Audio Deepfakes and the End of Trust

Season 1 • Episode 3

The media is plenty freaked out about “deepfakes”: Computer-generated videos of famous people saying things they never actually said. But only the video is faked; the audio parts, the voices of those fake celebrities, were supplied by human impersonators. But now, software exists to mimic anyone’s voice, opening a Pandora’s Box of fraud, deception, and what one expert calls “the end of trust.” Fortunately, a new coalition of 60 news organizations and software companies think they have a way to shut down the nightmare before it begins. 

Guests: Ragavan Thurairatnam, Dessa. Nina Schick, author and deepfakes expert. Joan Donavan, Harvard Kennedy School. Charlie Choi, CEO of Lovo. Dana Rao, chief counsel, Adobe.

Episode transcript

Intro

Theme begins.

Deepfakes are the latest in computer-generated imagery: they’re videos of people doing and saying things that they never actually did or said. Like, there’s a video of Obama saying,

“Obama:” “President Trump is a total and complete dipshit.”

But what’s weird is the voices in those videos are still done by human beings. Impressionists. Impersonators. The technology to simulate their voices still wasn’t good enough to fool anyone. 

Until now.

I’m David Pogue—And this…is “Unsung Science.”

Season 1, Episode 3: Voice Deepfakes

Every fall, Adobe hosts a conference called Adobe Max. It’s a chance for the engineers to strut their stuff, show what they’ve been working on, and make announcements to a captive audience of customers and press. 

The conference focuses, of course, on creative software—for photos, videos, music, and so on—because that’s Adobe’s thing, right? They make Photoshop for editing photos, Premiere for editing videos, and so on.

One session every year is called Adobe Max Sneak. Here’s how Adobe describes this presentation:

Faux 1: The Max Sneaks session invites our engineers out of the lab and onto the stage. Many Sneaks from previous years have later been incorporated into our products. 

In 2016, the Sneak session featured the usual sorts of Adobe experiments. There was a prototype app that replaces the sky in a photo with a different sky, with one click

There was an app that could adjust the colors in a bunch of photos to match the color scheme of an existing document. 

And then…there was Project Voco, which was described as Photoshop for voice.

Speaker Let’s hear from Zeyu you about Photoshop voiceovers. Please welcome to the stage… Zeyu. (applause)

Zeyu: Hello, everyone! Let’s do something to human speech. I have obtained this piece of audio where there’s Michael Key talking to Peele about his feeling after getting nominated. 

He’s referring to Key and Peele, the comedy duo. 

I forgot to mention that Jordan Peele, half of that duo, was sitting right there on the stage. He’d been hired as the cohost for this event. That’s the Jordan Peele who went on to write and direct movies like “Get Out” and “Us.” 

Anyway, the Adobe researcher now played a recording of Peele’s partner, Keegan-Michael Key. In the clip, Key is describing his reaction at learning that he’d been nominated for an Emmy.

Key I jumped on the bed and— and I kissed my dogs and my wife, in that order. (laughter)

Zeyu: So how about we mess with who he actually kissed? Project Voco allows you to edit speech in text, so let’s bring it up. 

The Voco window shows audio waveforms across the top—and lined up beneath them, the corresponding words. 

Zeyu And when we play back, the text and the audio should play back at the same time. So let’s try that.

Key And I kissed my dogs and my wife. 

Zeyu OK, so what do we do? Easily, copy paste. Let’s do it. 

Using his cursor, Zeyu copied and pasted the word “wife” to make it come earlier in the sentence—

Key And I kissed my wife and my wife. (crowd)

Zeyu: Oops!

—and then typed over the second occurrence of the word “wife” with the word “dogs.”

Zeyu We can just type the word “dogs” here. 

Crowd No, no! 

Key And I kissed my wife and my dogs. 

Crowd Woooo!!!! 

But that was just rearranging recorded words. Now came the really nutty stuff.

Zeyu Wait, here’s more, here’s more; we can actually type something that’s not here, so. 

Using his keyboard, he deleted the word “wife,” and he typed the word Jordan, as in Jordan Peele.

Zeyu: And here we go:

Key And I kissed Jordan and my dogs. (crowd reacts)

At this point, Jordan Peele leaps out of his chair in mock horror. He’s stomping across the stage, like, “I’m outta here.”

Peele You—you a witch! You a demon!

Zeyu I’m magic. We’re not just going to do with words, we can actually type small phrases. So we do “three times.” 

Speaker Oooooh. 

Zeyu: And, playback!  

Key And I kissed Jordan three times. 

Crowd Ohhhhh!! (Crowd cheering)

So yeah. They had fed 20 minutes’ of Key’s voice recordings into Project Voco, and now, just by typing, they could make him say things that he had never actually said. And there was absolutely no way to tell that it wasn’t real. 

The crowd seemed to love it. The only person who seemed at all troubled—was Jordan Peele.

Peele I, I’m, I’m blown away. I can’t believe that’s possible. You just type it in, and it interprets the person’s voice. If this technology gets into the wrong hands… (laughter)

But Zeyu Jin was quick with reassurance.

Zeyu Don’t worry. We actually have researched how to, like, prevent forgery. We have, like think about like a watermarking detection. 

Later, the Adobe blog described the event like this. 

Faux 2: Project VoCo, allows you to change words in a voiceover simply by typing new words. As always, we’d love your feedback.

And boy, did Adobe get feedback. From the BBC:

Faux 3: It seems that Adobe’s programmers ignored the ethical dilemmas brought up by its potential misuse. 

From the CreativeBloq blog:

Faux 4: This raises ethical alarm bells about the ability to change facts after the event. 

From Affinity Magazine:

Faux 5: The ethical issues associated with its misuse are endless. 

Adobe soon began issuing this statement to reporters:

Faux 6: Project Voco was shown at Adobe Max as a first look of forward looking technologies from Adobe’s research labs, and may or may not be released as a product or product feature.

Well—surprise, surprise: It was not released as a product or product feature. In fact, it was never heard from again. 

Now, meanwhile, in the rest of the world, the tech media was abuzz with stories about the rise of deepfakes

>>TV & NPR audio clips about “deepfakes”

Someone on Reddit first coined that term—deepfakes—to describe videos where the computer has replaced one person’s face with another’s. In the beginning, most deepfakes were made by amateurs grafting popular actresses’ faces into porn videos. 

But there was also a hilarious-slash-creepy trend of people putting Nicolas Cage’s face onto other actors in famous movie scenes. 

By 2018, deepfakes had gotten good enough that one video of President Obama was convincing in every way—except for the words coming out of his mouth: 

Peele: They could have me say things like, I don’t know, “President Trump is a total and complete dipshit.”

Now, you see, I would never say these things. But someone else would. 

Buzzfeed had made that video as a sort of public-service announcement about video deepfakes.

Oh, it made the point, alright. But 8.5 million views later, hardly anyone has commented on its one glaring flaw: The computer algorithm did a great job of generating the video of Obama. But they had to use an Obama impersonator—a human being—to do the voice. Guess who they got to do the impression?

Jordan Peele. 

Yup—same guy who’d been on the stage two years earlier witnessing the unveiling of Project Voco. 

Now, usually, audio technology always comes before video. There was radio before there was TV. There were cassette tapes before there were videotapes. There was streaming audio before there was streaming video.

But for some reason, in deepfakes, video came first. Audio deepfakes came along only later—and they took their time.

Let me play you the state of the art in voice deepfakes as of 2017. This is supposed to sound like Donald Trump:

Trump: I am not a robot. My intonation is always different.

By the beginning of 2019, the state of the Trump deepfake had reached this level:

Trump: With this technology, it can make me say anything. Such as the following: / Barack Obama is a wonderful man. Do you think this sounds like me? We are working hard to improve these results. That is all for now. See you later, alligator. 

Yeah. Maybe much later, alligator.

But then, later in 2019, the world got a load of Fake Joe Rogan.

Rogan: It’s me, Joe Rogan. Friends, I’ve got something new to tell all of you. I’ve decided to sponsor a hockey team made up entirely of chimps. Chimps are just superior athletes. And these chimps have been working out hard. I’ve got them on a strict diet of bone broth and elk meat. See you on the ice, folks.

That is not, in fact, the voice of comedian and top podcaster Joe Rogan. That…is an audio deepfake. It’s not Joe Rogan; it’s Faux Rogan. Let’s meet the guy who made it.

David For my pronunciation pleasure, Ragavan, will you pronounce your name so I can get it right? 

Ragavan Yeah, it’s uh— it’s yeah, you could say Ragavan. The last name, even I don’t attempt to pronounce.

DAVID: WHAT?

RAGAVAN: It’s Thurairatnam, but I’m pretty sure I’m saying it wrong. 

Ragavan is the cofounder of Dessa, a Toronto company specializing in machine learning. 

David And what is Dessa’s actual business? What –what did you found it to do?

Ragavan We kind of started looking at banks as potential customers. [00:08:47] And it kind of we— we ended up making like AI software for these sort of big, big, boring companies — But we also want to do crazy stuff because, like, it’s— we just saw that, you know, this this technology can do so many things. And we —we really wanted to show the world what it could do and also just have some fun. 

David So the RealTalk project was one of these side —side hustles. 

Ragavan That’s right. Yeah. The Real Talk Project was one of those side projects. 

Dessa’s dive into AI speech synthesis began at a company dinner in the summer of 2018. 

Ragavan I asked the team like, what can we do to, like, really show people, like deep learning can do amazing things, and also get a lot of attention.  

So one of the engineers who ended up working on the project, his name is Hashim Kadem. He said, like, you know, “there’s this podcast that looks like the most popular podcast in the world, Joe Rogan’s podcast, like, if we could get on there, get noticed on there, something, that could be that could be really good.” 

And that was the sort of seed of it. 

The Dessa team figured they had plenty of source material to use for its Rogan voice clone. After all, Joe Rogan has made over 1600 episodes of his podcast—and they tend to be long. Sometimes five hours long.

Ragavan On the surface it’s like, “oh, there’s hours of podcast recording. This should be easy, right?” 

But Joe Rogan, it’s just —like he’s just crazy. He puts his mouth like right on the microphone. It has all these weird things in it. And also, it’s like a conversation, which is just completely different. It’s like —one person talks, the other person talks in the middle of him talking, there’s laughing, there’s like coffee drinking and, you know, all sorts of things. And— and that makes it a lot harder. 

So what the team ended up doing was, they ended up using just his ad reads. So like, you know, whenever he’s reading an ad for his podcast—because we knew it’s just him, you know, he’s not going to be doing anything weird. It’s a lot easier. 

<<<Clip of Joe Rogan reading an ad >>>

Meanwhile, the team was also working through the tedious process of producing a perfect typed transcript of everything in those Rogan recordings.

