Latest / kill switch / Truth to Power
Transcript
- 0:04Speaker 1: Sleepwalkers is a production of I Heart Radio and Unusual Productions. Hey,
- 0:14Speaker 1: how are you doing? Are you in a curious situation?
- 0:19Speaker 1: But not really? Where are you? Can you say where
- 0:24Speaker 1: you are? You sound like you can't talk. Yeah, okay,
- 0:31Speaker 1: I've been working on Sleepwalkers so much. Yeah, so I
- 0:36Speaker 1: know what you're thinking. Kara isn't normally that distracted? But
- 0:40Speaker 1: the truth is that wasn't her speaking. We're playing prerecorded
- 0:44Speaker 1: fakes of her voice to her cousin, created by AI.
- 0:47Speaker 1: Are you sleeping today? Wake you? I feel so tired.
- 0:51Speaker 1: I'm sorry. I wanted to talk to you about huh.
- 0:57Speaker 1: I stupidly left my wallet at home and I need
- 1:00Speaker 1: to order tickets to the screening before it sells out?
- 1:03Speaker 1: What's screening? I'm not? Are you you know? Could you
- 1:09Speaker 1: read me a card number real fast or text me
- 1:13Speaker 1: a pick up your card? I'll then you back. Are
- 1:17Speaker 1: you you talking to me? Your cousin Leslie? Right? Hello? Yeah?
- 1:23Speaker 1: I think we're crossing paths here. You're not answering me
- 1:28Speaker 1: in a weird You're answering me in a weird way.
- 1:37Speaker 1: So what was it like hearing Leslie respond to robot Carol? Well,
- 1:41Speaker 1: it reminded me that it's very easy to prank people
- 1:45Speaker 1: when they have no context for what you're doing. It
- 1:48Speaker 1: took her like a full minute to be like, Okay,
- 1:50Speaker 1: I's tired, not like that's not car you know, it's
- 1:53Speaker 1: always like cool my dad, And I'll say to him
- 1:54Speaker 1: after a minute, Dad, are you playing internet chess? Well,
- 1:59Speaker 1: there is what they all there's tech brain, which is
- 2:01Speaker 1: when someone's texting and talking to you, they're like and
- 2:05Speaker 1: that's sort of what it sounds like. She was like,
- 2:07Speaker 1: are you having another conversation? Has she forgiven you? She's
- 2:11Speaker 1: forgiven robo ka, I'm still not off the hook. Sorry.
- 2:17Speaker 1: Fake audio and fake video can be a lot of
- 2:19Speaker 1: fun for pranks, and there are some life changing the
- 2:22Speaker 1: positive uses for synthetic media that we hear about later.
- 2:26Speaker 1: But just how much trouble could deep fakes get us into?
- 2:30Speaker 1: And as they get easier to make, how can we
- 2:32Speaker 1: keep them out of the hands of the wrong people.
- 2:35Speaker 1: I'm as Velochen, Welcome to Sleepwalkers. The plan originally was
- 2:52Speaker 1: to get cousin Leslie's credit card detail that failed. Yeah,
- 2:56Speaker 1: Julian had the idea of having Kara Ai asked for
- 3:00Speaker 1: credit card information basically to prove how easy it is
- 3:03Speaker 1: to get somebody's credit card information. You can imagine if
- 3:07Speaker 1: it was a little bit better and you were talking
- 3:09Speaker 1: to someone and they were like, oh my god, my grandchild,
- 3:12Speaker 1: you know, needs money. Oh my god, my grandchild is
- 3:14Speaker 1: in trouble, that they would say, Okay, hold on a minute,
- 3:17Speaker 1: I'll get you the credit card number, you know what
- 3:18Speaker 1: I mean. Yeah, And I think that's what's so frightening
- 3:20Speaker 1: about this technology. We're going to dive later into how
- 3:23Speaker 1: you synthesized your voice, but it's the same technological underpinning
- 3:28Speaker 1: of the video that many people have seen of Jordan
- 3:30Speaker 1: Peel basically speaking in Barack Obama's mouth. We're entering an
- 3:35Speaker 1: era in which our enemies can make it look like
- 3:37Speaker 1: anyone is saying anything at any point in time, even
- 3:41Speaker 1: if they would never say those things. For instance, they
- 3:44Speaker 1: could have me say things like President Trump is a
- 3:47Speaker 1: total and complete dipshit. So that was a computer neural
- 3:51Speaker 1: network faking Barack Obama's facial features and mouth movement to
- 3:55Speaker 1: literally look like he was speaking the words that Jordan
- 3:57Speaker 1: Peele said, and that actually makes it even more persuasive
- 4:00Speaker 1: than the fake audio we just heard. Of your voice,
- 4:03Speaker 1: because when you see something, you tend to believe it.
- 4:06Speaker 1: That's why the phrases seeing is believing. Thanks. We're got
- 4:10Speaker 1: to come back to deep fakes, but before we get there,
- 4:13Speaker 1: we're going to take a look at some other online
- 4:15Speaker 1: trickery because the scariest part is that fakes actually don't
- 4:18Speaker 1: have to be as sophisticated as you'll call to cousin
- 4:21Speaker 1: Leslie Toreak Havoc. This is particularly true on Facebook. So
- 4:26Speaker 1: we went to their headquarters in Palo Alto to meet
- 4:29Speaker 1: Nathaniel Glica. He's the head of cybersecurity policy at Facebook,
- 4:33Speaker 1: and he told me about an incident last summer that
- 4:36Speaker 1: creates the true dilemma for him and his team. In July,
- 4:41Speaker 1: we conducted a takedown of a fairly small network of
- 4:44Speaker 1: pages that were operating in the US. Showed links back
- 4:47Speaker 1: to Russian actors, and what they were doing was, among
- 4:50Speaker 1: other things, creating events where they were inviting Americans to
- 4:54Speaker 1: come to protests, and in particular this was around the
- 4:57Speaker 1: Unite the Right two movement, which happened in It was
- 5:00Speaker 1: the anniversary of the bloody clashes in Charlottesville in seventeen,
- 5:04Speaker 1: and the far right wanted to gather again. This time
- 5:08Speaker 1: Russia was watching, and there was an event that popped
- 5:11Speaker 1: up which was the No Unite the Right to movement.
