Ep. 94: Interview - AI Facial Recognition Technology and the Death of Anonymity. A conversation with Kashmir Hill, NY Times journalist and author of "Your Face Belongs to Us."

[00:00] Subscribe to the podcast. Subscribe to the newsletter. [00:02] We're really fun. We're so fun. We're the funnest. [00:32] We do that through newsletters and podcasts and events and a group chat where we hang out on Discord and talk about skincare and technology and all of the things that we do look at the intersection of technology and culture. [00:45] What do you think? Nice. Checks out. [00:49] Welcome to Boys Club Interviews. This is where we bring on people who are much smarter than us to talk about what's happening on the internet. And wow, today... [00:58] really delivered on that prompt. [01:02] Who do we have on? We had Kashmir Hill on the pod. Kashmir is a New York Times business reporter covering technology and privacy. She's been working at the intersection of technology and privacy since... [01:17] 2009, she's written a book. [01:20] called Your Face Belongs to Us [01:22] which is about facial recognition technology and a startup that is working on that. And she is insanely credible and it like just the real deal. So, so, so smart. So enjoyed talking to her. I was so nervous at first because I was just like, this person is so bright and knows something about... I was like...

[01:44] just struck by how much she knows about the world that I don't know anything about. And she was incredibly generous in her sharing in that knowledge. So really great episode. Truly [01:57] I make this joke in the podcast, but I had no concept for how pervasive and urgent... [02:05] facial recognition technology was to our lives today. And... [02:09] Now I am. I'm aware. Even if... [02:14] you are thinking about, man, I don't care about my privacy because I'm really careful about what I post, or I'm really careful about what I send. There's all of these other examples that really brought to life [02:25] what it's like to be able to be recognized online and what that does to your anonymity and your life. And it's really crazy. Here are just some rapid fire ones. You go to a protest. You think you're anonymous in a crowd. Someone takes a photo, scans the faces, is able to identify you and like dox you. You are walking out of Planned Parenthood for whatever reason. Someone takes a photo of you. [02:51] And... [02:52] uses this technology and doxes you. You're at a bar and you're just hanging out with some friends and there's a creepy stranger there who [03:01] takes a photo of you and then is able to find out all your personal information. You accidentally bump into somebody on the subway and they like to destroy people online and they find you. Have like a venge kink. Yeah. Yeah. And start to populate that on the internet. It's really dark, but I do think it's important to talk about because one of the things that we talked about is that the news cycle has been dominated by generative AI. And it's

[03:28] reading about facial recognition, AI software, I just cannot believe that this is not the thing that we're all talking about when it comes to policing, when it comes to average people who their identities are being scraped and imported into these applications. It's absolutely crazy. So big thanks to Cash. She's incredible. [03:48] Please follow her. [03:50] Buy her book. She doesn't need us to tell you that, but we like her. So buy her book and follow her and give it a listen. [04:20] why not finance? To get started, go to kraken.com backslash boys club, sign up in just a few minutes and see what crypto can be. Not investment advice. Crypto trading involves risk of loss. Cryptocurrency services are provided to US and US territory customers by Payward Ventures, Incorporated, PVI, DBA, Kraken. On today's show, we have Cash Hill. She is a New York Times business reporter covering technology and privacy. Welcome to the show, Cash. [04:50] I will say, you know the meme of like new fear unlocked? [04:53] That structure of a meme, I feel like in digging into your work, that is me now. I'm like, no fear. Sorry. But really, really excited to have you on. So Cash is the author of a book called Your Face Belongs to Us, a secretive startup's quest to end privacy as we know it, which details her investigation into Clearview AI, which is a facial recognition app.

[05:23] What is Clearview AI and how does their facial recognition software work? Yeah, I just want to say with Natasha's comment, I do feel like my job is to write real life Black Mirror episodes. [05:37] You're doing a great job, honestly. Success. [05:42] And Clearview AI is definitely one of those stories. [05:53] photos from the internet, from social media sites such as Facebook and Instagram, Venmo, LinkedIn, to build this facial recognition app that works with what they claim is 98.6% accuracy. You take a photo of someone, you upload it to the app, it searches through the 30 billion faces that they've collected to find any other photos of that person on the internet and will tell you [06:23] you know, their social media profiles, even photos they don't maybe even know are out on the internet. And I've had that experience when when the company has run a search of my own face. Well, first of all, the Venmo photo history is a hilarious little addition in there. I'm like, wow, what's my photo? Okay, so the book is a wild ride. It's part sort of profile on how on this company Clearview AI and and how it's come together and part sort of history on facial

