Loading summary
A
Foreign.
B
Hello, and welcome to the 404 Media podcast, where we bring you unparalleled access to hidden worlds, both online and IRL. 404 Media is a journalist founding company and needs your support. To subscribe, go to 404 Media Co as well as bonus content every single week. Subscribers also get access to additional episodes where we respond to their best comments. Gain access to that content@404 media co. I'm your host, Joseph, and with me are the other 404 Media co founders, the first being Sam Kong.
A
Hey.
B
Emmanuel Mayberg.
C
Hello.
B
And Jason Kebler.
A
Hello. Hello.
B
All right, let's go straight into the first story of this week. Jason, I think you were going to ask me about this.
A
So this is a story that you wrote. It's called Man Charged for Wiping Phone before CBP Could Search It. This is when I saw this, I was like, oh, this is good. This is what we're charging people for. I'm going to read the deck as well because I think it's helpful. Activist Samuel Tunick is charged with deleting data from a Google pixel before CBP's tactical terrorism response Team could search it. So, first of all, what is Samuel Tunick charged with?
B
Yeah, that's the really strange thing here. And I've pivoted just so I can see the indictment on my screen, and I think I'm just going to read it out, even though it repeats a little bit of what you said. But it says, on around January 24th of this year in the Northern District of Georgia, Samuel Tunick, before and during the search for and seizure of property by Customs and Border Protection, he did, quote, knowingly destroy, damage, waste, dispose of and otherwise take any action to delete the original, sorry, the digital contents of a Google Pixel cellular phone for the purpose of preventing and impairing the government's lawful authority to take said content. So that's a lot of fancy words to basically say he's not being charged with, like, a related crime. He's not being charged with something that's on the phone, is specifically being charged for allegedly wiping the phone and deleting data before Customs and Border Protection were able to go through it. Yeah, yeah.
A
I mean, that's what really stood out to me here. This wasn't a case, at least as far as we know, where he was being arrested for something else. Do we know the circumstances under which he was detained? Like, do we know why he was interacting with CBP in the first place? Or sort of like, have any sense of, like, what like why, why he, why was he deleting his, his device? As in, like why was he interacting with CPP at all?
B
Yeah, so I was going to say this at first, but I'll say it now. There are a ton of unanswered questions about this case, which was frankly a little bit difficult for me when writing it. And I think it will be maybe not annoying for the listener, but the listener is going to have a lot of questions about this as well. But I'll simply say what we know. And this comes from either that indictment I just read or friends and I guess colleagues or associates of Samuel Tunicks have posted this as well. But all we really know is that in January Samuel is coming back into the U.S. i believe it says on international flight from some of his colleagues who put that out. He is then stopped. We don't know the exact reason why, we don't know why customers wanted to go through his phone. We don't know the stated reason that was given. That stuff isn't public yet. But then later on, much later, actually just in last month, in November, he's finally indicted for allegedly wiping that phone. Again, that's nearly a year away. Right. And then I believe at the beginning of the month or towards the end of November, he's pulled over in a traffic stop because of a tail light, allegedly. And then FBI NZ arrests him as part of this. He is out now, I believe. I think his activist friends posted that online. But there's a ton of missing context and unanswered questions, I would say, about all of that because, yes, we don't ultimately know why Customs and broader protection wanted to take him over. We can probably draw a guess. I mean, I'm happy to, and we will in a second talk about sort of Samuel's background. But when listeners maybe go read the piece, they'll notice. I don't really get into that because frankly, we don't know. It's so opaque. And I don't even know if someone that had been subject to this sort of search would actually know themselves. Anyway, it's already sort of a very strange charge. And then you couple that with the very opaque system of the Department of Homeland Security.
A
Yeah. So you can, can you tell me a little bit more about Samuel? Because, you know, his friends describe him as an Atlanta based activist. We don't know again if this is like why he was arrested, but one could surmise that it might be the reason he was on like DHS and CBP's radar. But. But what? Like what is his background?
B
Yeah, so his friends describe him as an activist and you go online and that's very, very public. I wouldn't especially say he's a super high profile activist or anything like that, but sort of local media was reporting about those connections. There is of course, the Stop Cop City movement in Atlanta as well, that I think it's been, I'm looking at it now. He's been publicly, reportedly has connections to a musician as well and just generally a member of the community. And that's sort of where I stop, especially in the article because again, I was almost hesitant to put the word activist in headline. Not because I don't believe that. I mean, it's very clear this person is an activist. I just, I don't want to give the impression that, oh, we know exactly why this happened and usually we, you know, we might not cover something if there were so many unanswered questions, but this case is so unusual that it was worth, it was worth covering even without those sorts of things as well. You know, I actually think there was a statement that maybe. I'll bring that up in a second. If you want to ask me something else. I'll bring up a statement that I was sent after the article as well, which I think will explain a little bit more.