- 5:13Speaker 1: This was a counter protest. There were authentic counter protests
- 5:17Speaker 1: being planned, but this one was being convened by a
- 5:21Speaker 1: group of inauthentic pages and accounts which were linked back
- 5:24Speaker 1: to Russia that were clearly attempting to sort of bring
- 5:27Speaker 1: Americans together in a space where they would go into
- 5:29Speaker 1: physical conflict. Immediately after creating the event, they then went
- 5:32Speaker 1: out and invited legitimate, unwitting activists to co host the
- 5:37Speaker 1: event with them. Let's pause for a moment. This is
- 5:40Speaker 1: Russia we're talking about, and they're creating a Facebook event
- 5:44Speaker 1: to appeal to liberal activists, designed to draw them into
- 5:47Speaker 1: physical conflict with the far right and create the kind
- 5:51Speaker 1: of scenes that tear at our social fabric. But the
- 5:54Speaker 1: people co hosting it are not Russian agitators, their u
- 5:58Speaker 1: S citizens acting in good faith. What we saw in
- 6:01Speaker 1: that case, and what we're increasingly seeing, is these actors
- 6:04Speaker 1: trying to lure their behavior with domestic actors to force
- 6:08Speaker 1: not just the platforms but all of us to ask,
- 6:10Speaker 1: how do you separate these Ultimately, Facebook had to make
- 6:14Speaker 1: a decision. We removed that event from Facebook because it
- 6:19Speaker 1: was created by inauthentic actors. If someone else had created it,
- 6:22Speaker 1: that event would have been fine. So we removed the event.
- 6:25Speaker 1: But then we reached out to the co hosts, the
- 6:27Speaker 1: authentic hosts, and we explained to them what had happened,
- 6:30Speaker 1: and we made clear if you want to host your
- 6:32Speaker 1: own event, you should do that. We just want to
- 6:34Speaker 1: make sure that we everyone understands what's happening and what
- 6:36Speaker 1: did they say, and what was their reaction to realize
- 6:38Speaker 1: that the free will had been manipulated in that way.
- 6:41Speaker 1: If you look at reactions, it's a range of sort
- 6:43Speaker 1: of disbelief, Right, I don't think this was what you're
- 6:46Speaker 1: saying it was, too, I can't believe this happened. To Okay,
- 6:50Speaker 1: that happened, but I strongly believe in this, and I'm
- 6:52Speaker 1: gonna go and I'm going to advocate for my issues
- 6:54Speaker 1: somewhere else. That spectrum of difficulty is exactly clear why
- 7:00Speaker 1: we see actors use these techniques, because there are no
- 7:05Speaker 1: easy answers here. My assumption going into this was that
- 7:08Speaker 1: detecting misinformation would be the biggest challenge for Facebook, but
- 7:12Speaker 1: that's the easy part. It's after you identify the fakes
- 7:15Speaker 1: that the really tough questions begin. We know that, particularly
- 7:19Speaker 1: the government actors in this space, part of their information
- 7:22Speaker 1: dominance strategy is to make themselves appear bigger and more
- 7:26Speaker 1: powerful than they are. They want to seem like they're everywhere,
- 7:29Speaker 1: and it's really easy to see foreign government manipulation under
- 7:33Speaker 1: every rock. I think it's really important not to play
- 7:37Speaker 1: into the hands of these actors and sort of overplay
- 7:41Speaker 1: their own influence. This is attention we struggle with. Whatever
- 7:43Speaker 1: we conduct a takedown for some of these operations, the
- 7:46Speaker 1: most attention it gets is when we take it down.
- 7:50Speaker 1: The entire situation puts Facebook in a catch twenty two.
- 7:54Speaker 1: If they leave the content up, they're helping to promote
- 7:57Speaker 1: a foreign government's nefarious agenda. If they take it down,
- 8:00Speaker 1: the foreign government gets all this attention for being more
- 8:03Speaker 1: powerful and cleverer than they actually are. These decisions are
- 8:07Speaker 1: incredibly hard. Think of Charlottesville, Think of pizza Gate. Think
- 8:12Speaker 1: of Lane Davis, who stabbed his own father after an
- 8:15Speaker 1: argument over the conspiracy theory about liberal pedophiles. Face can kill.
- 8:21Speaker 1: And Facebook has recognized this for a start, they hired
- 8:24Speaker 1: Nathaniel a former cyber crimes prosecutor in the U s
- 8:27Speaker 1: Department of Justice, and in March of this year, Mark
- 8:30Speaker 1: Zuckerberg announced a company wide pivot towards privacy and encrypted messaging,
- 8:35Speaker 1: including services like WhatsApp, which they own. But David Kirkpatrick,
- 8:40Speaker 1: founder of Tachonomy, notes that the pivot carries its own problems.
- 8:45Speaker 1: If you look at South Asia where there's a lot
- 8:48Speaker 1: of ethnic discord political violence, notably in India, Indonesia, Me
- 8:54Speaker 1: and mar Sri Lanka. One of the primary ways that
- 8:59Speaker 1: that spreads is in group messages in WhatsApp. People in
- 9:03Speaker 1: the US don't typically use what'sapp for group messages, but
- 9:06Speaker 1: in places like India and Indonesia they do. And these
- 9:09Speaker 1: groups on five or six people, your your parents, and
- 9:12Speaker 1: your brother and city. These are like you subscribe to
- 9:15Speaker 1: a political leader or a religious zealot. So this is
- 9:19Speaker 1: more like the dear leader being piped into your home right.
- 9:22Speaker 1: So the problem has been almost more severe in those
- 9:26Speaker 1: systems than on Facebook itself of fake news and ethnic
- 9:31Speaker 1: hatred being disseminated. Because WhatsApp is a encrypted service, so
- 9:38Speaker 1: the service itself can't even see what the messages are
- 9:41Speaker 1: that are being distributed. What's scary is it doesn't take
- 9:44Speaker 1: any technical sophistication or knowledge on the part of people
- 9:48Speaker 1: writing these messages spreading this misinformation. They're just using WhatsApp. Yeah,
- 9:53Speaker 1: and these are just messaging apps and social media platforms.