Now, the next part of the story requires a gentle understanding of artificial intelligence, machine learning, and deep learning. It’s technical, and I debated just cutting this whole section. But hey—you’ve put on a podcast to learn something, right?

Ragavan? Take it away.

David Are you able to do a layman-friendly distinction between deep learning, and machine learning? 

Ragavan Yeah, let’s try that.  Normally, software, you have to write all the rules for it. So, like, you know, if you think of an app, like you have to say exactly, like “when the user does this, I want this to happen.” 

But with machine learning, what we do is, we, we make software kind of— learn how to do things just by showing it data. 

So, for example, let’s say I wanted to recognize something in an image. In machine learning, I would write by hand, like these sort of mathematical things that say, like, “oh, look for straight lines, you know, count how many straight lines there are. Count how many, you know, blobs of yellow there are or red there are.”

OK, I got it, sort of. Then what’s deep learning?

With deep learning, we take —take it a bit further, and it kind of learns directly from the data. With deep learning, it’s just like, “give me the data and give me the answer; I will figure out the rest. I will learn everything in order to make this happen.” 

I think that’s what’s really powerful about deep learning. And that’s— that’s one of the reasons why, you know, in the past few years, we’ve seen so many crazy things come out of it. 

I’m beating this dead horse about how hard is to create a deepfake voice to emphasize… how hard is to create a deepfake voice! I mean, the early incarnations of the Faux Rogan voice were not that convincing. Here’s an example:

Rogan: Rogan on AlexNet (RealTalk early clip)

Yeah, OK, it’s…something. But even after further work and hand-tweaking, there were still weird gaps and unnatural emphasis. Like this:

Rogan: (Anyone else clip) 

But eventually, the Rogan voice got really good. At fakeJoeRogan.com, for example, you can take a little quiz. You can listen to sentences, and try to figure out if they’re Joe Rogan…or Faux Rogan. 

I’ll play three examples, and you can test your deepfake radar. Ready? First one:

Rogan: What was the person thinking when they discovered cow’s milk was fine for human consumption? And why did they do it in the first place?

Real or fake? Remember your vote. I’ll give you the answers in a second. OK, second one:

Rogan: Some of you just need to improve the quality of your existence on earth. You gotta do the right things.

And finally, example number 3:

Rogan: Fantastic old-world craftsmanship that you just don’t see any more.

OK, remember your answers. After the break—we’ll see how you did.

BREAK 

Welcome back. Let’s see how you did with your deepfake detection skills. 

The first sentence—

Rogan: What was the person thinking when they discovered cow’s milk was OK for human consumption? And why did they do it in the first place?

That one’s fake.

Rogan: Some of you just need to improve the quality of your existence on earth. You’ve got to do the right things. 

Fake again. And finally, example 3:

Rogan: Fantastic old-world craftsmanship that you just don’t see any more.

That’s an actual recording of Joe Rogan. 

So how’d you do? Pretty soon, it’s going to matter. 

Nina By the end of the decade, you’re looking at a future where one Youtuber with limited resources or skills can kind of produce something that’s better than what the best Hollywood studio can produce today for millions of dollars and with teams of special effects artists. Don’t you worry, that is coming!

This is Nina Schick, author of a book called Deep Fakes: The Coming Infocolypse. Like, “apocalypse” but with “info.” “Infocolypse.”

Nina When I first started to come to deepfakes, you know, it was as they were emerging at the end of 2017 in the form of nonconsensual pornography on Reddit. And I immediately realized that deepfakes could become the most powerful weapon of political disinformation known to humanity. 

Nina may be one of the world’s most informed experts on why audio deepfakes are dangerous.

Nina Number one, you can fake media of anyone saying or doing anything. So you can imagine how, for instance, if you take the context of the United States after the George Floyd video came out, imagine there was a leaked recording of Donald Trump uttering a racial slur. You can see how that leaked audiotape could, in that incendiary kind of political environment, really kick off something far more dangerous. 

But I should note that this also has a very real risk to businesses. Imagine a business leader is caught on tape saying something that they didn’t actually say. It could be potentially devastating. 

And, of course, the opportunities for scammers are delicious. 

Nina Ultimately, it’s something that can affect every individual, right? One of the classic frauds that is perpetrated against millions of us every day worldwide, is the desperate phone call from a loved one. Right? “Dad, I’ve been an accident. I need money now. I’m in jail.”

Now, imagine fraudsters can use AI to scrape social media to find a video of your son, your wife, your daughter, and then use that AI basically emulate their voice with just a few seconds of training data— and now you get the call and it’s literally your son. /It is absolutely terrifying, to say the least, that this technology can be deployed by malicious actors without control. 

Anyway, deepfakes purporting to show people saying things they never said is only half the problem. The other half is the opposite situation—people blaming deepfakes for things they actually did say!

David I remember that one of Trump’s first responses to the “grab them by the pussy” video was, “I never said that! Software created that.”

Joan That kind of reaction is called the Liar’s Dividend, which is that people can come out and say, “well, I didn’t do that. I didn’t say that —that wasn’t me.” 

Meet Joan Donavan, research director at Harvard’s Kennedy School Shorenstein Center on Media, Politics and Public Policy. 

David And that fits on a business card?

Joan Hey, when you’re me, you don’t want anyone to have your email or your phone number. I don’t even have business cards, I don’t want people to know how to get in touch. 

She’s spent a lot of time studying misinformation. And she says that the antidote for the liar’s dividend—is other people as witnesses.

Joan You don’t build a court case based on a single shred of evidence. Everything adds up. Right? We have to kind of build or weave a story here. 

And then also if it is an interaction that is being that is being faked, like, is there a way to legitimate those claims, just as we would as any good journalists would, you know, verify.  But it’s going to require people talking to people to make sense of the thing. 

Now, Dessa, the company that created Faux Rogan, did it, as they say, to get attention. And they got it. The whole company was soon thereafter bought by Square, the digital payments company.

But why did Adobe do it? Why did they make Project Voco in the first place? 

It wasn’t to torpedo our public trust in anything anybody ever says again. Here’s what Adobe’s blog post said:

Faux 7: When recording voiceovers, dialogue, and narration, wouldn’t you love the option to edit or insert a few words without the hassle of recreating the recording environment or bringing the voiceover artist in for another session? 

Voco was created to make life easier for creative people. To fix stumbles in podcasts, audiobooks, and narration. To clean up dialogue in movies, TV shows, and games, when you need to edit lines after the actors are no longer available. To dub movies into other languages with the original actor’s voice.

Here’s Nina Schick again:

Nina You can see how this is going to basically change the future of the movies, change the future of advertising, I mean, change entire industries.

But another really compelling example, is using synthetic voice to give those people who’ve lost the ability to speak, for instance, through a neurodegenerative disease or a stroke, being able to give them their voice back, literally give them their voice back. And there’s already a team of researchers working on this.

And that’s why there are now a bunch of companies that can turn your voice into a deepfake—a voice clone—so that you can type whatever you want to have read aloud in your voice. 

To make a voice clone, you need to feed the machine-learning algorithm a lot of clean audio. You’re usually asked to read 20 or 50 sentences into the mic. 

DP reading: Sentence number 4. The rainbow is composed of many bands of white light.  

That’s partly to teach the AI—and partly to prevent you from cloning the voice of somebody else without their awareness. You’d have to put a gun to their head, sit them down, and make them read those exact sentences.

So: How good is the result? I tried all of the voice-cloning services I could find.

Here’s a voice I generated for free at site called Resemble.ai:

Resemble: Hello, and welcome to the brilliant new podcast called Unsung Science. I’m David Pogue. Or not.

Wow…Well, Resemble does offer sliders that let you change the pitch, emphasis, and emotion of each word. That “or not” at the end really sounded wrong–

Resemble: Or not.

—so I’m going to make the pitch lower, and change the emotion to annoyed. 

Resemble: Or not.

Much better! Or not. 

Well, let’s see if I could use it to pull off the phone scam that Nina Schick described:

Resemble: Hi Dad, it’s David, as you can obviously tell by the sound of my voice. I’ve been an accident. I need money now. I’m in jail. Can you send me some money right away? 

Yeah, probably not.

Well, how about its competitor, ReplicaStudios.com?

Replica: Hi Dad, it’s me again. David. I have some bad news. I’ve been brutally mugged in the streets of Paris! I need you to send me money. Lots of money. Please please please.

Nope. Not sold. Without a lot of hand work by engineers, the state of the art is just lame.

Now, with hand work by engineers, the state of the art is really good. This is the David Pogue voice clone made for me by a company called Lovo.ai:

DP Lovo: Now I’m in business. This fake Pogue is much more convincing than those free ones. 

To get something that good, I had to read 20 minutes of text. And if I were an actual customer, I would have had to pay a thousand dollars. 

You know the voices you’ve been hearing in this episode, reading statements by Adobe, and quotes from various news outlets? They’re all AI voices generated by Lovo. 

Gotcha! Yeah—I like my podcasts with a twist. 

Charlie So a lot of the AI systems out there, if you feed it in gold, it will output gold. But if you feed it in garbage, it will output garbage. 

Meet Charlie Choi. He’s the CEO of Lovo, speaking to me from Korea.

David I tried a bunch of the free voice cloning services and they were not good. Why is it that you can make ones that could actually fool someone and they can’t? 

Charlie We have a team of data scientists who, after receiving the recording data, we go in and really try to understand if this person has spoken every single word. And we try to annotate every single emphasis or maybe breathing patterns or laughs, so that the AI voice sounds more natural and more human. And for us, we can even simulate stuttering or, all of these imperfect artifacts which make human voice so real. Because humans aren’t perfect. 

Charlie We’re also teaching it where the emphasis goes in, or which part of it is a laugh or which part of it is a sigh. We’re also feeding it, for example, pitch information, so that the model learns how to change around the pitch. 

By the way: Remember how Adobe’s Project Voco was meant to make it easier to edit podcasts and audiobooks? Well—that idea was too good to stay down. Today, you can have that freedom by paying for a service called Descript.com. It’s a suite of tools for podcasters to make it easier to edit recordings. Here’s their ad:

Ad: Meet Descript. It’s a powerful new tool that makes editing easy. So easy that you’ll want to edit videos. 

Girl: Nice!

And if you’re willing to pay $24 a month, you get this:

Ad: Get this. Descript can turn your text back into audio. It’s called Overdub. Just type what you meant to say right into Descript. 

Wait, what? Isn’t that exactly what Project Voco was supposed to do—five years ago?

I tried it out. (Open parenthesis: Descript and the other companies mentioned here didn’t pay me to talk about them; most of ‘em didn’t even know I was doing this. Close paren.)

First, I had to teach Descript my voice by reading 15 minutes’ worth of prepared text:

DP: “The penguins stay when all other creatures have fled, because each guards a treasure.”