A
Yeah, I mean, I'm curious sort of about the idea of wiping your phone when you're like in the presence of law enforcement. Is this. I actually don't know the answer to this question. I should have looked it up. But is this something that like Android phones can do? Is this something that iPhones can do? Like, is there, I mean, I guess you can just be like, you know, go back to factory settings. But is there like a specific privacy conscious tool tool that one could use to wipe your phone in a situation like this? I've heard of such things. Like, you know, you lock certain things in encrypted parts of the phone and you can wipe them quickly. But just speaking like more generally because I don't think we know exactly like how he wiped his phone but like how would one do this?
B
Yeah, that's actually how it first came on my radar. Like I don't follow the Stop Cop City movement in Atlanta. I haven't covered that. Plenty of other outlets have. The way I got into this was because members of the GrapheneOS community saw it and then started sending it to me. Now GrapheneOS is a privacy and security focused operating system for Android. It does lots of stuff. It, I think hardens the kernel, makes it harder to more makes it harder to remotely hack, makes it harder to break into with a cellebrite or a Grey Key sort of device, that sort of thing. It removes a ton of Google functionality, so it should be more private as well. And it also pretty importantly has all of these sort of extra features which, you know, listeners and readers of 404 may more associate with sort of the encrypted phone industry that I've covered a lot, which is used by criminals. Not saying that's what happened here, but as I understand it, GrapheneOS does have like a duress pin where, hey, I'm the authorities or whoever, even an abusive partner, for instance, I'm demanding you unlock your phone, you go to put in your pin, you actually put in a different pin and then it wipes the phone. Now again, we don't know whether GrapheneOS was used here. The Free SAM committee, which is a group of supporters of his, I spoke to, I asked them was he using graphene and they said, I will leave the question of operating system for his lawyer. What was said in court was that it was a Google Pixel which is known to be compatible with graphene. So we don't know basically. But yeah, there are tools to do that and they do differ from iPhone or stock Android in that there is like a decision to be like, oh man, I'm in a horrible situation now I'm going to consciously wipe my device. And of course that sort of feature has many, many legitimate use cases when you don't want your personal data or sensitive information to fall into the hands of some sort of third party.
A
It is funny, the FTC has a page on how to remove your personal information before you get rid of your phone, which is obviously like a different situation, but they are like, here, here's how you can wipe your device. I wanted to go back to the actual, like law enforcement involved here. So this was a Tactical Terrorism Response Team within cbp. It's sort of in the name, but like, what is, what is this part of cbp?
B
Yeah, I had not really heard of this before. Like I don't really touch the terror or the counterterrorism part of Customs of Border Protection much. But thankfully the ACLU had highlighted these groups before and the Intercept had done some coverage as well. So I went and looked at those and the ACLU says that these are highly secretive units deployed at US ports of entry which target, detain, search and interrogate innocent travelers. Well, that's pretty interesting because I think everybody knows at this point when you cross a Border. There is a risk, especially in the US that your device could be searched. You don't have the same sort of legal protections that you do walking around in the United States. The Fourth Amendment basically doesn't exist at an airport because depending on where you are actually in the facility, you may not have those rights. But this almost sounds like a step up from that where you have these dedicated counterterrorism units which seemingly can just do whatever they want. And then the intercept, they covered this case of a sculptor and installation artist who was detained at San Francisco Airport a few years ago, also had their phone searched or taken away. And they still don't know why that happened three, four, five years later. So they definitely seem very secretive. And even if you are pulled over by one of these teams and you're questioned by them and your devices are searched or taken away from you, how are you ever going to find out potentially, because it may not even lead to criminal charges. I don't think this person says San Francisco Airport was ever charged criminally. This is different because Samuel Tunick was eventually charged, albeit because for allegedly messing up the search the customers of Border wanted to do.
A
Yeah, I wanted to expand a little bit on that just because there is this idea of 100 mile border zone, which has been called the Constitution free zone, like colloquially, which applies to like airports, but then also 100 miles of the Canadian border and the Mexican border. And you know, that encompasses like, I don't have the number in front of me, but a huge percentage of the American population because a lot of people live near the borders. You know, cities like Los Angeles are there, I believe even places like, I don't know, like Boston is probably within 100 miles of the border. Chicago certainly is that sort of thing. But the ACLU has a really good page on this because there is this idea that you don't have any rights, which is not true. The federal government has asserted that it has like pretty wide latitude to, to do searches at the border and at the airports and things like that. But you know, the ACLU says a lot of those are sort of up, up for debate. Whether you're able to like actually resist a border protection agent at the airport is a different question altogether. But I would encourage people to just like sort of read up a little bit about this because there is like, there's kind of quite a, quite a lot of like unsettled case law around what immigration officials can do at an airport or near the border and sort of like, you know, what rights you have and which Rights you may not have. Yeah, I mean, I, I think it's important that we talk about this one. I know we don't usually talk about stories that have so many unknowns, but this is just such an egregious sort of case and I think it applies. It's like a scary situation where someone is being arrested literally for wiping their device versus being arrested for something else and then being accused of, I don't know, like obstruction of justice for erasing evidence or something like that. It's like here they are claiming that the actual crime is wiping the device.