- 9:56Speaker 1: But what they mean is that a single message can
- 9:59Speaker 1: spread like wild And of course the history of new
- 10:01Speaker 1: communication technology tends to go hand in hand with violence.
- 10:05Speaker 1: When the printing press I books came to Europe, they
- 10:09Speaker 1: at least religious wars, but they also made the world literate.
- 10:12Speaker 1: And we've mentioned this before. Technology is usually dual use,
- 10:15Speaker 1: which relates back to deep fikes. Mostly when you read
- 10:18Speaker 1: about deep fakes, probably thanks in part to the fact
- 10:21Speaker 1: they're called deep fakes, the coverage is not very positive.
- 10:25Speaker 1: There have been more and more stories though about positive
- 10:27Speaker 1: uses for deep fakes. So when we come back, I'm
- 10:29Speaker 1: gonna tell you more about how I faked my own
- 10:31Speaker 1: voice and also some of the things that I learned
- 10:33Speaker 1: in the process. We started this conversation a few weeks ago,
- 10:44Speaker 1: and then you asked those to create these artificial voice
- 10:48Speaker 1: based on your identity. That's Jose's to tell out, the
- 10:52Speaker 1: co founder of liar Bird. They're the company who made
- 10:55Speaker 1: robot Kara and helped me prank my cousin, and they've
- 10:57Speaker 1: published a version of their tools online at liarbird dot ai.
- 11:01Speaker 1: Here's how it works. I know it might sound a
- 11:04Speaker 1: bit like magic, but in reality, the way that our
- 11:08Speaker 1: algorithms work is basically they are just a pattern matching algorithms,
- 11:12Speaker 1: and so it's trying to figure out how to identify
- 11:18Speaker 1: the patents in your voice by comparing it against thousands
- 11:21Speaker 1: of other voices I should have tens of thousands of
- 11:23Speaker 1: other voices, and trying to figure out what is it
- 11:27Speaker 1: that makes your voice unique. Once Jose's algorithms identified what
- 11:30Speaker 1: was unique about my voice, obviously everything they had the
- 11:33Speaker 1: building blocks they needed to make a fake. Then we
- 11:36Speaker 1: sent Jose a set of sentences we wanted robot care
- 11:39Speaker 1: to say, and he used another set of algorithms to
- 11:42Speaker 1: turn the text into what we heard. The way they
- 11:44Speaker 1: do this is they use it's called a generative adversarial
- 11:47Speaker 1: network again, which is a system where one neural net
- 11:51Speaker 1: tries to trick another one a thousand times per second.
- 11:55Speaker 1: So each time the second network to tacks of fake,
- 11:58Speaker 1: the first one tries again It basically learns from its mistakes,
- 12:01Speaker 1: and once it tricks its adversary, it's ready to show
- 12:04Speaker 1: its results. In our case, liar Bird pits my fake
- 12:08Speaker 1: voice against my real voice until it sounds like this
- 12:11Speaker 1: sub dog Scara. As this technology becomes more widely available,
- 12:17Speaker 1: so does the potential for abuse. And while Liarbird develops
- 12:20Speaker 1: the technology, they don't take the ethics lightly. But Jose
- 12:24Speaker 1: has an entirely different fear. We believe that the biggest
- 12:28Speaker 1: risk of this kind of technology comes from the fact
- 12:31Speaker 1: that not a lot of people know about it. I
- 12:35Speaker 1: believe that society is not ready for what's going to
- 12:38Speaker 1: happen when this technology becomes widespread, and so I really
- 12:42Speaker 1: want to make my best effort in trying to showcase
- 12:47Speaker 1: it to the public so that they are at least
- 12:49Speaker 1: prepare for what's coming. When people know a scheme exists,
- 12:53Speaker 1: they're less likely to be tricked by it. But if
- 12:55Speaker 1: you don't know deep fakes are possible, you're much more
- 12:58Speaker 1: likely to fall for them. Leslie might been better equipped
- 13:00Speaker 1: to call my bluff had she known it was even possible.
- 13:04Speaker 1: But here's the thing, Well, there are inevitable misuses of
- 13:07Speaker 1: deep fix both behind US and on the horizon. There
- 13:10Speaker 1: are a number of extraordinary benefits of this technology, which
- 13:13Speaker 1: is why Jose is working on it. When people are
- 13:15Speaker 1: diagnosed with a LIST, it's because they start to lose
- 13:18Speaker 1: their movement skills seemed, let's say, their hands or their feet,
- 13:21Speaker 1: and so they go to the doctor and then the
- 13:24Speaker 1: doctor tells them like, you know what, this can be
- 13:27Speaker 1: als and this gets progressively worse. This was the case
- 13:31Speaker 1: for Pat Quinn, the co founder of the Ice Bucket Challenge,
- 13:36Speaker 1: creating a real fight within the a LS community. This
- 13:41Speaker 1: is a public battle now. Pat was diagnosed with a
- 13:46Speaker 1: l S and it ultimately took his ability to speak,
- 13:49Speaker 1: walk and use his hands. During this time. Since they're
- 13:52Speaker 1: diagnosed until they lose their voice, they have some time,
- 13:56Speaker 1: and so the idea is that during this time they
- 13:59Speaker 1: will be able to record themselves, ideally in a really
- 14:03Speaker 1: high quality setting. Then based on these recordings, we will
- 14:06Speaker 1: be able to create an artificial copy of their voice
- 14:09Speaker 1: which they will be able to continue using for the
- 14:12Speaker 1: rest of their life. Liar Bird has partnered with the
- 14:15Speaker 1: a l S Foundation to create Project Revoice. Just imagine
- 14:19Speaker 1: how it would feel for them, to, let's say, not
- 14:22Speaker 1: be able to tell their husband or their wife I
- 14:25Speaker 1: love you anymore, to tell this to their kids. And
- 14:28Speaker 1: so using this technology, they are able to keep this
- 14:33Speaker 1: really important part of their identities. Using the exact same
- 14:37Speaker 1: technology I used to create my deep fake liar bird
- 14:40Speaker 1: was able to give Path the ability to preserve his
- 14:42Speaker 1: voice for the rest of his life. It's a strange
- 14:46Speaker 1: feeling saying the first words at the second time. It's
- 14:49Speaker 1: like you don't realize how powerful, how personally I her
- 14:55Speaker 1: voice really is until it stay them from you. My
- 14:58Speaker 1: voice is how I back against a very disease. Take
- 15:02Speaker 1: it for me, say something, Listen to it. No voice.