[06:53] the story is sort of about it being this very, very secretive, [06:58] company. How did it come on your radar? And for the listener, there's some really great twists and how you figured out the stories and like journalism, Easter eggs, I'd love to hear sort of the teaser version of how it came together. [07:10] Yeah, it started for me, I got a tip from a public records researcher who I knew from kind of privacy security circles. And he had been asking police departments around the country, what facial recognition technology they were using, and how much they're paying for it. And we all knew at that point that the police use facial recognition technology. But at that point, everyone thought it was pretty bad that the algorithms didn't work that well. [07:40] may be state driver's license photos. But what the researcher had gotten back from the Atlanta Police Department surprised him. It was this 26-page PDF about Clearview AI. It was written by this high-profile lawyer named Paul Clement, used to be Solicitor General of the United States. And he's describing, you know, what Clearview can do, that hundreds of police departments are using it, that it can search the entire internet. And he had written the memo basically to reassure [08:10] officers, that they could use Clearview AI without breaking the law, which was pretty remarkable. And I, Freddie sent this to me, this PDF, and he's like, do you want to look into this company? And I totally did. I mean, I immediately called him. It was, I got the email. It was in, I was in Switzerland. It was midnight. I was going to bed and I just called him right away. And I was like, Freddie, I'm, I'm out of the country, but I'll be back on Monday. And I want to look into

[08:40] what the company claimed it had done and that nobody would know about it. That this random, you know, company that when I looked online, there was nothing about them, that they were the ones to do this instead of a Google or a Facebook or a more well-known technology company. [08:55] And when I started looking into them, there were just so many red flags from a fake... [09:01] address on their website for where their office was based. It was three blocks away from the New York Times. And when I went there, the office didn't exist to no one from the company would return to [09:11] any of my calls or emails. And then I ended up finding police officers who had used the app and they really liked it. They said it worked incredibly well, but every time I would [09:22] talk to them, they would offer to run my face so I could see what the results would look like. And then they would stop talking to me. And it turned out the company had actually put an alert on my face and that they were monitoring me. And they were reaching out to these officers and telling them, don't talk to her, or you're going to basically get into trouble with us and [09:45] I just thought, wow, this is a very strange company with a lot of power. [09:50] Wow, that is... [09:51] So disturbing in so, in so many ways, very disturbing. I'm curious as, as you're talking about this, there's. [09:59] So many... [10:00] situations that come up for me that [10:03] are alarming and feel like they need to be regulated around facial recognition. But when I think about the conversation that's happening around AI, it's all sort of about generative AI and deep fakes. And we're sort of in this moment right now of all these

[10:17] really weird deep fakes of Obama and I spice sort of mashed together. And that's like, what is dominating the conversation? And I'd love to hear from you. Um, [10:26] Why do you think that this conversation around facial recognition is not at the forefront of regulation when it comes to AI? Yeah, honestly, I worry about that because facial recognition technology is, I mean, just like generative AI, it is AI. It is about, in both cases, what technology has made possible because of all these advances in pattern recognition and machine learning. [10:56] distracted by generative AI. Meanwhile, facial recognition technology has not been addressed, and it's very powerful, and it's spreading. And I do think we have these really important questions to answer about how widely deployed we want this to be, how accessible it should be. In the time from when I first reported about Clearview AI in January 2020, and privacy regulators outside of the United States looked at Clearview AI and said, hey, what they've done [11:26] We have privacy laws that say you need to get people's consent to use this kind of personal information. And they basically kicked Clearview AI out of their countries. But in the U.S., we just haven't really addressed it. And it's still being used. They just re-upped their contract with the Department of Homeland Security. And there are other... [11:47] facial recognition companies that are now doing what Clearview AI did. They created these big databases of faces and Clearview is limited to the police, but there's a subscription service called PimEyes. There's another one called facecheck.id online that anyone can use. You can just take a photo of anyone and upload it to these sites and find photos of people, find out who they are. And I just think the possibility is,