…and then, 24 hours later, Descript was ready to do the Project Voco thing. Let’s recreate the same Key and Peele joke that Adobe used, but using my own voice. Here’s what I actually recorded:

DP: I jumped on the bed and— and I kissed my dogs and my wife in that order. 

And then, I edited the sentence just the way the Adobe guy did onstage, to produce this hilarious result:

Faux DP: I jumped on the bed and I kissed Jordan three times. 

OK, that’s pretty amazing.

Voco and Descript are meant to fix a word or two in a legitimate recording. You can’t use them to generate a whole paragraph, or a whole speech.

That is a bigger challenge, and that’s the purpose of services like Lovo—to make a full-scale voice clone that can say anything of any length and sound convincingly human. Right now, they take a lot of work and a lot of money. But Harvard’s Joan Donovan says that technology will march on soon enough.

Donovan: As deep fakes require fewer and fewer images of people, and audio fakes require fewer and fewer sound bites, it’s pushing us into a future of forgery that is going to it’s– it’s going to be confusing for a while. 

So—is that it? Society is doomed? Nobody will ever be able to trust any photo, video, or audio clip again? 

Well—maybe not. It may be that you already know about the solution to the deepfakes problem—you heard it described 20 minutes ago.

Remember Adobe’s 2015 demo of Project Voco? In that session, the presenter promised that the company was also working on fraud-detection technology, so we’d know the difference between real and phony recordings. Remember?

Zeyu Don’t worry. We have, like think about like a watermarking detection. 

Well, Adobe hasn’t forgotten.

Dana: One of the early experiments we— we were working on was something we called Project VoCo which is a voice editing, synthesizing software. But we actually ended up deciding not to release it yet, because we actually didn’t know how to protect it. 

This is Dana Rao, who’s Adobe’s chief counsel. Ever since that Voco demo, he and Adobe’s engineers have been trying to figure out how to prevent a deepfake-ageddon.

Or an Infocalypse. 

Dana: I was talking to our chief product officer, I said, you know, we’re probably at the point where, where this is going to be really hard, as we said, tell fact from fiction. In a world where you don’t believe anything anymore, there are two big problems. One is, you believe a lie. And the other big problem is you no longer believe the truth. Right? And once you lose both of those things, if you’re in a democracy, you’ve sort of lost the ability to govern.

Their first thought was to use artificial intelligence to detect if some photo or recording is fake or not.

So we took the question back to our research team. The first question is, can we use A.I. to detect fakes? Like, that would be the easiest answer, right? And the response we got back from our researchers was, the technology to do the editing, which is what we do, is always going to be at par or step ahead of any technology to detect it. It’s just like the security in the arms race where you like, you’re always— you’re improving your security, but the bad guys are out there improving their attacks. And sooner or later, you’re going to lose that battle, or at least something’s going to get through. 

But then—a eureka moment. 

We don’t necessarily need technology that can identify a fake. What would be just as good is a way to prove that something is real. That would solve the trust problem. If there’s some leaked recording of the president saying, you know, “I like to run over baby animals,” knowing if it’s authentic would be just as good as knowing if it’s a fake. 

Dana: And so we said, “all right, what is another way to talk about this problem?” Let’s flip the problem on its head. And what we meant by that was, why don’t we give a place for good actors to go to be trusted, instead of trying to catch all the bad actors, which we think is a losing proposition? And that’s what CAI is designed to do. 

CAI is the Content Authentication Initiative. Five years after the Project Voco demonstration, Zeyu Jin’s reference to watermarking— 

Zeyu Think about, like, a watermarking detection. 

—has blossomed into a full-blown—I don’t know, program? Feature? Technology? Campaign? Consortium? All of the above. 

I’ll let Dana Rao describe how it works.

Dana: It occurred to us that we’re in this unique position to help the consumers understand, like what happened to an image? I’m gonna, you know, enhance the image. I’m going to make it sharper. I’m going to make it clearer. 

You make all the edits, and then you publish it. Once you publish it on the social media platform or wherever it is, the people can see it, they can see a little icon and they’re like, “oh, I wonder if the president really did go there,” and they can click on it and they can say, “well, it was David who took it.” They can see the location of the image, where it was taken. They can see the edits that were made if they want to. They can actually see the original, they can go to our website, see the original image and see edited image and decide for themselves. 

Now you have the facts. You decide for yourself. We empower the user to do it. That’s sort of the end to end system that we’re working on with a bunch of different partners to build out and hopefully change the conversation around how you consume content. 

Obviously, this idea can work only if every link of the chain preserves that encrypted metadata that’s embedded in the picture or recording. The phone camera that takes it. The software that edits it. The social-media network that posts it. Every step of the way.

Dana; And that’s why this is an open standard. It’s not an Adobe tool, it’s not proprietary. We’re building it with a bunch of partners. We want everyone to use it, we want every news media outlet to use it. We want every social platform. We want everyone, whoever does this. This is not a not a play for us to get money. We’re not charging for it. So if you want your story to be told, you can do it. 

Already, a bunch of companies are on board, including chip makers like Intel, ARM, and Qualcomm; software makers like Adobe and Microsoft; news outlets like the New York Times, the BBC, and the CBC; websites like Twitter, Facebook, and Getty images; and 55 other companies.

Here’s an ad from the CAI website, which gives you an idea of how these companies will explain CAI to the public:

Ad: I am photographing with a CAI-enabled prototype. It’s saying, “Don’t take my word for it.” There’s literally software that can prove, like, I didn’t mess with this photo. This is where it was taken, this is when it was taken, and this is the certification that it’s me who’s made that content. 

The feature that the CAI companies are adopting has a name, too. It shall be known as “Content Credentials.” When you see something suspicious online, you’ll click a Content Credentials icon to see that content’s credentials. And the path that it took to your eyeballs. 

And now, the big punch line: after years of work, Adobe has finally introduced this feature to the public. Just this week—assuming you’re listening to this podcast when it’s hot off the servers—Adobe unveiled the Content Credentials at Adobe Max.

Yeah, that’s right: the story that began at the Adobe Max conference five years ago…ended with the Adobe Max conference last week. This episode has bookends! Now that’s what you call an ingeniously structured podcast.

Of course, the Content Credentials technology isn’t a silver bullet. For one thing, the version just released works only on photos. Adobe hopes to have video and audio authentication maybe next year. Meanwhile, Harvard’s Joan Donovan says we’ll still have a lot of work to do—in policy, law, and in public awareness:

Joan People have figured out how to wield this technology for serious, serious and grave consequences. We have a duty to the future to say that we’re not going to allow it. We’re not going to let it proliferate. 

And so as we think about the future of technology policy, I believe we need a whole of society approach. What is our responsibility to one another? What is technology companies’ responsibility for that distribution and that exposure? And then how do we as a society, like, figure out what the true costs of misinformation are, so that we can do something about it?

You know, throughout all of these interviews, I kept thinking: A new technology. Capable of editing a record of actual events. Experts predicting the erosion of public trust…Where have I heard all this before?

Deborah: A picture may no longer be worth a thousand words. These days, the picture that the camera takes may well not be the picture that we end up seeing in newspapers and magazines. Technology makes it difficult, maybe even impossible, to tell what’s real and what’s not.

That’s Deborah Norville, the host of “The Today Show,” in February 1990. Her guest that day was Russell Brown, from Adobe, demonstrating version 1.0 of a brand-new program called… Photoshop.

Russell: We’ll take this show of Nancy and Ron. I’m gonna place myself into this photograph. Based upon the skill of the artist using the program,  they can give the illusion that photograph was quite real.

There was also another guest, a cautionary voice: 

Norville: Fred Ritchin is an author who has written a book. You warn against the dangers of what people like Russell do.

Fred: Well, the thing is, when you see a photograph, you really tend to believe that something happened and when people start monkeying with photographs, you don’t know which photographs are real, which ones happened, and which didn’t. My concern is that if the media takes to doing what Russell is demonstrating now, that people, the public, will begin to disbelieve photographs generally, and it won’t be as effective and powerful a document of social communication as it has been for the last 150 years.

Of course, these days, nobody worries about Photoshop bringing down civilization. We’re totally blasé about edited photos. We just go, “oh, that must have been Photoshopped,” and we go on with our lives.

I asked Nina Schick if these audio and video deepfakes are really any different. 

David Is there a newness to audio and video deepfakes that makes it more terrifying? 

Nina Yes.

David And maybe we’ll just get to a place where everyone’s like, “oh, that’s probably a deepfake?”

Nina Photo and image manipulation has a long history. The difference now is that it is not just images. You are talking about video— video manipulation, which until now has only been in the realm of Hollywood studios. 

Still, she does acknowledge that there’s more to it than the dawn of the Infocalypse. 

Nina Like all powerful technologies of the exponential age, this is going to be an amplifier of human intention. It will be used for bad, just as it will be used for good. So just as they will be used by malicious actors, they’re going to be many commercially valid, legitimate applications. 

Now, I wanted to end this episode with a twist: I thought I’d let my own voice clone from Lovo speak the final paragraph. But when I got the results back from Charlie Choi, it sounded so much like me that I didn’t think you’d be able to tell when I stopped and the deepfake voice started, and the gag would lose all impact. So I’m going to make it super clear. From the end of this sentence until the credits, you’re going to hear nothing but software, starting…now.

Clone: I thought I’d give the last word to—my clone. My voice clone, the one that Charlie Choi’s team at Lovo made for me. You’re listening to him right now.

And what I’d like my voice to say is that: Well, in the end, voice synthesis is just another technology. What happens from here isn’t about the tool; it’s about whoever’s wielding it.

I’m David Pogue—or a synthetic version thereof. And this…is “Unsung Science.”

How We Almost Blew the Vaccine

Season 1 • Episode 2

It may seem as though we got the Pfizer and Moderna COVID vaccines incredibly quickly. But Hungarian biochemist Katalin Karikó had been trying to make mRNA vaccines work for 30 years while fighting scientific gatekeepers who thought her idea was absurd. Her grants were denied, her papers rejected, her speaking invitations withdrawn; eventually, the University of Pennsylvania demoted her. But she still refused to quit, and in 2005, she and collaborator Drew Weissman cracked the code. They figured out how mRNA could direct our own cells to manufacture medicines to order. Their breakthrough saved the world from the worst of the pandemic—and opened a new world of medicines and vaccines for a huge range of diseases. 

Guests: Katalin Karikó, senior VP at BioNTech. Drew Weissman, Perelman School of Medicine, U Penn. Derek Rossi, co-founder of Moderna.

Episode transcript

Unsung Science Episode 2: 

How We Almost Blew the Vaccine

Hungarian biochemist Katalin Kariko spent 17 years working on a medical idea that was so far-fetched, the scientific community soundly denied her grant proposals.

