's release would undercut special counsel Robert Mueller's investigation. But a series of recent reports indicate that some of those Twitter accounts promoting #releasethememo were fake, and that some of them were linked to Russian interests. Thomas Rid is at the Johns Hopkins School for Advanced International Studies, and he's been examining how Twitter and other online platforms can be used by these so-called fake bots to spread information online. Welcome back to the "NewsHour."

Thomas Rid: Hi.

William Brangham: So, as I was describing, this hashtag on Twitter called #releasethememo became hugely popular. Lots of people, real live human beings, were using the hashtag because they wanted the memo to come out, but allegedly some of these hashtag were being promoted by these bots. Can you explain what we know about these bots and what they have — how they might have been connected to Russia?

Thomas Rid: So, a bot is an automated Twitter account that doesn't have a real human being behind it, but just a program. And, indeed, it seems that some automated accounts and some fake real human beings that aren't actually who they pretend to be, use the #releasethememo hashtag — a hashtag is a way to make a Twitter — a tweet circulate more — used that hashtag to give it some lift. Really, it's very difficult to distinguish between these real conservative activists and politicians.

William Brangham: Real people.

Thomas Rid: Real people — and fake personas that may be linked to Russian interests. And that is a problem in itself.

William Brangham: So, explain how this would actually work. For those who don't use Twitter or understand the mechanics of it, basically, I choose who I want to follow, some 50,000, 100,000 different voices, and I get messages from those people directly in their Twitter as they tweet out messages. If I don't follow these bots, why do I care what they are or are not saying?

Thomas Rid: So, indeed, a lot of users on Twitter think they don't follow any bots — and they may actually not follow any artificial accounts that are not human — and therefore think that doesn't concern me, this problem. But that's not how this works. For example, imagine you see a tweet, a post on Twitter, and that has been retweeted or liked thousands, tens of thousands of times. You never check whether these retweets, which make something appear very important and viral, whether they actually are real or not. So it's possible to give actual messages, like a hashtag, to give it more lift and more weight through automation and through automated abuse.

William Brangham: Now, it's like fake applause for a comedian. He plants people in there to clap or laugh extra hard, and it makes him seem funnier, at least to the audience, than he really might be.

Thomas Rid: Absolutely, except it's not just the specific line that the comedian talks about that can be amped up and amplified. It can be other messages that sort of filter up very slowly out of the vastness of all the Twitter posts and Twitter messages. So, Twitter, in fact, is a huge part of this problem here

William Brangham: Now, Kellyanne Conway and the White House have said this wasn't about a hashtag on Twitter. This was a vote in the House Intelligence Committee. They're the ones who released this memo. What is the evidence that exists, if any, that this was some coordinated campaign by non-American actors?

Thomas Rid: So, Russia — and in the Cold War, it was Soviet, of course, including East German and others — information operations, active measures was the term of art — always exploit existing conflicts, they exploit existing cracks in our political system, and then exacerbate these cracks and drive wedges into them. They don't create new conflict, which is a really important insight. It's not a distinction between fake and real, because they take a real conflict and make it worse.

William Brangham: So, I understand there was also some question as to whether or not the Black Lives Matter, the Confederate rallies, that those also were objects of attack by these Russian bots.

Thomas Rid: Very much so, but not just bots, just intelligence operators or sometimes contractors. And, again, there's a long history. We know from the Cold War that Soviet operators had a long history of exacerbating racial tensions. They would sometimes pose as the Ku Klux Klan, at other times pose as African-American activists. There's a long history of this.

William Brangham: I know you have been very critical of Twitter and other social media platforms, arguing that they could do more to stem this kind of fake messaging. What would you like them to do?

Thomas Rid: For example, Twitter gives the same level of privacy protection to you and possibly your children that it gives to Russian bots and fake accounts. That is just not OK. So, Twitter refuses to make the distinction and refuses to point out abuse, the fake bot accounts to its own users. It's impossible to, for example, opt out of bot traffic, to opt out of messages from bots and retweets from bots.

William Brangham: And that's something you argue that Twitter could easily do that to identify the fake from the real.

Thomas Rid: Absolutely. Technically, that is not a big problem. Indeed, Twitter claims that they can recognize bots automatically. But at the same time, the bots make Twitter appear larger and more engaging than it actually is. Twitter has never made money, and that's a way for them to appear bigger than they actually are.

William Brangham: Thomas Rid, thank you very much.