[12:15] already of how we're seeing them used. I write in the book about a guy who uses Pym Eyes to find out the real identities of women he sees in kind of pornographic videos. And he described going and finding this one woman's high school photos on Flickr. It's just kind of this kink he has. And then he kind of got sick of doing that. And he went through his Facebook friends list, all of his women friends, not his men friends, and looked for their faces. And he was able [12:45] to find naked photos of them like on revenge sites unsubscribe from this right like major so i'm i'm just worried about this world in which yeah yeah i mean we can be [12:58] linked to photos that maybe we don't want to be linked to on the internet. And then just this idea that you can't kind of navigate the world with any sense of anonymity, that anyone can find out your identity at any moment. I just can see so many ways in which that's going to go wrong. The current state of Clearview AI and sort of regulation in the US is that it can be used by law enforcement. Is there other restrictions that are placed on it? Curious sort of where [13:24] where things are at right now. [13:25] Yeah, so Clearview, you know, I write in the book about... [13:29] what their original idea was. And originally they just wanted to sell it to whoever would pay for it, but they ended up hitting on law enforcement as a very successful kind of use case for this kind of facial recognition technology. And now this has been a really successful defense for them when they have been sued, when the privacy regulators have come after them in Europe and Australia and Canada that they say, hey, we're only using this for the good of society. You know,

[13:59] troubles because I think a lot of people say, hey, if we're going to have facial recognition technology, this is the best use case. So with Clearview, with Europe, they said they shouldn't have put people in it. Like the approach in Europe is you can't put people into databases like this without their consent. In the US, in the few places where we have kind of privacy laws, there's a few states, they say that we have the right to get out of these databases once we're in them. And so [14:29] laws that allow you to go to Clearview AI and say, [14:33] hey, I want to see what you have on me and I want to be able to delete myself. And so there's a few places where you can get out of Clearview's database that way. There's also this law in Illinois that says that you can't collect people's biometrics without consent. So something like Clearview does violate Illinois' law. Illinois' law has protected [14:52] people already. Madison Square Garden, this big events venue, uses facial recognition technology to address security threats. Like people have been violent in the stadium or kind of the Taylor Swift model. Like if there's known stalkers of a celebrity, they might be on the ban list for a particular event. [15:08] But in the last year, Madison Square Garden, the owner, James Dolan, said, this would also be great for keeping out my enemies, people I don't like, namely lawyers who have cost us a lot of money. And so they banned all the lawyers who worked at law firms that had sued MSG. It kind of came to public light because this mother was taking her daughter and her Girl Scouts troupe to see the Rockettes at Radio City Music Hall.

[15:38] insane. And so yeah, lawyers have not been very happy about this. But it's a, you know, it's a very effective way for a company if they want to dissuade lawsuits, you know, ban the lawyers or ban the investigative journalists or ban the people with political opinions you don't like or ban anyone who's ever written a bad Yelp review or Google review. But Madison Square Garden has a theater in Chicago. And they can't do that in Chicago. They can't keep people out with facial recognition technology because of that Illinois law that says you need to get their permission. [16:08] It's just an example of how laws can work [16:11] can constrain this technology. We just have to choose to create those rules. Yeah, it's interesting when you talked about being able to go on and delete yourself from these databases. It sort of reminds me of cookies or GDPR compliance. You have the ability to take an action and remove yourself from specific databases or set your settings to a specific way in which your cookies are being tracked or whatever it may be. And unless you're a privacy [16:41] I don't think would really go to the effort of deleting their accounts. But I [16:46] it's only going to be until there's a problem for you that you realize, oh, shit, I want to go to all these places and remove myself. But then at that point, it's a little bit too late. So it's interesting because that feels like a nice loophole for them. But for the consumer. [16:59] who's actually going to take that action. Totally. And yeah, the companies love these kinds of laws, right? The opt out laws, because there's so few people who actually opt out. And with the California law, it requires that a company say how many people have asked to be deleted or asked to get access to their information. So I looked at Clearview's website and there's something like 34 million people who live in California. And in the last two years,

[17:29] from Clearview AI. So it really does show that those particular laws are not that effective. What works better is saying that you need to get people's consent to be put into the database. I'm immediately regretting all of the dumb ass things I have ever put on Facebook. So many images of myself, just images on images that no one needed to see and now are populating all these databases all over the world. It's just fodder for the AI databases. 500 wedding photos of me [17:59] Okay, so... [18:00] Clearly, this could be used towards very sinister ends, as you have laid out. What are some scenarios where there could be a positive use case for the technology? If everything was going right, what could be the case for this being used towards a good end? Well, it depends on who's defining what's good, but I think most people feel like the use of this for solving crimes is... [18:23] could be a benefit. I don't know if you've ever been a victim of a crime, but being in a scenario where it was caught on camera and there's an image of the person and that you can't track them down can be infuriating. And so to have a tool like Clearview, to be able to possibly identify that person, I think it's attractive to a lot of people. That said, it can go wrong. And I've written about at least one case where a doppelganger was arrested for a crime, a guy named