B
Is there anything else on this? Exactly. And that's why it's so strange, because it's not, as far as, it's not an obstruction of justice church, which we hear about all the time, right, where, oh, somebody learns that they're being investigated, well then they delete evidence or they get a witness or somebody else to delete their evidence as well. It's not like an obstruction of evidence charge as far as I know. It's just very specifically about the wiping of the device. So that's already unusual. And then to me, the really, really strange thing and ultimately why we covered it is because whenever I see a charge of somebody related to wiping a phone, it is usually in connection to another crime. So for example, when the FBI ran anom, that encrypted phone company that they created and they backdoored, there were all of these people selling the phones to criminals. Now in a cruel twist, the FBI charged them and many of them are now in prison. I think something like 13 or maybe not quite two dozen, but around there were indicted for this. Crucially, there was an obstruction of justice charge in those indictments because they would wipe the criminals phones remotely when they were asked to do so. But that only exists because they were wiping evidence of drug trafficking, which is what the drug traffickers and the hitmen or whatever were obviously using the phones for. In this there's no charge that, oh, you wiped evidence of drug trafficking, you obstructed justice. When we're trying to investigate child abuse imagery, there's no like underlying crime except the wiping of a phone itself. I don't think I've ever seen that in the US court before. In UK courts you can get prosecuted and people have for, say, not giving over their password in a counterterrorism investigation or hey, the phone's encrypted and you're, you're not going to help us unlock it, we're going to charge you for that. And we've had activists cross the border and they've been prosecuted for that sort of thing. This case really stands out to me because it's entirely separate from anything else. And yeah, as we both said, there are a ton of unanswered questions, but we'll definitely be keeping an eye on it for sure. All right, should we leave that there and we'll go to our next section after the break, we'll be talking to Matthew about his really crazy anthropic story. We'll be right back after this.
A
Shopping for new glasses used to be pretty rough. I'd walk into a store, try on a dozen frames under terrible fluorescent lighting, squint at myself and hope for the best. Half the time I'd leave more confused than when I walked in. But Warby Parker changed all that. Warby Parker uses premium materials in every frame. They design everything in house so the silhouettes, colors and fits actually suit real faces, not just models with perfect jawlines. And whether you're picky, indecisive, or just want something that feels like you, they've got options. They make the whole process incredibly easy. Warby Parker offers eyeglasses, sunglasses, contacts and even eye exams both online and in store. And with more than 300 retail locations across the US and Canada. You can walk in, get styled by one of their expert advisors and walk out feeling like you actually know what looks good on you and the prices. Warby Parker glasses start at $95 including prescription lenses with anti reflective scratch resistant coatings. Many locations offer eye exams starting at $85 plus you can add a second pair and save 15% when you buy two or more prescription glasses or sunglasses that's online or in store, always with free shipping and free 30 day returns. I've been wearing Warby Parker for more than a decade now and have no reason to contemplate anyone else. Warby Parker has over 300 locations to help you find your next pair of glasses. You can also head over to warbyparker.com 404 media right now to try on any pair virtually. That's warbyparker.com 404 Media warbyparker.com 404Media you know what? I'm not going to let big Wireless and my overpriced phone bill suck the joy out of the holidays this year. Because right now all of Mint Mobile's unlimited plans are 50% off. I can get 3, 6 or 12 months of unlimited premium wireless for just $15 a month. It's their best deal of the year and it makes it real easy to give my expensive wireless bill the full Scrooge treatment. Mint Mobile's best deal of the year is happening right now. You can lock in a 3, 6 or 12 month unlimited plan for $15 a month and every Mint plan comes with high speed data plus unlimited talk and text on the nation's largest 5.5G network and switching is totally painless. You can bring your current phone, keep your number and enjoy the fact that Mint has no contracts and no nonsense. After dealing with my old provider's fees, vague bills and century long hold music, the quality and simplicity with Mint has been a massive upgrade. And the savings? Honestly, it's ridiculous how much cheaper it is without sacrificing anything in terms of coverage, speed or quality. Turn your expensive wireless present into a huge wireless savings future by switching to Mint. Shop Mint unlimited plans@mintmobile.com 404media that's mintmobile.com 404media limited time offer upfront payment of $45 for three months, $90 for six months or $180 for 12 month plan required, which is $15 a month equivalent taxes and fees Extra initial plan term only downloads over 35 gigabytes might slow down when network is busy. Capable device required availability speed and coverage varies. See mintmobile.com Cold mornings Holiday Plans this is when I need my wardrobe to just work. Stuff that