JOHN DANKOSKY: This is Science Friday. I’m John Dankosky. Ira Flatow is away. 

Later this hour, we’ll be talking meteorology and hearing what volcanoes have to say. But first, it’s not always easy being an ant. You’re small. You’re fragile. You’re easily stepped on. If you leave the nest, you’ve got to worry about hungry birds, mammals, maybe clunky human feet. And if you thought that you’d be safe deep in your big nest with all your friends, think again. There’s something there too, disguised by smell, snacking on hapless workers, and even eating your young. 

Here with that tale of horror and other short subjects in science is Annalee Newitz, a science journalist and author based in San Francisco. Annalee, welcome back to the show. 

ANNALEE NEWITZ: Hey. Thanks for having me. 

JOHN DANKOSKY: So what is stalking these poor ants? It’s a terrible story. 

ANNALEE NEWITZ: [LAUGHS] It is a terrible tale of terror. So this comes from an article by a couple of scientists– Wendy Moore and Andrea DiGiulio. And they were researching a typical ground beetle called Ozaena lemoulti. And it turns out that this beetle has a really unusual relationship with ants. 

It’s not unusual for beetles to hunt ants. And you know, a lot of beetles are predators. But these beetles live with ants in oak trees throughout their entire lifecycle. So the females lay their eggs in the ant nest. They disguise themselves by covering their bodies in smells that the ants recognize as friendly smells, because ants are blind, and they kind of navigate the world through smell. And as these beetles grow older, as they become larvae and then adults, they feed on the ants by piercing their abdomens and sucking the fluids out. 

JOHN DANKOSKY: Ugh. It sounds terrible. So hold on. How did they discover this particular beetle doing this particular thing? 

ANNALEE NEWITZ: So these are scientists who were looking at beetles who are living kind of in concert with ants. And they discovered that this particular beetle, unlike others, wasn’t just sort of chomping down on the whole ant. They were looking at the inside of their guts and sort of looking at what was in their stomachs. And they found that these particular beetles had this odd paste in their stomachs, which showed them that they had evolved specifically to suck the guts out of ants. 

And also this gave them more clues about how much of the beetles’ life cycle was spent among the ants. And so they spent time in Arizona, right at the border with Mexico, observing these beetles in their natural habitat in oak trees with ants and came up with a lot of data from that and then brought one of the beetle larvae home with them to observe it eating ants up close. 

JOHN DANKOSKY: And the ants don’t just get together and attack these beetles because they smell really good? 

ANNALEE NEWITZ: That’s right, because they have the ability to kind of cover their bodies with a scent that the ants recognize as being kind of a home scent or a friend scent. Ants navigate the world by smell. And there are a lot of insects– beetles, mites, other creatures– that sneak into ant nests all the time and disguise themselves with smell. It’s kind of like putting on an invisibility cloak. And it’s a great way to protect yourself if you’re a beetle, because ant nests are really safe. 

JOHN DANKOSKY: All right. So let’s move on from that somewhat disturbing story to a story about Twitter, which can in its own way be disturbing. It turns out that our tweets are kind of predictable. Like I guess this isn’t news to some of us. But what are we learning about the predictability of our tweets? 

ANNALEE NEWITZ: This is a really interesting story. So a couple of researchers, data scientists and psychologists, looked at the Twitter streams of about 950 people. And what they did was they took– I’ll give you an example. Like, say your Bob, and you have a Twitter stream. You’ve tweeted a bunch of stuff. The question they had was, how easy would it be to predict your next tweet? So they take all of Bob’s tweets, they feed them into the algorithm that they’re using, which is a predictive algorithm, and they find out that by looking at Bob’s past tweets, they can predict with 53% accuracy what the first word of his next tweet will be. 

But that’s when things get weird. Because then they started looking at all of Bob’s closest friends on Twitter. So they said, all right, who are the 15 people that Bob tweets at the most? We’re going to call those Bob’s friends. And if they look at the combined tweets of Bob’s 15 friends, they can then predict the first word of Bob’s next tweet with 57% accuracy. So it’s actually like your friends tweets are more predictive of what you’ll say next than your own tweets are. 

And then it gets even weirder. Because that means, conclude the researchers, that we can even predict what you might tweet if you’re not on Twitter. 

JOHN DANKOSKY: We can predict what you might tweet? If you’re not even on Twitter, you might be able to tell just by looking at what your friends tweet? 

ANNALEE NEWITZ: That’s right. Because when Bob signed up for Twitter– we’re just going to keep picking on Bob. 

JOHN DANKOSKY: Poor Bob. 

ANNALEE NEWITZ: When Bob signed up for Twitter, Twitter said, would you like to upload your contacts list? And this is something that lots of social platforms do. It’s not just Twitter. You know, Facebook does it. WhatsApp does it. 

And so Bob uploads his contacts. And Alice is in his contacts. And Alice doesn’t use Twitter. But she now is known to Twitter because they have her contact information. And maybe 14 other people also have Alice’s contact information. So now even though Alice is not a member of Twitter and has never tweeted in her life, those 15 people’s tweet streams can predict what she might say next. 

And for now, you know, we’re only predicting kind of unimportant words that people might say. But the researchers say that soon this kind of algorithm could predict important words, like how you feel about a presidential candidate, or how you feel about a brand of soap. And that’s the kind of information that political parties want, institutions want, advertisers want. And that’s why there’s so much pressure to be coming up with algorithms that are even better at doing this. 

JOHN DANKOSKY: All right. Actually, that’s even scarier than the ant story now. Now I’m worried about this too. All right, we’re going to get to one more story here. You found something interesting about what happens in our brain when we’re enjoying music. So tell us– what’s happening? 

ANNALEE NEWITZ: So this comes from a study by neuroscientists who were actually just interested in whether our enjoyment of music is caused by dopamine, which is a neurotransmitter that’s associated with lots of different kinds of good feelings and a sense of accomplishment and things like that. So what they did was they gave test subjects a common drug that’s actually used a lot in Parkinson’s disease. It’s called levodopa. And it’s a precursor for dopamine that just allows dopamine to circulate more in your brain and be taken up by nerves in your brain. 

And so what they found was that people who were under the influence of levodopa were likely to pay more money for music that they enjoyed than people who were not on levodopa and in people who had actually had dopamine dampened, dopamine transmission dampened in their brains. So basically, we have a drug that makes you willing to pay more money for music. 

JOHN DANKOSKY: Very quickly– but only music that you like or any music? 

ANNALEE NEWITZ: Typically, this was music that people liked. So you couldn’t get, you know, someone who hated country music to suddenly be willing to spend tons of money on country music. 

JOHN DANKOSKY: OK. That’s the code they still need to crack, I suppose, in the music business. That’s all the time we have. I want to thank our guest once again, Annalee Newitz, a science journalist and author in San Francisco. Thanks so much for joining us. I appreciate it. 

ANNALEE NEWITZ: Thanks for having me.

Copyright © 2019 Science Friday Initiative. All rights reserved. Science Friday transcripts are produced on a tight deadline by 3Play Media. Fidelity to the original aired/published audio or video file might vary, and text might be updated or amended in the future. For the authoritative record of Science Friday’s programming, please visit the original aired/published recording. For terms of use and more information, visit our policies pages at http://www.sciencefriday.com/about/policies/