[18:53] arrested for basically stealing purses in New Orleans. He had never been to Louisiana before, but he wound up being in jail for a week because of bad investigation where they put too much credibility in the Clearview AI match. So that can go wrong, but possibly good use case. You know, facial recognition technology that we use to access services can be really beneficial to us, like opening your phone, you know, crossing a border where you can just easily match your [19:23] chip in your passport. I think a lot of people like those kinds of uses. The more challenging thing is companies having it to keep out shoplifters, but what if they get that surveillance creep, that function creep where they say, well, it's also good for these other things like Madison Square Garden with the lawyers or all of us having facial recognition. And I hear from people all the time who use it in online dating. When you have a picture of somebody, it is now possible to figure [19:53] if this is a catfish person who's repurposing someone else's image. Yeah, I mean... [19:58] I have written about people who do horrible things to each other online. I wrote about the serial defamer once who she held grudges against people going back to the 90s, people that she worked with at a real estate firm. And she would go online and just write horrible things about them, that they're fetophiles and thieves. And she would write about them and their families and their colleagues. It was kind of this. [20:19] The article is called A Vast Web of Vengeance. It was just like this spread. And so I think about someone like that having facial recognition technology and you bump into her on the subway or you say something that sets her off and she can know who you are from this brief real world encounter and follow you for the rest of your life. That is kind of where I worry it could go. But yeah, I mean, it really depends on how, again, how accessible we make it.

[20:49] Yeah, I think about... [20:51] On a lighter note, I think about, I don't know if you followed these TikTok girlies who are Gen Zers, who are masters at just very basic tools, not even facial recognition tools, but just LinkedIn and mostly social media accounts. I saw an incredible one a few weeks ago. It was... [21:11] this [21:11] Gen Z girl and she was like taking a picture of the sunset and this guy walks into her video and she's like, who's this daddy? And it's like this hot older man. And then the rest of the TikTok is her going through the internet and at the end of the day, finding him just through like reverse image search and Facebook and Instagram and then messages him on LinkedIn, like asking him out. And she's like, I don't know. [21:35] watching that, I was like, this is hilarious and so entertaining. This woman is like, so compelled by how handsome this man is to go find him on the internet. But, [21:43] That gets really sinister and really dark very quickly. And with these tools, that gets easier and easier and easier. So it's definitely terrifying. [21:52] I want to switch gears to your Twitter bio here and hear a little bit about your personal take on these technologies. You call yourself a privacy pragmatist. What does that mean for you? And how has that thesis either developed or evolved as you've been working on writing this book? Yeah. So when I first started writing about privacy in 2009, I started a blog called The Not-So-Private Parts. And it was about kind of the intersection of technology and privacy,

[22:22] about us, you know, what it means to be a public figure, how easy it was to find information about people. And I was kind of more of a privacy skeptic when I first started writing about it. I thought, you know, there are so many benefits that we get from free flow of information and being out there online that it was more good than bad. And so when I wrote about privacy, I would really try to find privacy. [22:47] victims who are really harmed by the use of the technology. I would say over time, I'm less skeptical than I was that I have seen so many, you know, I see more possible harms that come from these technologies now, but I try to balance it. I try to see the good and see the bad and try to find some kind of golden middle. And yeah, I mean, with any technology, kind of what you're saying, facial recognition technology will be used in all kinds of ways. And some of them may be frivolous [23:17] them may be sinister and some of them may be really beneficial and good, like helping us solve crimes. I think that is generally a very good thing. It's not so much about trusting a technology as trusting the people using it. And so ideally we can kind of create guardrails to avoid the worst case scenarios. [23:36] We come from the corner of the internet that's crypto and the privacy conversation in crypto dominates. And I would say within that conversation, Natasha and myself are like super moderate. It's just funny because it means a lot of different things to different people. When I think of the sort of privacy advocates in my circle, they're deeply technical. I always feel like they see something that I don't understand that they're like, oh, I'm anticipating some future that I just personally don't get yet or I can't see yet.