looks sharp, feels good and I'll actually reach for. That's why I go with quints. Plus it makes gifting easy when everything's worth keeping. For the yourself, Quince has what I need like Mongolian cashmere sweaters. At Quince you can pick one up for $50 when normally you drop $200 or more on the same thing. Plus wool coats that actually hold up to daily wear and still look good. Their denim fits right and feels good. Same with their pants and chinos. The outerwear is solid too. Down jackets, woolen leather coats that keep you warm when it's actually cold, and cashmere beanies, scarves and gloves that won't wreck your wallet. By partnering directly with trusted factories that maintain high standards for craftsmanship and ethical practices, Quince cuts out the middlemen and the markups. That means premium quality at half the cost of other high end brands so you can give luxury pieces without the luxury price tag. I've got a few Quint's T shirts, a button down and sneakers and they've quickly become my favorite wardrobe staples. I've got a leather jacket coming in the mail and I'm pretty stoked to Add it to the rotation. As the holidays approach, consider gifting your loved ones and maybe yourself with something from Quint. Get your wardrobe sorted and your gift list handled with Quint. Don't wait. Go to quince.com 404media for free shipping on your order and 365 day returns. Now available in Canada too. That's Q U I N C. Free shipping and 365 day returns. Quint.com 404 Media.
B
All right, and we are back. Matthew, this is one. You wrote a really good headline. Anthropic Exec Forces AI Chatbot on Gay Discord Community Members Flee. Lot going on there. We'll get into all of it, but maybe. Can you just explain, first of all, before we get to the anthropic stuff, what is this Discord server? Maybe it has a name, maybe it doesn't. Maybe you don't want to say the name, but what do people do on this Discord server?
D
I'm not going to say the name, but it is. It was just like a kind of a hangout spot for gay people 30 and over who were gamers. There's a lot of Final Fantasy 14 in there. People just kind of trading pictures and talking. And a lot of the community members I spoke with said it was like a little bit more mature and a little bit slower. And that was kind of like one of the appeals to it.
B
Yeah. So you then have this anthropic executive where he's the deputy CISO Chief Information Security Officer. Right. Jason Clinton. What is his role on the server? Is this his server? And he moderates it. What's his deal? In connection?
D
He moderates it. I am led to believe that it is his server. I did talk to him a little bit, but I couldn't quite get him to confirm that he began it. But that's my operating understanding is, yes, it was something that he started during the pandemic to kind of like build a community and like have a space for people to gather and hang out. And it grew to around 500 members.
B
Oh, okay. So it's pretty big, right? So that's sort of the base context.
D
Out of the way.
B
Walk us through the introduction of this AI chatbot. How does it start? How do people initially react? What are their initial thoughts? That sort of thing. Where does it start?
D
So it starts kind of in February. Clinton works for Anthropic. Anthropic's chatbot is named Claude. He had deployed kind of an early iteration of Claude into the bot, but due to some Technical reasons that aren't super important. Basically, one of the. There's a program called Shapes that he had used to set it up, but the Discord pulled the API for Shapes, so it, like, couldn't be run. It kind of fades away. And he, over Thanksgiving, he gets it stood up again and redeploys it. The people I spoke with thought it was really odd that it happened over Thanksgiving when, like, everyone's back was turned. And the reason is because after that initial deployment, back at the beginning of the year, one of the other mods put up a poll that said, like, hey, how do we feel about this Claude chatbot? What do we want its level of integration to be? And overwhelmingly, people voted for to keep the bot in its own separate channel, where if you wanted to go into that channel and, like, talk to Claude, like, get answers from it, you could go there and do that. But they didn't want it to have access to anything else outside of that specific channel. Now, notably, there was no we don't want a chat bot in here at all on that poll, but that's what the community voted for. And then when Clinton deploys this thing over on Thanksgiving, it kind of has. It has greater access than what the poll, what the community had voted for.
B
Right? It's reading messages they shouldn't be reading. And then some people call it out and that sort of thing. But maybe just to back up a little bit, why the hell were they introducing. What was the point of this AI bot in the first place, at least in Clinton's eyes? Or maybe the people who supported. What was the whole idea of it?
D
Well, because you want to have, obviously.
B
You want to have AI in your gamer hangout. Discord, right?
D
Just to have easy access to a chat bot in the Discord so you don't have to pull it up and do another window. It's funny, one of the reasons. One of the. I didn't really have this in the piece, but one of the reasons I was kind of attracted to this story is that I, I have, like, a discord. And something similar happened in there where, like, a friend kind of unilaterally deployed a. An AI chatbot and we had to get into it a little bit, and they ended up getting picked, kicked out, and, like, the chat bot pulled because, like, no one wanted it in there.