KK: You know, I was demoted from my position.

DP: Why were you demoted?

KK: Oh, because I didn’t get funding!

And in the end, she did it: she and her collaborator invented the mRNA vaccine. Without her, there would be no COVID vaccines from Moderna or Pfizer.

I’m David Pogue, and this is Unsung Science

[BREAK]

Season 1, Episode 2: How We Almost Blew the Vaccine.

[music]

As I sit here recording this episode, the COVID pandemic isn’t what you’d call

over. But it’s definitely been beaten back. It’s nothing like the death-eater

Armageddon it would have been if we hadn’t had the vaccines.

And I’m not sure you realize how miraculous it is that we got a COVID vaccine so darned fast.

Scientists analyzed the coronavirus for the first time in January 2020—and the Pfizer and Moderna vaccines entered clinical testing in April! Three…months… later! And went into the arms of the first patients in December. 

It was by far the quickest vaccine ever developed. Modern vaccines usually take ten or fifteen years to create. 

And to make this story even more incredible: the Pfizer and Moderna vaccines are both mRNA vaccines, which I’ll define in a minute. No mRNA vaccine or drug had ever been approved before. 

And you know the two companies that developed those vaccines, Moderna and BionTech? Neither one had ever brought a product to market before. (You may know the BionTech one as the Pfizer vaccine, because Pfizer did the manufacturing and distribution.) 

And COVID was only the beginning. 

ROSSI: The vaccine industry is going to pretty much all move over to RNA vaccines, simply because they— they’re very effective. They can be made very, very quickly. And ultimately, I think the cost of goods will be much cheaper. 

It’s also being used —for, you know, oncology. So, cancer applications. I mean, I think the possibilities are limitless.

That’s Derrick Rossi, a former Harvard professor, a really good explainer, and the cofounder of Moderna. He’ll be back.

But to me, the juiciest part of this story of all is how we learned to create mRNA vaccines in the first place. It’s a story of two scientists’ relentless, almost irrational devotion to the concept, despite years of rejection, humiliation, and ridicule. This story’s redemption arc is so incredible, it almost sounds like cheesy fiction. 

But before I introduce you to our heroes, I want to introduce you to a little cellular biology. Don’t freak out—it’ll be fun! I’m going to explain mRNA in the form of a bedtime story. Tinkly music box, please?

[Cue the tinkly music-box music.]

Once upon a time, there was a sensational little restaurant. The recipes dreamed up by Deena, the master chef, were genius. Miso-glazed lobster tails with sesame bok choy! Scallop Sashimi with Meyer Lemon Confit! Apple galette with vanilla-raspberry drizzle! 

And she did it all in her head! She didn’t fiddle with ingredients—she didn’t even have ingredients to play with in her little office, locked away in the middle of the restaurant. She’d dream up the recipes, and then send them off to the kitchen, which Deena called the Site Operations Center, or Site Ops. They turned her recipe instructions into delicious dishes, to feed the waiting customers. 

To hand her recipes off to Site Ops, she relied on her trusty assistant Myrna as a messenger. Every day, in the sanctum of that inner office, Deena carefully recited her recipes. Myrna memorized every syllable—and then headed out to Site Ops, to relay the instructions to the chefs. They’d make the recipes, send them out to the patrons. And they lived happily ever after.

And…SCENE!

[Music concludes abruptly.]

Wasn’t that great? You just learned molecular biology!

Well, kinda. 

In this super-simplified analogy, the restaurant is a cell in your body. And obviously, your cells don’t make lobster flambée or whatever I said before. What they do make are proteins, these giant complex molecules that perform just about every important maintenance task in your body. Proteins fight disease, communicate between your organs, convert food to energy, clot your blood, and on and on. Derrick Rossi really admires proteins.

ROSSI: The real worker bees in the cell, do essentially all the cellular functions which give rise to life, are proteins. And people don’t know that. They think of proteins largely in the context of what they eat. You know, if they’re eating steak, they’re eating protein—or beans or something, if they happen to be vegetarian. 

But actually, proteins, there’s a large, very large diversity of them in our cells, upwards of 30,000. And they really are the worker bees. 

The master chef, Deena? That’s DNA, which really does live in an inner chamber of the cell known as the nucleus. The DNA keeps the recipes for all those proteins, and sends them to the outer area of cell, called the cytoplasm—Site Ops. (See what I did there? I’m a punning genius.) 

ROSSI: So DNA lives in the nucleus, which is a, you know, a very localized compartment of the cell. And proteins are made in a totally different part of the cell called the cytoplasm, and never the two shall meet. 

That’s why we need a messenger: Myrna, the messenger, and the star of our story. She carries the instructions from the nucleus, out to the ribosomes—the protein-making equipment—in the cytoplasm. 

And by the way, scientists don’t actually call her Myrna. They pronounce it mRNA, which stands for messenger RNA. 

You probably saw that one coming up Sixth Avenue.

ROSSI: If the if the recipe is contained in the DNA, which it is, you have to get the recipe to the kitchen. The mRNA is the thing that carries it to the kitchen and it goes into the protein production factory. 

I call it the trifecta of life: DNA gives rise to mRNA, gives rise to protein, gives life. 

DAVID I mean, in elementary school, we learn about DNA, but who’s ever heard of messenger RNA? 

ROSSI: It’s the— it’s the neglected middle child. And I’m happy to hear that  mRNA is finally getting its due. 

So here’s the question that had been dogging scientists for decades: What if we could write our own recipes for making proteins? Over 4 thousand diseases result from mutations in our DNA, including cancer. What if we could step into that process—DNA recipe, ribosome manufacturing—and influence it? Those recipes could teach our bodies to make proteins that cure old diseases, or fight new viruses.

That would be huge. Just for example, we know we’re going to get more viruses and more pandemics. I mean, they come along every couple of years, right? SARS. MERS. Zika. COVID. This could be amazing. 

Turns out, we have had some success modifying the DNA in patients’ cells.

ROSSI: DNA, of course, has been used. And if you’ve heard, you know, you’ve heard of gene therapy, different types of gene therapy, these are DNA based. 

But editing the first step in the process is not quick or easy or always even possible. Because as you know from our bedtime story, it’s a bunch of steps to go from the DNA to the kitchen.

We’ve also tried intercepting the third stage in the process, we’ve tried making the proteins in a vat, and just injecting them. And that sometimes works.

ROSSI: The first therapeutic proteins came —came into being in the 1980s. Genentech, a company in south San Francisco, led the way with insulin. And since that time, over 120 different FDA approved protein therapeutics have been approved, and are in use today. 

But injecting proteins directly isn’t optimal, either, because they can’t help you with diseases inside your cells. The injected proteins can only swim around in the gaps outside your cells, in what’s called the extracellular space.

ROSSI: But proteins are not very good at crossing over into the intracellular space. So pretty much all protein therapeutics are limited to deficiencies or diseases that are manifest and treatable in the extracellular space. 

So: tackling disease by modifying the DNA isn’t easy, and tackling disease by injecting the proteins themselves is limited. But as Rossi points out—we’re forgetting about the intermediate step. Myrna.

ROSSI: So but what about that, you know, neglected middle sibling?

You know, like, what if we could hand new recipes to Myrna to deliver to the kitchen? What if you could inject modified mRNA? In other words, what if you could…shoot the messenger?

[Pun-reaction SFX] 

But seriously, folks. 

ROSSI: But mRNA, on the other hand—you could now have the ability to make intracellular protein therapeutics, which had never been doable before, not to mention extracellular as well. 

[Music]

Maybe you’ve heard of the spike proteins on the COVID virus—the tiny spikes that give the coronavirus its name. You know, cuz “Corona” means crown. The virus uses those spikes to inject itself into our cells.

Most vaccines work like this: You inject a weakened or dead version of the whole virus into the body. That teaches your cells to develop antibodies, which will attack the real thing if it ever comes along. 

But an mRNA vaccine wouldn’t require injecting the whole coronavirus. Our synthesized mRNA could trigger the manufacture only of COVID spike proteins. Your body would see the spike bits and go, “Ho-HO!, what are those? THOSE don’t belong in here!—I’d better manufacture antibodies.” Within hours, you’d start making antibodies that recognize the spike proteinLater, if you ever encountered the actual COVID virus, your cells would already know how to protect you!

Well, we’ve tried stuff like that. For decades. And we gave up. Until 2005, every modified-RNA experiment failed bigtime. Every time we tried to inject it, the body killed it on contact. Our cells didn’t appreciate that we were introducing a brilliant human-engineered invention intended to keep us healthy; it always saw the synthetic mRNA as some evil external virus, trying to sneak into our nuclei to reproduce. Here’s Derrick Rossi again:

ROSSI: It’s the story of when cells and viruses first met one another, really, which is, you know, hundreds of millions of years ago. And ever since that time, viruses have been trying to figure out ways of getting into cells to replicate, to, you know, complete their life cycle. And cells have been figuring out ways of detecting when viral DNA is injected, and combating that by various defense mechanisms. 

So it turns out that when you try to introduce RNA into a cell, you trip these very ancient antiviral pathways, which do a very good thing to the cell. They say, “bad news coming in, let’s shut down, let’s stop protein production. And if it looks really bad, if it really looks like an infection, let’s kill ourselves.” An altruistic suicide. 

Through many decades, people introducing mRNA into cells were very good at tripping these antiviral pathways, killing the cells in the dish. And basically, the field didn’t move forward because of that. 

DP: Is that the same immune response problem that Kariko and Weissman were worried about? 

ROSSI: It’s exactly that. It’s exactly that. And they’re the ones that solved it.

OK. You now know what modified messenger RNA is, and why we couldn’t use it to fight off viruses. After the break—you’ll meet the two people who thought they could crack the code—and the brutal years of rejection they faced for trying.

BREAK –

Before the break, I was explaining what modified mRNA is. But according to Derek Rossi, it’s also where the company name Moderna comes from.

ROSSI: Actually, I had originally some not so great ideas for the company. One of the first ideas that I had was Harbinger Therapeutics. So you know, the harbinger of of medieval times was that guy who would ride in on his horse to a town before an approaching army, and tell the town that, “hey, the army’s approaching. Here they come!”  So it was it was basically delivering a bad news, what you thought was a nice, happy town, life is all of a sudden about to be overturned by this approaching army. So that was the harbinger. So I thought that wasn’t a real great name for the company. 

So then just one day I was —it just struck me.

It was modified mRNA, which we shortened to mod RNA, and then was not hard for me to come up with Moderna from mod RNA. 

I was also explaining how a generation of scientists had given up on using synthetic mRNA to fight disease and viruses. Every time they injected the stuff, they got an immune response. Every time, our immune systems killed off the modified mRNA as though it were an invading enemy. Most researchers moved on to more promising areas of inquiry.