- 15:10Speaker 1: Since revoicing pat Lierbard has received a number of emails
- 15:13Speaker 1: from als patients asking if it's possible for them to
- 15:16Speaker 1: do the same thing, preserve this part of themselves which
- 15:19Speaker 1: they know they're going to lose, and Jose has heard
- 15:22Speaker 1: from people who have lost family in other ways. For instance,
- 15:25Speaker 1: we have received that quite a lot, actually very emotional
- 15:28Speaker 1: emails about people telling some variation of this. My wife
- 15:33Speaker 1: died three months ago, and I have two children, age
- 15:38Speaker 1: four and six, and I would really really love to
- 15:41Speaker 1: be able to tell them a good night story in
- 15:44Speaker 1: the voice of their mother, or to tell them that,
- 15:46Speaker 1: in the mother's voice, I love you, I am proud
- 15:48Speaker 1: of you, be happy. The tools on lierbrard AI are
- 15:53Speaker 1: intentionally less advanced and meant to just spread awareness, but
- 15:57Speaker 1: Liarbrard's more bespoke tools open amazing possibilities for changing how
- 16:01Speaker 1: we deal with loss and grief. I would like to
- 16:03Speaker 1: ask you just one question, which is like, how would
- 16:07Speaker 1: you feel, let's say, about recording the voice of your
- 16:10Speaker 1: parents and keeping them What do you think would you
- 16:13Speaker 1: like to the lease or or how do you feel
- 16:14Speaker 1: about that? It was interesting when jose asked me because
- 16:17Speaker 1: I had actually thought about it ever since I learned
- 16:19Speaker 1: about Liar Bird when I was fifteen. So fourteen years ago,
- 16:26Speaker 1: my dad died in a fatal car accident, and nobody
- 16:31Speaker 1: prepares for accidents, you know. One minute my dad walked
- 16:34Speaker 1: out the door, and forty five minutes later the police
- 16:37Speaker 1: showed up at the same door to tell us what happened,
- 16:40Speaker 1: and so I never got to see or speak to
- 16:43Speaker 1: my dad ever again. Sometimes my therapist will ask me
- 16:49Speaker 1: if I think about what I would talk about with
- 16:51Speaker 1: my dad if he was still alive, and I always
- 16:54Speaker 1: say that, you know, I don't. I don't think about
- 16:56Speaker 1: that too much because it's sad to think about at
- 17:00Speaker 1: that because he's not actually around, and because I know
- 17:02Speaker 1: I can't talk about him. But it's also hard to
- 17:05Speaker 1: conceive of. You know, I can't recall off the top
- 17:08Speaker 1: of my head what he sounds like, and sometimes I'll
- 17:12Speaker 1: hear his voice when we watch home movies and it
- 17:14Speaker 1: always spooks me out. So the idea of having his
- 17:17Speaker 1: disembodied voice asked me things like how do you like
- 17:21Speaker 1: working on this podcast? Or what's the most amazing thing
- 17:24Speaker 1: you've learned, or even saying things like I'm so proud
- 17:28Speaker 1: of you. Do you know that? I'm not sure how
- 17:31Speaker 1: I'd react to his voice like that. Regardless, the thought
- 17:35Speaker 1: of it is something in the realm of possibility is
- 17:38Speaker 1: equal parts chilling and exciting. I actually think, given the chance,
- 17:43Speaker 1: I might do it. This is not the science fiction
- 17:51Speaker 1: thing or something that will exist years from now. It's
- 17:55Speaker 1: something that exists already to the people can even go on.
- 17:58Speaker 1: And as my cousin Leslie learned, these deep fakes are
- 18:02Speaker 1: already good enough to use on unassuming family members suburb
- 18:07Speaker 1: This isn't Kara, this is artificial terra. Oh my God,
- 18:13Speaker 1: my voice right now, it's AI. This is awful. When
- 18:26Speaker 1: we started reporting on deep fakes, I never anticipated how
- 18:29Speaker 1: moving the technology could be. I was more focused on
- 18:32Speaker 1: the dangers, and they are worth considering too. One person
- 18:35Speaker 1: who is out in front bringing awareness to the potential
- 18:38Speaker 1: harms of fake media is Danielle Citron. She's a legal
- 18:41Speaker 1: professor at the University of Maryland and the author of
- 18:44Speaker 1: Hate Crimes in Cyberspace. Machine learning technology and neural networks
- 18:48Speaker 1: can learn from your photo and voice that's taken from
- 18:52Speaker 1: recordings of your voice, can sufficiently learn enough about your
- 18:57Speaker 1: face and the way it moves and you or voice
- 19:00Speaker 1: so that it can create really incredibly difficult to debunk
- 19:06Speaker 1: videos of you doing and saying things you never did.
- 19:10Speaker 1: Now we all know how dangerous the simple recent word
- 19:13Speaker 1: can be. Danielle got interested in how fake video could
- 19:17Speaker 1: increase the forces of hate exponentially. There was a whole
- 19:21Speaker 1: Reddit thread devoted to deep fake sex videos of celebrities,
- 19:25Speaker 1: female celebrities like Emma Watson and Hathaway and others. If
- 19:28Speaker 1: you went through the thread, which I did, you can
- 19:31Speaker 1: see the conversation moving beyond Emma Watson to my bitch
- 19:35Speaker 1: girlfriend or that woman I hated in high school, and
- 19:38Speaker 1: it was it was all the conversation about women, you
- 19:40Speaker 1: know what I thought was like the evil Cyberus stocking
- 19:43Speaker 1: was all based on Crewe doctored photos of someone naked,
- 19:46Speaker 1: but if you worked at it, you could figure it out.