[24:06] scenario where it's like, okay, we all know it's important. And we all kind of think about it. And in tech circles, like maybe there's more advocates than it than in sort of mainstream circles. But I can't really get my neighbor to care about privacy, or some mom that I meet on school pickup. Why do you think that that advocacy is lacking for this as like a human right? I think that everybody cares about privacy. It's just you have to find the right way in. [24:36] your information get misused. And there's so many different ways. I used to think about this in the context of targeted advertising. Like most people don't really care about targeted advertising. Sometimes they find it creepy. But where I think it gets very uncomfortable is when a data broker is trying to mine your pregnancy status. And they figure out kind of that you are pregnant because of what you're searching for or what you've bought. And they start targeting you [25:06] you know, so common. And then you just keep getting these ads that basically assuming that you've come to term. I mean, I wrote about a woman once who got baby formula in the mail, you know, a week after a baby should have been born, but her baby hadn't been born. I mean, just kind of these harms that come. So that's what I try to focus on. What creeps you out? Like, do you mind that everyone has a ring camera in their doorbell and they kind of see you walking by? Is that okay with you? One of the things I've written about, [25:34] is a lot of the tech giants now scan all the photos that are uploaded looking for abusive images of children. There's something called known images. It's CSAM or child sexual abuse material, and they look for known images and they will flag that and report that. Google a few years ago developed an algorithm that looks for unknown images. So it would hopefully find victims that

[26:04] kids that are being sexually abused if those images are being shared online. But I wrote about these dads who during the pandemic took photos of their sons for the doctor because they're having medical problems in their personal area. And they got flagged by Google, had their account shut down, lost all of their Google data, their emails, their photos, I mean, everything, and were reported to the police and got investigated by the police, who in both cases, just said, Oh, this is nothing, [26:34] But they couldn't get their Google accounts back, which was pretty devastating, especially for one, like his whole life was on Google. So just looking at, okay, you can set up this thing, which sounds great, but these things can go wrong, right? You can misidentify and there's going to be some significant number of people when we're talking about, you know, scanning millions, hundreds of millions, billions of accounts where it could go really fast. [26:57] poorly for people. And so trying to point those out to create more grace in the process. [27:03] Yeah, there was like a, this was, this happened over, I think it was in Brooklyn, in my neighborhood in Williamsburg, there was like a wheat pasting campaign that was images of children and it was like, Meta knows, or I think it was Meta or Facebook or something, knows that these children are being abused and doing nothing about it. And it was something like really inflammatory and... [27:23] then it was basically like sign this petition to force them into action. And I remember seeing it and being like, [27:30] this is crazy. This is a crazy, first of all, wheat paste campaign, like what in the world? And then second, this is surveillance. Like that's what that that's what that campaign is for. And I feel like I would only have recognized that because I've been working in crypto the last two years and people are obsessed with their privacy and being anonymous and all this sort of stuff. But it's interesting. That's sort of the other side of the aisle of people who are seeing this as a real

[28:00] because this technology is in many ways still very nascent. One of the things that I've heard you talk about is how big tech had this technology before Clearview or these other applications and decided to not... [28:13] release their tool. [28:14] Could you talk a little bit more about that and why they came to that decision? [28:18] Yeah. So when I first heard about Clearview AI and kind of the extraordinary project they had done, I thought, wow, there must be some real... [28:27] technological masterminds involved in this company that they're able to do what Google and Facebook couldn't. But I mean, the story of the company is fascinating. It's almost felt like writing fiction sometimes, the background of them. And I would not say that they were necessarily technological masterminds and that their breakthrough was more of an ethical one. I discovered through my reporting for the book that both Google and Facebook had kind of developed a Clearview AI-like technology internally and that both companies [28:57] they didn't want to be the ones to put it out into the world, that it was just too dangerous. It could be used in positive ways, but that just clearly had downsides. And I don't think they wanted to face the reputational, legal and ethical risk of it. And so... [29:14] Clearview was just [29:15] willing to break that taboo to do what these other companies weren't willing to do. That was kind of their advantage as this startup that didn't really have anything to lose. Unexpected hero. [29:28] That's a twist. I know they're not known as being particularly conservative when it comes to use of data. So I thought that was really quite striking that for them, facial recognition technology was one of the first places to really draw a line.