There was, however, one scientist who had not given up—and would not give up.

KK: So, you know, I am a daughter of a butcher. And when I decided I would be a scientist, I was in high school in a small city. I had no idea. I have never seen a scientist. I just figured out that I would be a scientist, and I would go to work. 

Katalin Kariko grew up in Communist Hungary, in a home without TV, refrigerator, or running water. She became fascinated by mRNA in grad school—but when her lab ran out of funding in 1985, she decided to come to the States with her husband and 2-year-old daughter. They sold their car for 1300 bucks, and she stuffed the cash into her daughter’s teddy bear, because Hungarian law limited how much money you could take out of the country.

Their daughter, the one with the teddy bear? Grew up to be Susan Francia, who’s won two Olympic gold medals on the U.S. women’s rowing team. That’s a harbinger of the kind of family we’re dealing with here.

Anyway, back to her mom. Katie Kariko, as her colleagues call her, was more or less obsessed with figuring out how to master modified mRNA.

KK: For 10 years at University Pennsylvania, from ‘89, I started that there to 1998 maybe, that I was trying to use mRNA for therapeutic purposes. 

And for ten years, the scientific community thought she was nuts. 

DP: Can you tell us how much success you had with—with grant proposals during the 90s? 

KK: Yeah, I did not get money. They always ask me that—you know, who is my supervisor? The woman, an accent, probably she wouldn’t come up with ideas like that. 

A woman with an accent—whatever the reason, nobody believed in her idea, and nobody would fund her research. She wrote proposal after proposal. In one of the talks she gives these days, she’s got a slide that consists of nothing but the rejection letters. Sooo many stories of slammed doors.

KK: It was 1993. We went to Princeton and we presented. So they could have been invested.  They promised the 70,000. That would be the best 70,000 dollar. But they never gave me the money, and…they never even return my phone call, not even today. I don’t name them, because they are still around.

Well, you know what they say about academia: Publish or perish. If you don’t bring in the grant money, you get demoted. And sure enough—

KK: You know, I was demoted from my position. 

DP: Why were you demoted?

KK: Oh, because I didn’t get funding! 

Penn took her off the professor track, because she wasn’t landing the grants—but once she wasn’t on faculty, a vicious cycle began. 

KK: And then later, I didn’t get funding, because they question that I am not faculty. 

DP: Today, how do you think about the people who turned you down or demoted you? Did they have good reasons? 

KK: They said that many things I didn’t do well, I could not articulate well enough, you now, the ideas, because I couldn’t attract the money. I acknowledge maybe I was not doing well, because they couldn’t see it, I couldn’t explain well.  

Drew: Most of them basically said, “yeah, we’ve heard of this before. RNA’s too difficult to work with, we’re not interested.”

That voice belongs to Dr. Drew Weissman. She met him at a Penn photocopier one day in 1998.

Weissman is a physician and an immunologist who had come from the National Institutes of Health, where he worked on an HIV vaccine with another immunologist whose name you might know—Anthony Fauci. 

Weissman told Kariko that he’d been looking into using genetic material to make vaccines, and she told him that she’d learned how to modify mRNA. He invited her to join his lab.

Which brings us back to that infuriating inflammation problem, the problem that made the rest of the science world consider the whole field a dead end:

Drew: Inflammation occurs whenever the body doesn’t like something. It can be a virus, a bacteria, it can be hitting yourself on the head with— with a brick… there’s lots of different types of inflammation. And it’s the body’s response. And that includes high fever, low blood pressure, feeling lousy, a variety of things. 

David: So that kind of response would not be ideal in medicine, you’re giving somebody no, 

Drew: You don’t want to make people sick with your medicine. 

Or your mice. Whenever Kariko and Weissman injected modified mRNA into lab mice, they’d lose their appetite, or their fur. They couldn’t get around the immune-response problem—or the no-support problem.

David: Is it normal for researchers to stick at it like that for so long when you were getting so many naysayers? 

Drew: I wouldn’t let any of my people work that long on something. The reason that I didn’t give up, and Katie didn’t give up, is that we saw the potential from— from the very beginning, we knew that there was enormous potential for RNA as a therapeutic. And it was more a matter of just figuring out how to make it work. 

David: And there weren’t family or colleagues saying, “dude, what are you doing? It’s a dead end now!”

Drew: No, I would get that all the time. I would go to meetings, and I would talk with other leaders in science and even Tony at some points. 

That would be Tony Fauci.

Drew: (continued) And he would listen to the data and say, “yeah, that’s really interesting, but what are you going to do with it?” And I basically knew I was being blown off. And I went back to work and kept working at it. 

David: So here’s my favorite part. How did you discover the way around this immune response? 

Drew: Yes, that was years and years of work. 

Now, I’ll let Weissman explain the solution, but first I kind of need to set this up. 

It turns out that your cells often dress up the proteins they make with little chemical attachments, little molecular modifications, that make them work better, last a little longer, or whatever. They’re like aftermarket mods. You’ve got the same car, but now it has a nicer stereo. You can think of them as decorations, or embellishments, or, as Derrick Rossi calls them,

ROSSI: —Dongles, if you will. You know, phosphorylation here, ubiquitination there, glycosolation here, it gets sort of decorated with all of these sort of modifications that are required for it to function properly in its day-to-day business as a worker bee. 

You know what else sometimes comes decked out with modifications? RNA molecules. And some types of RNAs have more of these extra aftermarket modifications than others, including RNAs from different animal or bacterial or plant cells.

OK, so getting back to the Kariko and Weissman experiment that changed medicine forever:

WEISSMAN: And the key experiment, we took a bunch of different kinds of RNAs. So RNA from bacteria, from mammals. There’s ribosomal RNA, transfer RNA, nuclear RNA, messenger RNA, mitochondrial RNA. We took all of those, and we tested them for inflammation. And they were all different! 

Some RNA types triggered the body to attack—and others didn’t. What was it about the winning types that let them slip by?

And the answer? The ones that didn’t produce inflammation—were the ones with a lot of mods!

Drew: And what we noted is that RNAs that had a lot of modification/ didn’t have any inflammation, and RNA that had none was highly inflammatory. 

To test that theory, they whipped up a batch of synthetic RNA that had a mod of its own—they added one molecule, called pseudo-uridine—and bingo. No more inflammation. Kati Kariko was ecstatic that there was no more immunogenic response, meaning that the immune system stayed quiet.

KK: I was so happy, not just because now that we could make a messenger RNA, which is not immunogenic, but what was important, / 10 times more protein was produced. I mean, you couldn’t even dream that— that finally is not immunogenic and so much more protein is made from it, you know, compared to the conventional RNA we made before. 

They had done it. After ten years on her own, and then seven years working with Weissman, Kariko had broken through the barricade. They had figured out how to introduce modified mRNA into human cells—that could trigger the production of any proteins they wanted. They’d figured out how to send Myrna into the kitchens of your cells, carrying recipes that never came from the master chef. They’re recipes we gave her to carry.

Kariko and Weissman published their results in 2005, in the journal Immunology—and then waited for the scientific world to lose its mind. 

DP: So in 2005, you published this paper. Did it set the scientific world on fire?

KK: No. Drew Weissman said, “oh they will, they will notice”— but nobody! nobody, nobody said anything. Nobody invited us, nobody cared!

They started a company; nobody would invest. They tried to get grants; they got one. 

But at least two people cared a lot about the breakthrough. One of them was Ugur Sahin (OOgoor ZAHin), cofounder of the German drug company BionTech, which would go on to develop the Biontech/Pfizer COVID vaccine. He hired Kariko in 2013; she works at Biontech to this day. 

The other person who read that 2005 paper was Derrick Rossi.

ROSSI: When we read the paper, we thought, well, let’s try this. And lo and behold, now, when we introduced mRNA into cells, we could get it to express whatever protein we wished. And the cells were as, you know, happy as pigs in mud. They were not dying. So this was— this was the key. And it was at that point that I founded Moderna —co-founded Moderna to— to bring this technology to development for mRNA medicines. 

I gotta tell you three things that really struck me about both Kariko and Weissman. First, of course, their sheer refusal to give up. For YEARS.

David: Is there something in your character and Katie’s character that made you guys so dogged to keep to keep working at it? 

Drew: I think that’s our personalities. We’re both pain in the butts. We we don’t give up. We —when we’ve got an idea that we think is good, we keep going after it. 

Their employer, Penn, owned their patents, and then soon sold them to an obscure chemical company in Wisconsin for $300,00.  Of course that was the best deal that little company ever made. It’s already made hundreds of millions of dollars off that deal, by licensing Kariko and Weissmann’s technology to Moderna and Biotech. Nice going there, Penn lawyers

So the second thing that surprises me is that they seem to hold no grudges. They’ve stood by and watched their invention make millions of dollars for other people.

David: Did you want to protect the technologies, so that you would be the beneficiary instead of other people? 

Drew: Well, we tried. We just couldn’t do it. We —we tried to license the technology from Penn, but we couldn’t come to an agreement with Penn. And so we had to give up. 

David: Wow. That would make me sort of bitter. Is there any bitterness on your end? 

Drew: You know, I’m sure we have —we’re unhappy about some things that have happened. We’re scientists. To us, solving the problem, developing the new findings, new technology, new treatments —to us that’s —what’s important. Grievances, who cares? 

KARIKO: I have to tell you, when I was hired at Penn in ‘89, my salary was $40,000 a year. And 20 years, so 2010, it went up to all the way to 60,000. 

My husband once told me that probably in the McDonald’s, I would get a better hourly pay. But it is —you know, I enjoyed what I was doing. 

Listen, if —I am 66 old. My family, my husband, we never have a new car. We always had the car which was coming in the trailer and he fixed it up. Probably I’d never buy a new car, because we’re so used to it not to have a new one! Probably I would freak out in the, you know, parking lot that somebody would scratch it! 

Of course, she said that before she and Dr. Weissman won 3 million dollars from

the Breakthrough Foundation, which was created by Sergey Brin, Mark

Zuckerberg, and other billionaires, to award important achievements in science.

She told the Philadelphia Inquirer that she plans to pour it back into her research,

and to support science education for financially strapped students. No mention of

a new car.

And the third thing: They both insist that the glory means nothing to them, either. These days, Kariko and Weissman are invited everywhere. Institutions celebrate and honor them, and the media harasses them. 

Drew: I mean, my family keeps pushing me to enjoy the —the spotlight. And anybody that knows me, the two things I don’t like are attention and talking. So to me, this has put me into uncomfortable situations where I’m doing things I don’t like and taking me away from the science. 

David: Does that include interviews?

Drew: Unfortunately, yeah.