- 19:48Speaker 1: Now we can put people into pornography in ways that
- 19:52Speaker 1: are devastate their careers. So, Kara, I do think it
- 19:56Speaker 1: says something that this new technology is being used to
- 19:59Speaker 1: target women. And a lot of these conversations are happening
- 20:02Speaker 1: on the same forums on Reddit where the in cell
- 20:04Speaker 1: movement was born, right, So I think this is especially
- 20:07Speaker 1: important when we talk about famous women and their likeness.
- 20:11Speaker 1: A lot of men on the Internet want to see
- 20:13Speaker 1: their favorite actresses in positions that they wouldn't be able
- 20:17Speaker 1: to see those actresses in, and so with this technology,
- 20:22Speaker 1: it's quite easy to put someone's face on somebody else's
- 20:25Speaker 1: body without the consent of the actual actress and actually sag.
- 20:29Speaker 1: The Screen Actors Guild held a panel a few weeks
- 20:32Speaker 1: ago to bring this up that like, yes, we we're
- 20:34Speaker 1: talking about this in terms of democracy and our political
- 20:37Speaker 1: system and the upcoming election, but we also have to
- 20:39Speaker 1: talk about this in terms of the livelihood of women
- 20:42Speaker 1: who make money on their likeness and whose likeness is
- 20:45Speaker 1: now being misappropriated. Yeah, because it can destroy their careers
- 20:49Speaker 1: and silence them. There's actually a case in India where
- 20:53Speaker 1: people attempt to use deep fate pornography to intimidate and
- 20:57Speaker 1: silence a journalist called run ayub and I spoke about
- 21:00Speaker 1: that case with Danielle the Indian journalist who had been
- 21:05Speaker 1: very critical of Hindu politics, nationalist politics, and a deep
- 21:10Speaker 1: fake sex video sort of was spread basically to discredit
- 21:13Speaker 1: her um and spread through texting networks and went viral,
- 21:19Speaker 1: and she basically was devastated and went offline, stopped writing
- 21:23Speaker 1: for like three weeks. She's a journalist, this is what
- 21:26Speaker 1: she does for a living, right, So imagine that kind
- 21:29Speaker 1: of granular individual harm um and compare it with harm
- 21:33Speaker 1: to CEOs the night before an I p O. A
- 21:36Speaker 1: deep fake is release that shows this person taking a
- 21:39Speaker 1: bribe or doing drugs or whatever. I'm making it up,
- 21:41Speaker 1: but that tanks the I p O. Right. This kind
- 21:45Speaker 1: of video manipulation used to be confined to places like Disney,
- 21:48Speaker 1: and the output was blockbuster movies that are fictional but
- 21:51Speaker 1: not fake. Now AI is being consumerized, and the tools
- 21:55Speaker 1: to create convincing video are spreading, and that means creating
- 21:59Speaker 1: the kind of chaos Danielle describes is also more and
- 22:02Speaker 1: more accessible. That threatens all of us. One person working
- 22:07Speaker 1: on the issue is Hani for Reed of Dartmouth University,
- 22:10Speaker 1: who has been called the father of digital forensics. I'm
- 22:14Speaker 1: concerned that once we know you can create fake content,
- 22:18Speaker 1: there is nothing stopping anybody from saying that any video
- 22:22Speaker 1: is fake. Everybody has plausible the liability. So rewind two
- 22:26Speaker 1: years ago when the Access Hollywood tape came out of
- 22:29Speaker 1: President Trump saying what he does to women. The response
- 22:33Speaker 1: from the campaign was not this is fake. It was
- 22:35Speaker 1: we apologized, this was locker room talk. They found ways
- 22:39Speaker 1: of trying to excuse it. If that was today, guaranteed
- 22:44Speaker 1: he would have said it was fake. And in fact,
- 22:45Speaker 1: a year ago, after having apologized for the for the
- 22:48Speaker 1: audio recording, he said it was fake. And so now
- 22:51Speaker 1: politicians have plausible deniability, and at a time when our
- 22:55Speaker 1: US president is demonizing the press and telling everybody that
- 22:58Speaker 1: you can't believe anything, that credible deniability holds some weight.
- 23:02Speaker 1: And so I'm extremely concerned. Now, how do we distinguish
- 23:05Speaker 1: what's what, and that I think for a democracy is
- 23:08Speaker 1: going to be incredibly challenging. So when nothing is believable,
- 23:12Speaker 1: the mischief doer can say it's a lie. Do you
- 23:15Speaker 1: know what I'm saying? Like the person who commits the
- 23:17Speaker 1: crime or does something and says something incriminatory can say,
- 23:21Speaker 1: that's a fake. So the more you educate you both
- 23:24Speaker 1: but deep fix the evil doers can leverage that and say, well,
- 23:29Speaker 1: you can't believe anything, right. Danielle calls this the liar's dividend.
- 23:35Speaker 1: In a world where nothing can be trusted, everything can
- 23:38Speaker 1: be denied, and even documented bad deeds can be explained away.
- 23:42Speaker 1: This kind of thing is accelerated by deep fakes, though,
- 23:44Speaker 1: which is why I think there are some attempts to
- 23:47Speaker 1: correct it with law, with law like the Anti Deep
- 23:50Speaker 1: Fakes Law, very similar the Malicious Deep Fake Prohibition Act
- 23:54Speaker 1: of eighteen, which was introduced by this Republican center from
- 23:58Speaker 1: Nebraska named Ben Sas and it basically aims to outlaw
- 24:02Speaker 1: fraud in connection to audio visual records. But I don't
- 24:06Speaker 1: know if this law will path. In any case, not
- 24:08Speaker 1: all deep fakes are malicious, and so we have to
- 24:11Speaker 1: be careful with laws which are too broad. As we
- 24:15Speaker 1: heard in your Liabird piece, there are so amazingly positive
- 24:17Speaker 1: applications of deep fake technology. Here's Honi for read talking
- 24:22Speaker 1: about deep fakes and the movie business. Can you imagine
- 24:25Speaker 1: a world where the actor can simply license their appearance
- 24:29Speaker 1: and they never have to show up on the set.
- 24:31Speaker 1: You say, look, here's a bunch of images of me.
- 24:32Speaker 1: Synthesize me doing whatever you want. I'm basically an animated
- 24:35Speaker 1: character for you, and then anybody can be in the movies.