[29:43] Wow. Okay, one last thing. I know you reported today some news from PimEyes. PimEyes is the desktop version of Clearview AI. You can access it from your browser. You could go there right now if you wanted to. I saw in your report that PimEyes has a database of nearly 3 billion faces and enables about 118,000 searches. [30:03] Per day. [30:06] According to the CEO. And you can put, I tried it. I put my photo in and it pulled up all these places that my photo is being used. And it was really disturbing. I think if anyone's listening and wants to understand sort of the breadth of this technology, I would recommend going there and trying that out because it really brought it to life for me where I was like, man, [30:24] Oh, okay. Here I am on all these places that I had really no idea about. But anyway, you reported today that there was some news that they finally banned searches for children's faces. Is that right? Can you say more about that? [30:36] Yes. So I reported last week about kind of AI based threats to children from searching for their face using these public face search engines to the use of their deep fake versions of their voices to try to convince their parents that they've been kidnapped, which is like just crazy. [31:06] files by child molesters. And they said they were working on a fix. And after the story came out, you know, some people in the story were parents who were looking for photos of their children to see, hey, is my kid on the internet? And I don't know about it. And so I started hearing from parents, hey, I just tried to run my child and the search is blocked. And so Pimmies has now blocked basically minors, kind of anyone who looks like they're under 13 is blocking them.

[31:36] days where they're probably four or five years old in the photo. And it did block one of their faces, but the other one, she was kind of turned sideways in profile and it did run her face. So they obviously still have some kinks to work out in the system to make sure it works. Yes. [31:52] Really, really interesting stuff. And so grateful for you to come on and share with us about this. I just want to end on a happy note. What is one thing on the internet that is just like bringing you a lot of joy right now? [32:06] it can be anything it can be a meme it can be a tweet what is bringing me joy i don't know you're like i'm not on that side of the internet i have one i have one that relates to her yeah i'm just like looking at my tabs and it's like all bad stuff oh no [32:26] One thing that I have is that Cash's book is called Your Face Belongs to Us. And people have been using the meme. You explain it. [32:33] been saying that you should change the title. I mean, maybe this isn't happy for you, but. No, this has been fun. And so, yeah. So the name of the book is Your Face Belongs to Us, but everyone, and I know this meme because I'm like an early blogger, but all your base are belong to us. So it's just this bad translation in a Japanese video game of what they're saying. And so I did a bunch of book signings, like a dedicated book signing where I like cross, I like add all and cross out the S and add R and make it all your face are belong to us. [33:03] And yes, I have enjoyed that. Most of my happiness comes from being off the internet right now, like looking at the fall trees and hiking. I respect that. On that note, Cash, thank you so much for coming on. We really appreciate it deeply. And yeah, just thank you so much. Thank you. What a fun podcast. When you're first getting started with crypto, it can be scary. Am I doing this right? Is this just like my bank or trading app? How is it new

[33:33] we love Kraken. They have a 24-7, 365 customer support team that's there to hold your hand all along the way. This isn't a nine-to-five Monday to Friday bank. This is crypto. It's all the time. Anyone's welcome. Open door policy. Come one, come all. Try something new at kraken.com backslash boys club. Not investment advice. Crypto trading involves risk of loss. Cryptocurrency services are provided to U.S. and U.S. territory customers by Payward Ventures Incorporated, PBI, DBA, Kraken. [34:03] Draft tweets? Yeah. I have one here. Okay, go ahead. Okay, so a couple days ago, I posted this tweet. Weird how we all post whatever idea comes into our head at any hour of the day on here. And the reason I... [34:16] Thought that was because of the second tweet that I almost replied to and started a thread, but it felt too unhinged. So I didn't. But this is a perfect medium for it. So for instance, something mildly good happened to someone I kind of don't like today. And I got annoyed. [34:33] and feel compelled to share that thought with the whole internet, that's mental illness. [34:40] that is mental illness like [34:44] Why would I share that? [34:47] That I was like dealing with that in my own head. That's so funny. That is so funny. That is scary. [34:56] It's scary. Okay. Thanks for listening. Bye. [35:03] There it is. That's Boys Club. We have two podcasts, two online, where we talk about internet culture, and this podcast where we report on some Web3 or emerging tech story and then talk about our feelings, the feelings check-in. So thank you for listening. Please like, subscribe, share, send to your friends. Anything else? That's it. Thank you so much.