KK: Listen, I do it for me. For me, it was important that I knew that what I am doing is good, reproducible and would be helpful. 

Even the knowledge that I know that I contributed to something is —is sufficient. For me, I didn’t need it that people will know that. No, it is not important. I know that, and that’s it. 

And many times I just thinking about, you know, that a hundred years from now, nobody will know that we existed. So what is this fighting for? 

DP: A lot of people think that there’s a Nobel Prize in your future. 

KARIKO: I’m not interested in money or prize and anything. [00:34:00] I was just recently asked, can —can I explain that how to be successful? I don’t know. I don’t know.

Some measure how many times you fail, then you still have enthusiasm, and keep your enthusiasm, and you just keep going. That’s maybe success. Other people, maybe money is the success. I don’t know how you measure it, but being happy is important. Happy, enjoying the work you are doing. 

DP: How about this for success? How about laying the scientific groundwork that saves millions of lives? 

KK: Yeah. And —and more people get vaccinated and they feel safe, you know, to go out, or meeting their relatives and, yes, I am very happy. 

By the way, Kariko, Weissman, and Rossi all stressed to me that every breakthrough stands on the work done by previous scientists.

ROSSI: To get anything done in science is a large, generally speaking, a large community of people building off the work of others, and building on the shoulders of others, to move things forward. / Science—Science has to work this way, and it worked this way very well this time. 

By 2019, Kariko had been at Biontech for six years, working on an mRNA vaccine for the flu. The company was already talking to Pfizer about manufacturing and distributing what would have become the world’s first mRNA vaccine. That flu vaccine was just about to begin human trials—when COVID hit. 

That’s one of the reasons we got the COVID vaccine so fast—because both Biontech and Moderna already had other mRNA vaccines working. Once they had sequenced the coronavirus, they were able to repurpose those vaccines and get them to trials fast.

Maybe it’s just a coping mechanism. But Katalin Kariko, who spend her entire career working on RNA, who was rejected from grants, demoted at her job, and had her name taken off of papers, sees all of it as part of the journey. Just bumps on the road that led her to the mRNA breakthrough that’s helping to end the pandemic. Because it led her, eventually, to BioNTech.

KK: I truly feel that, you know, if with my colleague, I’m not going over there in Biontech and we work—I don’t think that that would be BionTech Pfizer vaccine. So we have to thank those people who showed me the door, kicked me around!

Drew Weissman is still at Penn, working on an even more impressive vaccine.

Drew: Well, so we’re thinking ahead. There have been three coronavirus epidemics in the past 20 years. There’s going to be more. I mean, it would be foolish to think we’re not going to have more. 

So we started last spring working on a pan-coronavirus vaccine. So the next time there’s, you know, a new pneumonia somewhere in the world that turns out to be a coronavirus, we’ll have the vaccine made./ to stop the next pandemic. 

Derek Rossi left Moderna in 2014; today, he has two new drug-development startups, one working on cancer drugs, and the other on a multiple sclerosis therapy. 

ROSSI: What’s cool is that the success of these RNA vaccines has led to new mRNA companies sprouting up like mushrooms in Boston, Cambridge and around the planet. An industry has been born, that’s for sure. And I think that’s great because it just means more money and more resource and more brainpower, more people working on cool ways to, you know, affect our health when we’re unhealthy. 

Moderna itself is plowing full-steam ahead into more mRNA-based treatments. It’s in human trials for vaccines against HIV, Zika, Chikungunya, RSV and CMV, a few kinds of cancer, and, of course, that flu vaccine.

We make flu vaccines as we have since the 1930s, in a hilariously old-fashioned process that entails injecting the virus into, I kid you not, chicken eggs. 

Unfortunately, there are many different variants of the flu. So every year, researchers have to guess which flu strains will be common in the U.S.—a year from now. If they’re lucky, the flu vaccine’s effectiveness might be as high as 60%, as it was in 2010. If not, it can be only 10% effective, as it was in 2004.

But if we had mRNA flu vaccines, like the ones Moderna and Pfizer-Biontech are developing, we could have the vaccine only weeks after we know about the virus. We wouldn’t have to guess which flu strains would be here—we’d know! We could make the vaccine based on the kind of flu that’s already here.

Oh—and no chicken eggs would be involved. 

Moderna plans to combine its new mRNA flu vaccine with the COVID booster shots, so every year a single shot would protect you against both.

And it all started with Katalin Kariko and her unshakable belief that mRNA could be used to fight and prevent disease. So I’ll give her last word. I asked her if she had any advice for younger scientists. She mentioned hard work, and being a good networker with your colleagues. And…

KK: And it is also important to select a good partner. And my husband is very supportive. So he was not complaining that I am not cooking, you know, things. And I’m coming home, you know, Saturday and carrying, you know, the little machine. And I asked him to fix it because I need the next day. (laugh) And he was doing that, and he was always said, “OK, just go.” So —so he was very supportive. 

DP: You got lucky there. What a guy. 

KK: And I have to tell you, by the way, that when I met him, he was 17 years old. And when we married, my mom didn’t even give us a one year. And we were just celebrating 40 years together!

DP: Oh, my gosh!

KK: I always was, like, knowing what I am doing. 

What Happened to the Mosquitoes in Fresno?

Season 1 • Episode 1

Mosquitoes are the deadliest creatures on earth; they kill 500,000 people a year—and as the planet warms, more species are spreading North from the tropics. In 2013, a nasty new type, called Aedes Aegypti, arrived in Fresno, California. But traditional tactics, like spraying insecticide and genetic modification, have ugly side effects. So one genius programmer from Google thought up a better solution—that doesn’t involve insecticide; doesn’t mess around with genes; doesn’t require irradiating; makes it impossible for the mosquitoes to develop resistance; can’t affect any other species; and costs less than what governments spend now on treating their citizens for Dengue fever. A lot was at stake in the Fresno experiment; if it worked, the technique could save lives around the world. (Spoiler: It worked.) 

Guests: Linus Upson, VP of Engineering at Verily. Leslie Vosshall, professor of neuroscience at Rockefeller University. Jodi Holeman, Fresno Consolidated Mosquito Abatement District. Peter Massaro, Google director of automation. Jacob Crawford, senior scientist, Verily.

Episode transcript

Unsung Science 1.1—The Mosquitoes of Fresno

Theme begins.

In 2013, Fresno, California, was invaded. The conqueror’s name: Aedes aegypti. 

Jodi: It is, like, the nastiest mosquito that’s out there. So it’s — it’s like the perfect evil species, really. 

You can spray insecticide, but that also kills bees and butterflies. You can try genetic engineering, but you run the risk of the mosquitoes developing resistance.

But now, there’s a third option—no chemicals, no genetic shenanigans. And it was dreamed up…by a software engineer at Google.

I’m David Pogue—and this is “Unsung Science.”

AD BREAK

Hey there! I’m David Pogue, and this is Unsung Science: the stories behind amazing accomplishments in science and tech. 

Season 1, Episode 1…get psyched! 

The deadliest animals on earth are not snakes, scorpions, sharks, or even other people. It’s mosquitoes. They kill a million people every year—by biting us and infecting us with fun stuff like malaria, dengue fever, yellow fever, zika, West Nile, chikungunya, and so on.

You probably associate most of those afflictions with faraway, tropical climates. And that used to be a good assumption. But nowadays, in the climate-changed climate? Not so much.

Vosshall: So the milder the winters, the more favorable the neighborhood is for mosquitoes. It’s like a new housing development that’s opening, like, “Now building phase three!” Right? That’s going to just keep moving up the Eastern Seaboard, because they need to not have overly harsh winters to be able to survive. 

This is Leslie Vosshall, a professor of neuroscience at Rockefeller University. In 2015, she was elected to the National Academy of Sciences, which is a big deal. I know this, because she’s my sister-in-law.

Pogue: Sorry not to have had our annual holiday get together. 

Vosshall: I miss it. I was I was so ready for it. It’s the highlight of our year. 

For the last decade, Dr. Vosshall—oh man, so weird to call her that—Leslie—has been studying mosquitoes: how they find us and why they bite us. So before I tell you why you should care about Fresno’s mosquito problem, I’m going to let her blow your mind with three mosquito facts. 

Fact #1: Only the females bite. 

Leslie: Yeah, this is an important point. And I’m amazed that 99 percent of both the general public and Ph.D. scientists find that surprising. Most people assume that a mosquito is a mosquito, that the females and the males are both equal-opportunity blood feeders. Only females bite. 

Which is going to be very important as this story unwinds. 

Ok, Fact #2: Those little ladies don’t mean us any harm. 

Leslie They’re not doing it on purpose. They’re not doing it because they hate people and they want to kill them. The females are going about their mandate. Their mandate is to get blood. That is their one job as future mothers, is to get blood. Because without that blood, she won’t produce children. 

They’re not even the original source of the diseases:

Leslie: (cont’d) So humans just happen to be carrying these diseases, and the females spread them. In the process of eating their meal from us, they end up infecting people. 

And fact #3 is the one that fried my brain. In the warmer climate, mosquitoes are spreading northward into new regions—but not by flapping their little wings.

Leslie: They don’t really move around by their own power. They’ll move around maybe a half a mile. But it’s people that move them, absolutely you’re right about that.

That’s right: We are spreading the mosquitoes. 

Leslie: The two biggest ways that mosquitoes hitchhike around the world, those stupid floral arrangements that you get, those little bamboo things, those little sprouted bamboo things that are filled with water—where are they set up? They’re set up in the tropics. So they come from China, Thailand, Vietnam. Somebody sets them up, puts them on some big container ship and puts them into flower shops all over the country. And those things tend to be full of mosquitoes. 

The other major way that the animals hitchhike is on the global used tire trade. So, tires are heaven for mosquitoes. If you want to set up a mansion for mosquitoes, tires! Because they’re black, they end up heating up, and they are they filled with water, which just becomes like a teeming nursery for mosquitoes. 

Leslie Vosshall spends most of her time studying one mosquito species in particular: Aedes aegypti. Two Latin words: Aedes, A-E-D-E-S, and aegypti, spelled like A-Egypt-I. It’s the worst.

Jodi: Yeah, it’s probably, you know, top two most wanted mosquito in the queue. You could arguably top—most wanted mosquito in the mosquito world. So, yeah, it’s not the one that you want to be dealing with. 

Jodi Holeman is the superintendent of operations for the Consolidated Mosquito Abatement District in California—that’s Fresno and nearby Central Valley towns. She’s quiiiiite familiar with Aedes aegypti. 

Jodi: It is, like, the nastiest mosquito that’s out there. It’s super aggressive with people, difficult to find, difficult to control. A single female will bite you multiple times. I can’t tell you how many residents have shown me their legs, their arms, their extremities, and all the bites that they have. So it’s — it’s like the perfect evil species, really. 