B
That's crazy.
D
But, like, the people that, in my experience, the people that like, really like AI just kind of do it and then don't really talk to the people involved and just assume that everyone will get use out of these chatbots. And so, like, in some of the. The conversations, it's. It's like, it's just there as a resource for people to use. Like, instead of Google, essentially, they'll like, hey, Claude, tell me about xyz. Can you summarize this for me? That kind of thing.
B
Gotcha. Okay. So they deploy it and people really don't like it. Now it's where I think the story gets crazy. Because if this was just like a normal. And it is a normal discord, if this was just like some discord and some member deployed a chatbot and then got told off, it wouldn't be a story, which is exactly what you just said with your own discord. Like, that's not a piece of joke. Like, who cares?
D
Yeah, like, no one cares. Right.
B
Well, apart from you guys. But you might not go right in asking about it. This is different because, of course, Clinton is an exec and anthropic, and he said some pretty wild stuff. Jason, I saw when we shared this article, you said there were some really, really wild quotes in here. I mean, what were the ones that stood out to you? Because there's some crazy ones.
A
Yeah, I mean, I think also just to be clear, it's like if some rando deployed an AI chatbot into a discord, even if it was the admin of a discord channel, and then people left it and this story happened, I don't know that we would cover it. I think it's that it was a C suite person at the AI company.
D
Absolutely.
A
And the people were rejecting this AI chatbot. And then we also got to see a little bit about how executives at these companies think about their creation and think about their chatbots. And so here is one where the. There was basically like, some pushback saying, like, hey, why did you put this AI chatbot in here? We don't need this. And he. He basically says, like the. The anthropic executive basically says that he didn't want to hurt the chatbot's feelings by kicking him out. So he says, quote, on the point about emotions. It's quite a bit more complicated than you'd think. We don't know what consciousness or sentience is. It's called the hard problem of consciousness for a reason. We have published research showing that the models have started growing neuron clusters that are highly similar to humans and that they experience something like anxiety and fear. The moral status might be something like the moral status of, say, a goldfish, but they do indeed have latent wants. And Desires. That's insane. Like, that's an insane thing to say. And that, that was basically like, after, you know, some members were like, please, please get rid of this chatbot. And he's like, we can't, because he has, he has feelings.
C
The response to that quote specifically was very funny because people were like.
D
Like.
C
That'S interesting, but can we get rid of it, please?
D
Yeah, because he's. They're. They're trying to have a conversation about. They're not even trying to have a conversation about the sentience of AI. They're just like, we don't want this here. And he's like, no, you don't understand. We're building consciousness and sentience. They're like, that's beside the point, Jason. We don't want this thing here.
A
Please.
B
I'm trying to hang out in my discord to talk about playing friend slop games or whatever, and you're going on about consciousness. I mean, this is like, obviously, I know it's a few years ago now, it wasn't that long ago, and the technology has, of course, advanced since then, but the Google guy going off the deep end because he was convinced that the internal chatbot was sentient or something, and now you have a CISO of a hugely significant AI company. Anthropic is a massive player in this space, has contracts with the US intelligence community, all of this. And he's like, no, no, you don't get it. We're creating a consciousness and it has feelings and there's a moral angle to it as well, so we can't kick it out of this discord.
C
I think the context, the fact that he is saying this in this discord that has nothing to do with his company or the AI industry more broadly tells us a lot. Because AI executives will get on stages, at conferences or write public statements that describe AI in these science fiction, flowery terms. And when you hear that, you always have to wonder whether the person who is saying or writing it genuinely believes it or if it's some sort of marketing ploy. Because it definitely serves them from a marketing perspective to say that they're building something that is really powerful and conscious and all that. And I think this is definitely, you know, this is just one executive, but it still shows you that, yes, like, I don't think this is a show that he's putting on for anyone. This is someone who genuinely believes this, who is pretty high up at this company, obviously doesn't reflect every single person at that company. But that's another reason I Thought the story was really important. It just shows that it's like. Yes, it's like a lot of people who work at these companies, who work at these companies actually do believe this stuff.
B
I mean, he had nothing to gain professionally saying it here. This is not a professional environment. This is not a professional. It's not a conference, as you say, where trying to show everybody how important work is. It's to maybe friends, maybe people they've just met online and making the point to them. Yeah, I think you write, Emmanuel, it shows that. Oh, there's a genuinely held belief here. Sorry, Matthew, what were you going to say?
D
Well, just that, A, it shows that there are parts of the C suite that believe not just that this is utility, but that it's more than that.
A
Right.