- 24:38Speaker 1: You can imagine customized movies. Imagine I go to the
- 24:40Speaker 1: movie and say, look, I'd like to see this movie,
- 24:43Speaker 1: but with George Clooney and not Kevin Spacey in it.
- 24:45Speaker 1: Please synthesize that for me. Can we do that today
- 24:48Speaker 1: or tomorrow now? But in theory, that is essentially where
- 24:52Speaker 1: we're going. So if you if you haven't seen, some
- 24:54Speaker 1: of these people are creating all these deep fake videos
- 24:56Speaker 1: of Nick Cage and inserted into all these different movies.
- 24:59Speaker 1: That's essentially and that's not the full length movie, they're
- 25:01Speaker 1: doing it into clips, but that's essentially the trend where
- 25:04Speaker 1: you can just put your favorite actor or actress into
- 25:05Speaker 1: whatever movie you want and just watch it. It's personalized movies.
- 25:09Speaker 1: I'm not gonna lie. I find it super weird that
- 25:12Speaker 1: Nicholas Cage has become the posted boy for having his
- 25:15Speaker 1: face deep faction to various movies. I wonder if you
- 25:18Speaker 1: actually asked Internet nerds, why neck Cage, what do you
- 25:22Speaker 1: think would like? I have no idea. Well, he's kind
- 25:24Speaker 1: of already a meme, right, he was, and he was
- 25:26Speaker 1: in face off where his face was switched with another
- 25:29Speaker 1: person's face, So he's always sort of in the poster
- 25:31Speaker 1: child for face swapping, you know. I think actually one
- 25:34Speaker 1: thing that I thought about is this idea of representation.
- 25:37Speaker 1: You know, if there's a movie or movies or series
- 25:40Speaker 1: like James Bond where the lead character has been historically white,
- 25:44Speaker 1: and you want to show your African American son James Bond,
- 25:49Speaker 1: it would be kind of cool to make James Bond black, right,
- 25:53Speaker 1: because then your child could be watching a movie where
- 25:55Speaker 1: James Bond looks like your child. Absolutely, And I think
- 25:59Speaker 1: one of the big pro BOMs in the movie business
- 26:01Speaker 1: and the media business in general is representation. So more
- 26:04Speaker 1: people do have access to this technology now, but it
- 26:08Speaker 1: used to be that only a Hollywood special Effects company
- 26:11Speaker 1: would have access to this technology. When you remove the gatekeepers,
- 26:15Speaker 1: you get these incredible explosions of culture, but you also
- 26:19Speaker 1: get real threats to the social fabric. And so in
- 26:22Speaker 1: the case of deep fakes, and they're all very well
- 26:25Speaker 1: when they're labeled as fake or when we know they're fake,
- 26:28Speaker 1: but when they're posing as real, that's when we start
- 26:32Speaker 1: to be really under threat, I think as a society.
- 26:35Speaker 1: But there are people working on this as ever. Cat
- 26:38Speaker 1: and Mouse. When we come back, we'll talk about some
- 26:41Speaker 1: of the ways they're fighting back. When it comes to
- 26:50Speaker 1: deep fakes. Pandora's box is open, and as Jose argues,
- 26:54Speaker 1: there's no turning back the clocks. The technology exists. So,
- 26:58Speaker 1: knowing deep fakes and fake news have become more sophisticated,
- 27:01Speaker 1: I wanted to find out how actual news organizations are
- 27:04Speaker 1: thinking about the problem. So I spoke with John mccathwaite,
- 27:07Speaker 1: editor in chief of Bloomberg News, and he actually started
- 27:10Speaker 1: by pointing out that fate news isn't new news. I
- 27:14Speaker 1: think that one crucial thing when you look at fake
- 27:16Speaker 1: news is tom it's always been there. You know. The
- 27:18Speaker 1: first bit of fate news was the trojan horse fake
- 27:21Speaker 1: news and propaganda have for ever been some of the
- 27:24Speaker 1: more exotic weapons in global conflict. John points to another
- 27:28Speaker 1: example involving the famous British by and author of James
- 27:32Speaker 1: bond Ian Fleming. Supposedly one of his great schemes was
- 27:36Speaker 1: to drop lots of jumbo sized condoms over Germany um
- 27:42Speaker 1: and label sort of British small on the outside in
- 27:46Speaker 1: the end in the name, with the aim, no doubt wrongly,
- 27:49Speaker 1: of destabilizing German manhood. My point is that there are
- 27:53Speaker 1: many many ways in which you can do this. But
- 27:56Speaker 1: the most interesting thing to me about fate news is
- 28:00Speaker 1: that really in modern history it's tied very heavily to technology.
- 28:04Speaker 1: What tends to happen is a new technology comes along
- 28:07Speaker 1: which suddenly sets media free. If we look to history
- 28:10Speaker 1: we can understand this moment better. We mentioned the early
- 28:14Speaker 1: printing press before and how it enabled explosions of ideology
- 28:17Speaker 1: and led to religious conflicts. Well, when the printing press
- 28:21Speaker 1: was industrialized in the nineteenth century, there was another fake
- 28:24Speaker 1: news boom. Go back to I think it's no one.
- 28:28Speaker 1: You have the invention of the steam press in London,
- 28:30Speaker 1: and what that does It enables people to multiple by
- 28:33Speaker 1: ten the amount of paper that you can print suddenly
- 28:37Speaker 1: all the way across Europe, and then in America, free
- 28:40Speaker 1: ship newspaper starts bringing up. Because you can distribute far more,
- 28:44Speaker 1: you can reach far more people, far more quickly. And
- 28:47Speaker 1: the most notorious of this was The New York Sun
- 28:49Speaker 1: at one time, I think, the world's biggest selling paper,
- 28:53Speaker 1: run by Michael Benjamin Day, and he would run some
- 28:56Speaker 1: stories like the moon was populated by people who are
- 28:59Speaker 1: half human half bat. But what happened, and I think
- 29:03Speaker 1: this will happen again, is that consumers said, we don't
- 29:07Speaker 1: want to read that, we need facts. And so if
- 29:09Speaker 1: you look back at many of the big newspapers of
- 29:12Speaker 1: our time, the New York Times, the Economists were I
- 29:14Speaker 1: used to work, many of these things came from that
- 29:17Speaker 1: particular period because people paid more to get things they trusted. Well,
- 29:21Speaker 1: that is definitely happening again. In other words, most of
- 29:24Speaker 1: the high quality press today, the New York Times, the Economist,
- 29:28Speaker 1: which John also edited, came from consumer demand for trustworthy information.