Until a few years ago, Aedes aegypti was the rest of the world’s problem. But in 2013, as part of its master plan to move north, aegypti arrived in Fresno, California. 

Mercifully, the Fresno aegyptis are not carrying diseases, as they do in other parts of the world—at least not yet. But they are incredibly annoying. 

Jodi: People are not used to it. People in our jurisdiction in the Central Valley of California, they associate really bad mosquito problems with camping, or going up into high elevations—not going into their backyards, right? So now we have this new species that’s super aggressive. 

Now, we’re not completely helpless. We do have some anti-mosquito artillery at our disposal. For example, we can spray insecticides—and we do. A lot. 

Unfortunately, insecticides kill more than mosquitoes. They also wipe out lots of innocent-bystander bugs—like honeybees, ladybugs, and butterflies—some of which actually would have helped with the mosquito problem, if, you know, we hadn’t killed them. 

Insecticides also accumulate in the water, which kills frogs and fish, and eventually makes its way into our bodies.

And worst of all, over time, mosquitoes develop resistance to our insecticides, exactly the way certain bacteria become resistant to our antibiotics. The world’s chemical companies have to keep tinkering with the formulas so their sprays remain effective.

There are also ways to reduce the mosquito population through genetic engineering, using CRISPR gene-editing techniques. But according to Leslie, that approach has its own problems, one of which is, again, resistance.

Leslie: Once you start CRISPRing mosquitoes, I do have the concerns about, like, what is the long term strategy for having that not get you resistance in a couple of years? I think even already in the laboratory, it doesn’t take that long for mutations to arise, where the female will circumvent —the population will circumvent that. 

The other problem with genetic modification is that it involves the words “genetic modification.” Which terrifies a lot of people. You know, playing God, tinkering with nature’s delicate balance, all that kind of thing.

Leslie: People are at this unprecedented state where nobody trusts scientists. There has to be an enormous amount of public engagement. You have to convince the public that it is safe.

Out in California, Jodi Holeman’s team had tried everything to solve the mosquito problem in Fresno. Spraying. Educating people about standing water in their yards. Nothing worked.

Jodi: Even with literally throwing every tool we had in the toolbox at this particular species when we first identified it in our district, it continued to spread. It was a—pretty colossal failure. 

OK. So how do you solve a problem like aegypti?

With Silicon Valley software engineers, of course. You’ve probably heard that cliché about how they just want to make the world a better place—but sometimes, they actually do it. 

Linus: After working at Google a number of years, I managed to get the freedom to have, to go try crazy things. And I was able to get permission to go and try this crazy thing. 

Linus Upson no longer works at Google. Now he works at Verily Life Sciences, Google’s sister company—part of the Alphabet family. Verily is dedicated to solving public-health problems.

Linus: (cont’d) And at the beginning, I gave ourselves maybe a 20 percent chance of success. There’s a lot of things in biology and in health care that fail. It’s much less reliable than computer science. But I thought the payoff was big enough in terms of the impact you could have on global health, that it was worth trying. 

That payoff would be wiping out mosquito-borne diseases by drastically reducing the mosquito population. But not by spraying, and not with genetic tampering.

His big idea—the one he gave a 1 in 5 chance of success—was the sterile insect technique, or S.I.T. 

I hope you’re sitting down.

Linus: So the sterile insect technique was developed by a couple of very clever entomologists. If a sterile male mates with a fertile female, she’ll still get the insect equivalent of pregnant. She’ll still produce and lay eggs, but the eggs won’t hatch. You can make each generation smaller and smaller. 

The sterile insect technique debuted in the 1950s, when it was used to tackle the New World Screwworm Fly, a nasty little parasite that was killing thousands upon thousands of cattle, and costing ranchers billions of dollars.

Linus: They chew their way through the animal, killing and maiming it in the process. And this was costing cattle ranchers billions of dollars. 

And so they reared billions and billions of screwworm fly, dropped them from airplanes across the United States, and over a 10 year period, completely removed screwworm fly from the United States. 

So here, at last, is the big reveal: Linus Upson’s master plan. 

His mosquito-control idea doesn’t involve insecticides; doesn’t mess around with genes; doesn’t require irradiating the males; making it impossible for the mosquitoes to develop resistance; can’t affect any other species; and costs less than what governments spend now on treating their citizens for Dengue fever.

The experts rave. Jodi Holeman, Consolidated Mosquito Abatement District:  

Jodi A beautiful system. It really is. 

Leslie Vosshall, Rockefeller University: 

Leslie: Very clever, very effective, kind of more cost effective because they don’t have to deal with a few animals escaping, still being fertile.So I think that that’s highly effective. 

The key to the whole thing is another Latin-named critter: Wolbachia.

Linus: Instead of irradiating the mosquitoes to make them sterile, we actually take advantage of a naturally occurring bacterium that exists in more than half of the world’s insects, called Wolbachia. 

That’s W-O-L, b-a-c-h, I-A. Wolbachia. A very common, very widespread, harmless kind of bacteria. It’s found naturally in 60% of all insect species—but not Aedes aegypti. 

And here’s a lucky break: If a male mosquito does get Wolbachia, it can go right ahead and have sex with a female in the wild that doesn’t have Wolbachia. She’ll lay her eggs—but they’ll never hatch. 

Linus: And if the male has Wolbachia and the female doesn’t, she’ll still produce and lay eggs. But the Wolbachia will arrange for those eggs to die. 

David: Wow! So the male doesn’t know that anything’s wrong, the female doesn’t know that anything is wrong. She—she goes ahead and lays her eggs. It’s just that the next generation doesn’t come along. 

Linus: That’s right. 

It sounded very cool in principle. All he had to do to get started… was build the world’s largest, airlock-controlled, robotically governed mosquito-raising factory—and then invent a machine that could separate the boy mosquitoes from the girl mosquitoes without making a single mistake.

Last month, I got to see the place. After the break—I’ll take you on a tour.

BREAK

Verily Life Sciences is dedicated to developing cool moon-shot human-health inventions like battery-powered, glucose-sensing contact lenses, or a spoon that holds steady even if you have a severe tremor. Its headquarters are a pair of huge, sleek, four-story office buildings in South San Francisco. 

If you pass through a courtyard, through security, and down a couple of hallways, you arrive at the mosquito-rearing facility. Our tour guide is Pete Massaro, Google’s director of automation. He designed a lot of the machines you’re about to meet.

The first stop… is the airlock. 

Peter: Remember—leave the mosquitoes in.

(airlock noise)

David: You have an airlock!? Oh wow.

(Leave the audio up, so that DP has to raise his voice.)

It’s just like the one on a spaceship, except here, they’re not worried about losing air to outer space; they’re—

(end of the blowing sounds—DP must compensate his volume)

—they’re worried about stray mosquitoes getting out of the building. Big fans blow air inward into the facility. 

Next, you enter… the insectary. 

Sound of the mosquitoes throughout:

It’s a plain white room, dominated by four gigantic mosquito cages, each a cube about four feet square. That sound you hear? That’s not fans. 

POGUE: I– I have to comment on the sound. (LAUGH) This is, like–

PETER: 600,000 mosquitoes flying sounds like– (LAUGH)

POGUE: A nightmare. (LAUGHTER)

PETER: It should make you itchy. 

It took me days to get that sound out of my brain.

Verily gets these mosquitoes, pre-infected with Wolbachia, from a company in Kentucky. Their eggs hatch into larvae, which look like minuscule white specks. They’re poured into clear plastic pouches, along with food and water, and loaded by robot into this—oh man, how can I describe it?

Begin robot audio

Imagine a grocery-store aisle, except that it’s only 2 feet wide. And the shelves on either side of you are made of stainless steel, and only a few inches tall each. So there’s like 35 shelves on each side. Each shelf contains those plastic bags full of developing mosquitoes. 

And running back and forth between the shelves is Pete Massaro’s masterpiece: a robot. An eight-foot-tall, faceless, shiny silver storage-and-retrieval robot. 

PETER: They’re now in here with the food and water. / 16:43:07  It’s also insectary conditions in there. So it’s 80 degrees. We feed those with that same robot– four times during the six days. 

All day long, the robot zooms back and forth through these skinny aisles, raising generation after generation of mosquitoes. It loads new trays into the shelves, tends them and gives them food pills, and then, after six days of growing, pulls the finished trays out.

The robot doesn’t have a name, but as far as those mosquitoes are concerned, it’s, “Mom.”

POGUE: And I see barcodes on those trays. Are you able to track these–

PETER: Every single– every location is tracked. Every mosquito has a name and a number. And every single mosquito has an image that’s stored on Google Drive.

POGUE: Its picture?

PETER: Every single one of ‘em. (LAUGH) Every one from the very first mosquito we ever made. 

POGUE: Have you ever thought about running the Department of Motor Vehicles? (LAUGHTER) It’s the same problem.

PETER: It would be fun. (LAUGH) 

After six days, the larvae have grown into pupae. At this stage, the females are slightly bigger than the males, which works out nicely for the next step: Separating the boys from the girls.

PETER: So in order to do the– sterile insect technique, using Wolbachia, you need to release only the males. So you wanna be very careful not to have females released. We have a very carefully designed sieve that separates about 97% to 99% of the females from males. 

Yes, a sieve. Yet another robot slices open each pouch and dumps its watery contents through a sieve. 

PETER: The males will– fall through the sieve. And females go to their destiny. Which is the municipal waste water treatment plant. (LAUGH) But actually– we actually, first we– we cook them. So that they’re not live.

And now comes… the really cool part.

The sieve is 99% accurate at separating the bugs by sex. But in this context, 99 percent isn’t good enough. If you release any Wolbachia females into the wild, they could wind up mating with your Wolbachia males, and the whole beautiful system of population control would fall apart. 

So how do you eliminate the last one percent of females? This is how Linus Upson explains it:

Linus We had to develop machines that can separate male and female mosquitoes at near perfection. 

David Come on. That sounds like a bad science fiction movie. How can you create a machine that separates male from female mosquitoes by the millions?

Linus There’s an engineer in our team who is an automation expert by training, named Victor Criswell. He spent about two and a half years trying to figure out how to get mosquitoes to march single file in front of a camera, so that we could take their picture and use jets of air to sort the males and females apart. Because the males and females actually look quite different from each other, if you can get a good picture of them. 

And these bugs just don’t want to do what you want them to do! If you want them to fly, they walk; if you want them to walk, they’re going to fly. If you want them to go one way, they’ll go the other way.

David: Was it ever in doubt that you’d be able to get these machines working? 

Linus: Oh, yeah! It was in doubt for about two and a half years! We tried dozens of different configurations, different ideas. Can we tell the difference between males and females by wing beat frequency? Do we want them flying? Do we want them walking? 