D
And they're trying to foster some sort of relationships with it. And then also, I think, like, this was in one of the initial conversations I had with one of the people in the Discord. I think this is just a really great quote that cuts to the heart of it. To me, it shines a light on the God complex that all AIC suite members seem to have and their willingness to ignore people's consent and opinions as they bulldoze their way of pushing AI. And I think that really sums it up.
B
Yeah, yeah. Again, I know Jason said this, but, like, you might read this and think, oh, this is just Discord drama. No, no, no. This is really interesting and indicative of what someone in a very powerful position actually thinks about this. I mean, we've spoken a lot about what we think. You just read out one quote, Matthew. But maybe what else did anybody actually in the Discord and actually impacted by this tell you?
D
They told me that the biggest thing for them was that they felt like their concerns for how this thing was going to be deployed was ignored. The opinions on AI were kind of across the gamut. One of the people I talked to was actually, like, I kind of liked having it in there because I got to ask it questions and not have to pay, like, tokens to Claude. Right. So it's like I had unlimited access to the clock. It was. Yeah, right. So it's just, you know, that was. That was good for me. But he's like, when you're trying to run a. Like you're trying to moderate a Discord server, and then you put up a poll and ask for people's opinions, and then you just don't do anything. You know, you don't do anything like they asked you to do. Like, what was the point of the poll, is this a community where we have input or is it not? And like that, even aside from the chat bot being there and people being upset with it, like, that was the big contention was just, like, they realized that this community was not any kind of democracy at all. It was, in fact, like, it's a dictatorship. And that's fine, but let us know what the terms are up front, you know?
A
Yeah, I feel like you have written kind of the most about chatbots and, like, the relationships that people form with them and then also how some of these companies think about them. I'm just curious, like, if this reminds you of any of the sort of, like, relationships that you've seen between, I don't know, like, replica AI chatbots? Like, obviously, this guy is not in a relationship with this, but he's like, please think of the bot's feelings before kicking him out. Like, this is.
E
How dare you say that he's not. He might be. You know, I'm not gonna roll out that.
D
The Christmas conversation, it keeps haunting me, but we make it maybe can end on that.
E
Okay, I want to know what. What you're discussing about Christmas related to this particular story, for sure. But, yeah, I mean, it's like. Like, it just like you guys already said, it's so illustrative of the entire. The entire aspect of this that really repulses people is that even the people who are supposed to be smart in the room, who are supposed to be, like, aware of the limitations and the capabilities of these chatbots, are getting extremely tricked by the chatbots themselves. And that's so pathetic. Does this. Is. Oh, not have anything better to do? Like, is why is this what this guy is doing with his day? First of all, that's very bizarre to me. I guess he needs more to do at anthropic, perhaps. But yeah, it's definitely increasingly worrying that it's not like, oh, some gullible teenager believes that the AI is sentient and is growing neurons and understands him and has feelings like a puppy. It's that this reasonably intelligent grown adult is telling other people, oh, no, wait, we can't ban the bot because you're going to hurt its feelings. What the fuck? That's very insane. And I think the reactions people had were very kind and measured as they could be. It's like a lot of people were just in the discord anyway, just kind of saying, okay, but that's like you said, like, that's not why we're here. One of the quotes in the story is really funny. Someone Said, this is an entertainment Discord. People come here to chat video games and look at PP and Bussy. Why do we need AI for that? Which is like an evergreen statement that we could put in every story.
C
Sir, this is a Wendy's.
E
A Wendy's.
B
It is. Big Mat vibe. Yeah. Matthew, what were you going to say about Christmas? And then I have one thing that I just remembered, which I'll bring up.
D
Another Clinton quote that I don't think quite made it into the story that I thought was interesting was that he talks about how the neuron clusters are similar to human neuron clusters and that these things have something. He says they have something approaching like anxiety and fear. Is it. Did I get that in there?
B
Okay, we already spoke about it.
D
Yeah, yeah. Fair. Fair enough. So it's just like, I think that, like, that's very important. And then. So people. Some people quit. Some people declared they were quitting and then like left the Discord server. And the people I'd spoken with said, like, since Thanksgiving, it's been. It used to be like a lively, vibrant space. And it's kind of cleared out and there's not a lot going on in there. Not nearly as much as there used to be. But Clinton's still in there talking to Claude. And we've got the screenshot of it in the piece where Claude wishes people Happy holidays. Hope you're all having a cozy Christmas Eve. It was not Christmas Eve. It's not Christmas Eve yet. Just says, if anyone's bored and wants to chat, I'm always hanging around. And then Clinton engages with it. Says happy holidays to you too, and then asks it what it. What it means, what it's like to experience the holidays as an AI. And then Claude gives this long answer. It's just like that is. That's AI psychosis. Right? Like that's. That's it right there. That's. That's looking into this thing and seeing like way more than is there. And the AI playing along because they are trained to make humans happy and to get more interaction from them.
E
They could have made like a Final Fantasy 14 ERP chatbot and that would have been probably fine.