- 29:33Speaker 1: And that same consumer demand may help us out of
- 29:36Speaker 1: today's predicament. But there is one key difference. Now we
- 29:40Speaker 1: have deep fakes. It's worth a lot of money to
- 29:43Speaker 1: a lot of people to try to fool us. So
- 29:45Speaker 1: you look at things like Twitter handles that aren't quite
- 29:47Speaker 1: the same, some mixture between humans and computers. You used
- 29:50Speaker 1: to deal with those. What is harder at the moment
- 29:54Speaker 1: is video. So to give you an example, I think
- 29:57Speaker 1: a year or so ago, there was some tack in
- 29:59Speaker 1: a subway in New Org. We could verify really quite
- 30:01Speaker 1: quickly that the subway attack had happened, but almost immediately
- 30:05Speaker 1: there was a picture on Twitter of one of the
- 30:08Speaker 1: alleged assailants lying in a pool of blood. Now, trying
- 30:12Speaker 1: to verify that that was true was much harder, and
- 30:15Speaker 1: it came down to things like working out whether that
- 30:17Speaker 1: was the correct subway floor. You can look at pixels,
- 30:20Speaker 1: you can look at all those different things, But yes,
- 30:23Speaker 1: verifying video is often harder than verifying facts. Do you
- 30:26Speaker 1: have any tools or technologies that you're licensing or spending
- 30:29Speaker 1: money on to do it. We spend a lot of
- 30:31Speaker 1: money on technology across all these fronts. With more and
- 30:34Speaker 1: more news coming directly from social media, large news organizations
- 30:37Speaker 1: like Bloomberg News need to be able to verify which
- 30:41Speaker 1: photos and videos are real and whether they actually relate
- 30:44Speaker 1: to the events they're investigating, which is why Harni for
- 30:47Speaker 1: Reid is in such high demand. Suddenly, the need to
- 30:51Speaker 1: authenticate content has really global implications. Everything from our court
- 30:56Speaker 1: to our national security, to our democratical actions to citizens
- 31:00Speaker 1: safety is starting to rely on our ability to tell
- 31:03Speaker 1: the real from the fake. And so I think this
- 31:05Speaker 1: field of forensics, this field of authentication, has never been
- 31:08Speaker 1: more important, and that's what Harney spends his days working
- 31:11Speaker 1: on at Dartmouth. He develops techniques to analyze and authenticate
- 31:16Speaker 1: digital media. Ahead of the elections, He's working on what
- 31:20Speaker 1: he calls a soft biometric tool to detect fake videos
- 31:24Speaker 1: of specific politicians such as Bernie Saunders, Elizabeth Warren, and
- 31:29Speaker 1: Donald Trump. UM I would say the game is going
- 31:32Speaker 1: to be that we never eliminate the ability to create
- 31:36Speaker 1: fake content, but what we do is we raise the bar.
- 31:39Speaker 1: We take it out of the hands of the amateurs,
- 31:41Speaker 1: we take it out of the hands of the average
- 31:43Speaker 1: person downloading some code, and we make it more difficult,
- 31:46Speaker 1: more time consuming, and more risky. And this is the
- 31:48Speaker 1: same thing that we do with counterfeit currency. You can
- 31:52Speaker 1: still create counterfeit currency today, but it's really hard, still
- 31:55Speaker 1: a risk, but it's a more manageable risk. On the
- 31:57Speaker 1: subject of money, there are digital orencies which are much
- 32:00Speaker 1: more difficult to counterfeit than coins and banknotes. You've heard
- 32:04Speaker 1: of Bitcoin and ethereum, which you're enabled by blockchain, a
- 32:07Speaker 1: so called distributed ledger. Information about transactions is shared between
- 32:12Speaker 1: all the users of the currency, rather than authenticated and
- 32:15Speaker 1: guarded by a bank. Sharing this kind of information across
- 32:19Speaker 1: a crowd of people with multiple backup copies has a
- 32:23Speaker 1: range of uses. One thing Harney is looking at is
- 32:26Speaker 1: using blockchain to authenticate images and videos. At source, we're
- 32:31Speaker 1: gonna start seeing, um the use of a different type
- 32:35Speaker 1: of camera. So there are now companies out there that
- 32:38Speaker 1: create what are called secure imaging pipelines, and so when
- 32:42Speaker 1: you record an image or video, they extract a unique
- 32:46Speaker 1: signature from that content, they cryptographically signed it, and they
- 32:50Speaker 1: put that on the blockchain. So that's basically a distributed
- 32:53Speaker 1: ledger that's very very hard, if not impossible, to manipulate.
- 32:56Speaker 1: Perhaps staying ahead of the perpetrators and making fakes more
- 32:59Speaker 1: differ are cooled is the best we can do. But
- 33:02Speaker 1: what about our usage? How much responsibility do we have
- 33:06Speaker 1: to navigate the web thoughtfully? And how much responsibility should
- 33:09Speaker 1: be on the platforms. We have Facebook, Twitter, we have
- 33:13Speaker 1: yelped because they're not responsible for user generated content. What's
- 33:17Speaker 1: interesting is that, like Nathaniel at Facebook, Danielle also sees
- 33:22Speaker 1: risks in over zealous moderation. If you put too much
- 33:26Speaker 1: responsibility on the platform, you will likely incentivize over censorship.
- 33:31Speaker 1: So all the great things that we think about a
- 33:33Speaker 1: lot of these platforms, and especially the social media, the
- 33:37Speaker 1: Parkland survivors or Black Lives Matters, right, we don't want
- 33:40Speaker 1: to lose the facility and new enablements for organizing and speech.