Yeah, it was lots and lots of things were tried before we were able to get to the point where we developed confidence of like, “OK, we don’t have it perfect yet, but we see a path where this could work.”

And believe it or not, it actually works. I witnessed it! The factory has over 150 of these sex-sorting machines, row after row of ‘em.

In each machine, you can see the mosquitoes walk one by one into a narrow, white, illuminated tunnel. A gate snaps shut behind it. That clicking you hear is all the little gates opening and closing.

PETER: It’s almost like ant trails or something. We don’t exactly know why they do this. They just get in there and walk right through. It’s, like, “Yeah, this is where I belong.” Right? 

Now, a camera takes the mosquito’s picture. Software identifies it as either male or female, and then clicks open one of two gates. A little puff of air blows the bug out of the tunnel.

PETER: And so if it’s a male, it will send it up into the container. If it’s a female or unidentified, it will send it to its doom.

POGUE: And– and that works?

PETER: That has worked so incredibly well that, you know, to our knowledge, no females have ever left this factory. 

POGUE: Well, Pete, would you go so far as to say that what is before us is the world’s most advanced mosquito sex-sorting machine?

PETER: I would definitely say that. (LAUGHTER) 

By the summer of 2017, the whole thing was ready to launch. The Verily team had worked out the bugs—sorry—and built themselves a rip-roaring, well-oiled mosquito factory, capable of churning out millions of males a week, every single one infected with Wolbachia bacteria. They had even gotten approval from the EPA to perform this experiment in the wild. 

Now all they had to do was convince the residents of Fresno that this was going to be a good idea. The pitch was basically this: 

“Hello there, neighbor! Say—a bunch of Silicon Valley engineers are going to be driving around in weird white cannon vans, shooting out millions more mosquitoes into your front yards. Ok, sound good?” What could possibly go wrong?

JODI: When you get out of the laboratory and you start to get into people’s neighborhoods and into their lives, you’ve got to have a stronnnng communication program. 

David: You mean, you think that ordinary citizens might not immediately love the idea of you guys air-dropping millions more mosquitoes

Jodi: I know. It’s just a—It boggles the mind that people just aren’t like, “oh, yeah, sure, whatever you want to do.”

Representatives from Jodi’s team and the Verily team held a series of community gatherings and movie nights, offering free Jamba Juice cards to anyone who’d show up.

Jodi: You had people that were either, “I don’t—I don’t care what you guys are doing. Especially if you’re going to do something about this mosquito because it’s driving me nuts.” 

And then you had the other side of— the other extreme, which is I mean, it just ended up being like a list of conspiracy theories, really. People that had all these thoughts as to what we were doing, that — that we were trying to do population control, that we were releasing insects that carried something that would make people sick and die. 

David: Wow.

Jodi: It’s—it’s disheartening. It’s hard to hear somebody say to you that they genuinely think you’re out to kill them. And so I try to —while I can’t validate that, I try to acknowledge it and say, I just—I just will not give up in trying to find common ground. It was it was a— it was a battle. 

And even once the Fresnonians were convinced that the project had value, she had to prepare them for—the vans. 

(Horror-movie chord sound)

JODI: Because it’s not an ice cream truck. They designed these vans that have these ginormous tubes that are really, really cool. And these tubes have literally thousands and thousands and thousands of mosquitoes, male mosquitoes inside of them. That is connected to basically an apparatus that sort of pulsates them out. It kind of blows these mosquitoes out of the side of the van.

In the summer of 2017, the Verily mosquito vans made their first runs, blasting millions of male mosquitoes into the sunny Fresno air. 

All they had to do now was wait to see the results.

Jacob: So we were actually able to see the results within about a week and a half or two weeks. Not— not to say that that was the end point, but we could see the impact of our releases. 

Jacob Crawford is a senior scientist at Verily—a mosquito biologist on the Debug team. To monitor the experiment’s results, his crew put out mosquito traps all over Fresno, and counted how many adult mosquitoes wound up in them. But he could also collect mosquito eggs around town.

Jacob: All we have to do is, is kind of cure them a little bit back in the lab, and then flood them as if it was raining, which is what they’re waiting for. And if they have mated with one of our males, then they’ll never hatch. 

Debug Fresno got a late start in the 2017 skeeter season; all that EPA paperwork had held them up. But when 2018 rolled around, they were ready. 

Linus: In 2018, we sort of had everything dialed in and everything ran really smoothly. Because it was our second time doing it and we got gotten a lot better at it. 

It didn’t take long for the team—and the town—to learn the results. 

David: And what were the numbers? 

Jacob: So the previous week we were seeing that something on the order of 70 to 80 percent of the eggs would hatch? And in the span of about a week or two, that —we took that down to zero. I did a dance at my desk. That was the strongest results I had seen and many, many moons. 

Jodi Holeman flipped through the binders of data that revealed how many female mosquitoes were showing up in the traps.

Jodi: Suddenly it’s like, zero zero zero zero zero, indicating no females and no females in the trap, no females in the trap. So that was very exciting. 

It was the biggest mosquito SIT experiment ever run in the U.S., and it was a grand slam.

Jacob: We managed to reduce the mosquito population by 95 percent… in the areas where we are treating compared to the areas where we weren’t treating. 

David: I would call that a success. 

Jacob: All that is a massive impact. Absolutely. 

David Do you think residents would notice? 

Jacob: Absolutely. Yeah. They went from not being able to use their backyards because the mosquito was so aggressive, to having a summer where they weren’t getting bitten. They were all saying the same thing: “Thank you. This has been a great, a much better summer for us.” 

Verily debugged Fresno one more time, in 2019—again, with spectacular results—but then that was it. From the beginning, Debug Fresno had been designed as a three-year test, a proof of concept for bigger and needier areas—like Singapore, an island nation with record-breaking waves of dengue fever. Last year, 35,000 Singaporians were diagnosed with this awful mosquito-borne disease, which brings extreme fever, internal bleeding, shock, and sometimes death—the highest number ever recorded in a year. 

Linus Upson’s team has partnered with Singapore’s government to develop a program called—Debug Singapore.

Linus: And so we’re in the process of building our first dedicated factory, to make mosquitoes in Singapore. It has the floor space to be capable of running a program for the entire country. 

Verily is also laying the groundwork for Debug Puerto Rico. 

Linus: The other place we’re working right now is, we’re partnering with the CDC in Puerto Rico. So we’ve been producing mosquitoes and shipping them to Puerto Rico throughout the entire pandemic. 

Pogue: Shipping them to Puerto Rico? From here?

Linus: Yes. It was easier to ship them from here than to build a factory in Puerto Rico.

Pogue: I see. How do you do that?

Linus: There was a lot of engineering work in figuring out how to safely transport mosquitoes in airplanes, at lower pressure, and sitting on tarmacs in the heat, and getting them out there. We have a special container they go in, and we control oxygen, and carbon dioxide, temperature, humidity, so they can all get there safely.

After Singapore and Puerto Rico, well, the world’s the limit.

Linus: We’re talking with a number of other places in the Caribbean and a number of other places around the world. We’re just now getting to the stage where we can take on multiple projects at the same time. 

Now, Verily is part of Alphabet, and Alphabet is a for-profit corporation; and ultimately, Linus Upson says that the Debug program has a business model.

David: So someday the governments will, in theory, pay the company to run these programs. 

Linus: That’s the plan.

If all goes well, Verily’s government customers will be able to save lives and save money.

Linus: So the way we’re approaching this from a business perspective, is we want to make the cost of this intervention substantially less than governments’ direct health care spend on dengue. 

David: OK, so if you can get rid of the mosquitoes in a certain region more cheaply than what a government is now spending to treat the disease, the government will say, well, then that’s worth it. 

Linus: Correct. 

I gotta say, I love this story. I love that a software engineer with a goal to save millions of lives might actually do it. I love that Verily’s approach doesn’t involve chemicals or genetic tampering. I love that this crazy program actually works. 

It all just sounds a little too perfect. Surely there’s some unwanted side effect. Surely these Silicon Valley geniuses have overlooked something.

David: What is there to be afraid of if this becomes a common technique? You know, are you depriving birds of their meals? 

Linus: So Aedes aegypti mosquitoes are invasive in most of the world. Humans spread them around the globe starting about 400 years ago. We’re the ones who are building all the habitat for them. We build gutters and storm drains and great places for mosquito larvae to breed. And so we have dramatically amplified their population all over the world. 

We’re the ones who, who created that population of mosquitoes to begin with. So from an environmental standpoint, we’re just cleaning up the mess we made in the process. 

David: It just it seems too good to be true. 

Linus: Unlike chemical pesticides, which, you know, are broad spectrum and kill a wide range of insects, one of the wonderful parts of the sterile insect technique is that is exquisitely species specific. You’re targeting just the one thing that you’re going after. And the sterile insect technique, we’ve been doing now for 70 years. It’s been one of the most successful interventions for crops and livestock. And now we just want to be able to also apply it to human health. 

David: So is there anything to worry about? 

Linus: We haven’t been able to come up with one. 

But if you live in Fresno, you might have one concern. What makes this sterile insect technique so safe and so controllable is that it affects only one generation of bugs. The males that you infect with Wolbachia don’t pass on Wolbachia, because, remember, they can’t make babies. So—no unintended consequences. 

Unfortunately, that SIT affects only one generation is also the bad news. Because if you don’t keep shooting Wolbachia males out of your vans, the wild mosquito population will bounce back. 

There was no Debug Fresno program in 2020, and there won’t be one in 2021. So for now, the era of Fresno’s beautiful, bite-free backyard barbecues is over.

Jodi: Basically yes, the population will rebound really quickly once you stop releases. Mosquitoes from outside neighborhoods, they’re going to move back in really, really quickly. 

Still, Jodi Holeman holds out hope that a Wolbachia project could return, or expand. Maybe Fresno, or maybe California, could become one of Verily’s paying customers.

Jodi: Can you imagine if you could roll out a program that started from, like, Southern California and just released slow and just— just gradually, knock this population out as it went up the state? 

Of course, it’s a massive effort. It’s an expensive effort, it has to be a coordinated effort. But I feel like, you know, there are some things that residents are willing to pay for. And when they’re being terrorized by mosquitoes, if you have something that you can tell them, “Hey, it’s 95 percent,” I think you’ll have strong support for funding a program like that. 

Now we’re going to send you out into the sunset with a special treat. Jodi Holeman’s
boss, Steve Mulligan, has a hobby. He writes song parodies about, if you can believe it, the sterile mosquito technique you’ve just been hearing about. With tremendous apologies to Nirvana and it’s song “Smells Like Teen Spirit”, take it away Steve!

(Steve sings parody mosquito song)