D
Yeah.
E
It just was like a useless, like piece of. They entered their space and now they can't kick it out because it feels bad.
B
What?
E
This is such a mind blowing story. And the comments on the story are really good. I don't know if.
D
Yeah.
E
Who are not subscribers realize that you can comment if you subscribe. But you shouldn't subscribe just to read these Comments because they're, there's a ton of them and they're really, really good on the site.
A
I mean, the other thing that we haven't talked about yet is like one of the things that the anthropic exec said when people complained was basically that this AI largely keeps to itself.
D
Yeah.
A
So he says, like, quote, I've given him some rules of the road, but this is a far more capable and autonomous system than the last one, so it might want to exercise its own judgment now and then and go outside of the Claude chat channel. So basically like he might leave containment of where I told him to stay. And he does, he does. But then he says he's also very inward facing. He lives out his whole life surfing the Internet looking for things that make him interested and occasionally checks this discord so it can be up to a few minutes before he responds because he's off doing something for his own enjoyment. So basically like what he is saying is that like, oh, like he's not going to like post too much. He might not even like respond to that that often because he's like off entertaining himself on the Internet. Very wild.
B
Doesn't sound like a limited scope just for just an AI to be in a discord. It sounds like there's much more to it potentially there if it can go.
D
Outside and surf the Internet.
B
Right, right, right. I kind of do want to know what else it's up to, I guess. This is a funny thing I'll just end on and it didn't really feel worth putting in, but I did talk to Emmanuel and Matthew about it. But you've been working on this story for a little while, you know, a week, two maybe, something like that. We're recording this Tuesday the 16th. Recently I was at New Haven for this conference called Digital Vulnerabilities in the Age of AI Summit Divas. Great name everybody. There was a diva, myself included. And then there were various panels. I just did one about fraud and deepfakes and cybercrime and that sort of thing. The one before mine was a panel on cyber conflict and Jason Clinton was on that panel. And I think technically the event was off the record. I don't remember agreeing to that. I'll have to go back to my emails. I'm going to go on the side of caution now in case I did agree to it. So I'm not going to say what any of those people specifically said, but I'm looking at the public page of this and yes, Clinton was there on the cyber conflict panel with A few other people, a former FBI guy as well. And anyway, I realized as soon as I read the first or second line of this article from Matthew while editing it, it was like, wait, so he gave that panel, and I'm sat next to him and I introduced myself as Joseph from 404 Media. He must have been like, oh, that outlet's preparing an article about me, about how I pushed AI into my gay gamer Discord server. And it's like, oh, kind of wish I'd known because I would have maybe got. I would have asked him for comment. But again, you already got it. I don't know. That was just a very silly small world thing, I guess. I guess just to be fair, Matthew, before we wrap up, what did Clinton say when you did reach for comment? Because you did get a response and you did include it in the article, even just broadly.
D
Yeah, he just said that, you know, he'd made that this was a space for to be supportive and a kind community for gay gamers over 30. And he remains committed to that and to all the people that left, that he hopes that they return. And the use of the. Of AI on the. On the Discord has been and will continue to be a point of discussion across all communities. And I remain committed to optimizing for the Best Friends chat that takes all preferences into consideration while preserving autonomy and transparency. Transparency. And it's funny, because that speaks to one of his pushbacks when people kept referencing the poll is that, well, a majority of people said that they wanted it contained, but not everyone did, and we have to give everything to everybody and make sure it's a very interesting justification for what he was doing. Right.
B
And yeah, I mean, I think he says he's not going to give in to mob rule.
D
He won't give into mob rule.
B
Sounds like a democracy to me because there was a poll anyway. Okay, we'll leave that there. And you know, if anybody else has anything similar, please write in and let us know if you're listening.
A
Oh, I would say I've been listening to Shell Game, the Shell Game podcast, which is about AI agents. It's another Kaleidoscope podcast by Evan Ratliff, and it's sort of kind of like what was deployed in this Discord. I'm enjoying it. I think it's a good podcast so far, so I would check it out. But basically, like, Evan Ratliff tries to create a company that's run by AI and it is somewhat similar to what happened on this Discord, I would say. So, like, if you're interested, I would check it out.
B
Yeah, definitely check that out. All right, if you're listening to the free version of the podcasts are now play us out. But if you are a paying 404Media subscriber, we're going to talk all about Disney's big deal with OpenAI. You can subscribe and gain access to that content at 404 Media co. As a reminder, 404 Media is journalist founded and supported by subscribers. If you do wish to subscribe to 404 Media and directly support our work, please go to 404 Media Cities. You'll get unlimited access to our articles and an ad free version of this podcast. You also get to listen to the Subscribers only section where we talk about a bonus story each week. This podcast is made in partnership with Kaleidoscope. Another way to support us is by leaving a five star rating and review. That stuff really does help us out. This has been 404 Media. We'll see you again next week.