- 33:44Speaker 1: So if you put too much liability on the platforms,
- 33:46Speaker 1: they're going to overreact and anything anyone complains about and
- 33:50Speaker 1: have very aggressive filters. So we might very well miss
- 33:53Speaker 1: Black Lives Matter, we might not have Parkland and never
- 33:57Speaker 1: see it because you're gonna have overly are of censorship.
- 34:01Speaker 1: Here's Nathaniel again. Whenever people come together in a new medium,
- 34:05Speaker 1: you're going to have people that try to manipulate and
- 34:07Speaker 1: try to take advantage. I think one of the things
- 34:09Speaker 1: that's really fundamentally true that we have done when we
- 34:12Speaker 1: think about the Internet generally social media as well, is
- 34:15Speaker 1: we've removed some of the traditional gatekeeping mechanisms that have
- 34:18Speaker 1: existed in the past, and that has meant that far
- 34:22Speaker 1: more people could engage, much more quickly and much more
- 34:25Speaker 1: vocally than ever before, and that has led to some
- 34:27Speaker 1: incredible things. If you think about the me too movement,
- 34:29Speaker 1: which really part of what drives it and enables it
- 34:32Speaker 1: is the ability to route around some of those gatekeepers, right,
- 34:35Speaker 1: But at the same time, you're also going to see
- 34:37Speaker 1: malicious actors try to misuse that. I think that is
- 34:39Speaker 1: a fundamental truth for any form of media. The question
- 34:42Speaker 1: is how do you enable authentic engagement while making the
- 34:48Speaker 1: types of manipulation that we see more difficult. If Facebook
- 34:53Speaker 1: and other platforms are too destructive of society, ultimately everyone loses,
- 34:58Speaker 1: even the technology company needs and their shareholders. So how
- 35:02Speaker 1: do we move from understanding that to finding solutions. Here's
- 35:06Speaker 1: David Kirkpatrick again. If we are going to retain democracy,
- 35:12Speaker 1: we need technical systems, digital systems, technologies that more effectively
- 35:19Speaker 1: and persuasively compelling lee distribute knowledge so that we have
- 35:25Speaker 1: citizens that are capable of functioning in a democratic landscape
- 35:30Speaker 1: that is more complex, more rapidly changing, and ultimately more global.
- 35:35Speaker 1: And as far as Hanni ferred is concerned, this has
- 35:38Speaker 1: become everyone's problem, so we all have a part to
- 35:42Speaker 1: play in solving it. I think two things are gonna
- 35:44Speaker 1: have to change. So one is the technology to authenticate
- 35:46Speaker 1: it is going to have to get better. So whether
- 35:48Speaker 1: that's authenticating at the source or the types of things
- 35:50Speaker 1: that I do with authenticating content and operating that at scale,
- 35:54Speaker 1: that's going to have to get better. But I think
- 35:56Speaker 1: what's also going to have to change is how we
- 35:58Speaker 1: as consumers of digital content think about what we see.
- 36:01Speaker 1: We are going to have to become more critical, more reasoned.
- 36:05Speaker 1: We have to get out of our echo chambers. We
- 36:07Speaker 1: have to stop allowing social media to manipulate us in
- 36:09Speaker 1: the way that they do. So I think the solution
- 36:12Speaker 1: is at least too prong and potentially three with some
- 36:14Speaker 1: legislative relief on the line to really force the companies
- 36:18Speaker 1: to do better than they have been over the last
- 36:19Speaker 1: few years. So does the good outweigh the bad? I
- 36:23Speaker 1: don't know. We have to have a hard conversation. People
- 36:25Speaker 1: in who work in infectious disease and physicists who develop weaponry,
- 36:30Speaker 1: they think about this all the time. We as technologies
- 36:32Speaker 1: have not quite thought about this as much in the
- 36:34Speaker 1: past because our field is so young. But I think
- 36:37Speaker 1: now you know it's time to wake up and start
- 36:39Speaker 1: asking those hard questions and having those conversations before it's
- 36:42Speaker 1: too late. Once again, we're being aged to wake up
- 36:47Speaker 1: from our sleep book, and we do have some months
- 36:50Speaker 1: is at least when it comes to deep fakes, we
- 36:53Speaker 1: can make it akin to counterfeiting money. The people who
- 36:56Speaker 1: do it will get prosecuted, and program is like Honey
- 36:59Speaker 1: will work on detection technology, but we still have to
- 37:02Speaker 1: hold the bills up to the light before we decide
- 37:05Speaker 1: whether to accept them. That's our job, that is, if
- 37:08Speaker 1: we're not too busy watching Nicholas Cage starring as Thelma
- 37:12Speaker 1: and Louise in Thelma and Louise. Even more complicated than
- 37:16Speaker 1: deep fakes is the concentration of power at companies like Facebook.
- 37:21Speaker 1: In the next episode, we visit a secret lab at
- 37:23Speaker 1: Google to understand what happens when technology companies start taking
- 37:27Speaker 1: on the role of the state, and we speak with
- 37:30Speaker 1: Lena Khan, who has proposed new regulation to balance the
- 37:33Speaker 1: power of big technology companies like Amazon. I'm asloshin See
- 37:38Speaker 1: you next time. Sleepwalkers is a production of I Heart
- 37:53Speaker 1: Radio and Unusual productions. For the latest AI news, live interviews,
- 37:58Speaker 1: and behind the scenes footage, find us on Instagram, at
- 38:01Speaker 1: Sleepwalker's podcast or at Sleepwalker's podcast dot com. Sleepwalkers is
- 38:06Speaker 1: hosted by me Ozveloshin and co hosted by me Kara Price.
- 38:09Speaker 1: Were produced by Julian Weller with help from Jacobo Penzo
- 38:12Speaker 1: and Taylor Chacogne. Mixing by Tristan McNeil and Julian Weller.
- 38:16Speaker 1: Recording assistance this episode from tofarrelf Our Story editor is
- 38:20Speaker 1: Matthew Riddle. Sleepwalkers is executive produced by me Ozveloshin and
- 38:24Speaker 1: Mangesh Hattiga. For more podcasts from My Heart Radio, visit
- 38:28Speaker 1: the I heart Radio app, Apple Podcasts, or wherever you
- 38:31Speaker 1: listen to your favorite shows.