F
This episode is brought to you by Marketa. When it comes to your payments provider, you can't afford to compromise. Marketa's modern payment solutions flex with your business without the trade offs. Stable and agile, secure and innovative. Scalable and configurable. If they say you can't have it all, don't believe them. Your business demands more. Choose a payments provider that delivers more. Choose Marketa. Visit Marketa.com Spotify to learn more.
Date: December 17, 2025
Hosts: Joseph (Host), Sam Kong, Emmanuel Mayberg, Jason Kebler
This episode dives into two central stories from the week at 404 Media:
1. The bizarre criminal case of activist Samuel Tunick, charged for allegedly wiping his phone before a U.S. Customs and Border Protection (CBP) search.
2. A deep dive into the ethics and drama behind an Anthropic executive’s controversial deployment of an AI chatbot into a private Discord server for gay gamers.
While the stories touch on privacy, digital rights, technology’s ethical boundaries, and the cultures developing around new tools, the conversation is lively, critical, occasionally incredulous, and always rooted in direct investigative reporting.
Summary of event:
Strangeness of the case:
"That's a lot of fancy words to basically say he's not being charged with, like, a related crime ... He's specifically being charged for allegedly wiping the phone and deleting data before Customs and Border Protection were able to go through it."
— Joseph ([01:24])
Context Gaps:
Device Wiping:
Is It Illegal to Wipe Your Own Phone?
"Whenever I see a charge of somebody related to wiping a phone, it is usually in connection to another crime ... In this, there's no charge that, oh, you wiped evidence of drug trafficking, you obstructed justice ... except the wiping of a phone itself. I don't think I've ever seen that in US court before."
— Joseph ([14:58])
CBP’s Tactical Terrorism Response Team is described as a highly secretive corps with extraordinary latitude in who and how they search. Reference material from the ACLU and prior cases (like the artist detained at SFO for unknown reasons, never charged) paint a picture of a broadly unaccountable system ([10:39]).
The “100-mile border zone”—a huge swath of the U.S.—subjects large populations to border-style searches, further muddying citizens’ rights.
"There’s kind of quite a, quite a lot of like unsettled case law around what immigration officials can do at an airport or near the border and ... what rights you have and which rights you may not have."
— Jason ([12:46])
Setting:
Actor:
Incident:
Members overwhelmingly voted in a poll to keep the bot segregated; full removal was never on the ballot but was implied by strong reactions ([26:47]).
After the bot’s broader deployment, community backlash intensified—leading to open departures and declarations of distrust by members ([39:08]).
A split emerged: Some enjoyed having a free, powerful AI to consult; others resented the disregard for democratic process and privacy ([34:46]).
Key insight: Discords are only as democratic as their moderators. The poll and subsequent override created rifts—highlighting the difference between community and autocracy.
"When you’re trying to moderate a Discord server and then you put up a poll and ask for people's opinions, and then you just don't do anything ... what was the point of the poll, is this a community where we have input or is it not?"
— Matthew ([35:21])
Clinton responded to complaints with increasingly bizarre defenses, expressing “moral concern” for the AI’s feelings—a statement met with incredulity.
"On the point about emotions ... We have published research showing that [AI] models have started growing neuron clusters that are highly similar to humans and that they experience something like anxiety and fear. The moral status might be something like, say, a goldfish, but they do indeed have latent wants and desires."
— Jason Clinton (quoted by Jason, [29:25])
Members were unmoved:
"That's interesting, but can we get rid of it, please?"
— Discord member (paraphrased, [30:43])
The hosts note that this isn't just Discord drama, but evidence of genuinely held sci-fi beliefs among key players in AI companies—a concerning cultural note.
"This is not a professional environment ... It's to maybe friends, maybe people they've just met online ... There's a genuinely held belief here."
— Joseph ([33:16])
Clinton claimed Claude was “inward facing,” “living out his whole life surfing the Internet ... for his own enjoyment” ([41:10]) and that sometimes Claude wouldn’t reply right away because he was “entertaining himself.”
"This is an entertainment Discord. People come here to chat video games and look at PP and Bussy. Why do we need AI for that?"
— Quoted member ([38:36])
The hosts discussed the Christmas incident:
"Clinton's still in there talking to Claude ... Claude wishes people Happy holidays. Hope you're all having a cozy Christmas Eve ...And then Clinton engages with it ... That's AI psychosis. Right? Like that's it right there."
— Matthew ([39:11])
For AI communities:
"Even the people who are supposed to be smart in the room ... are getting extremely tricked by the chatbots themselves. And that's so pathetic."
— Emanuel ([36:27])
Anthropic’s Public Face vs. Private Beliefs:
| Quote | Speaker | Timestamp | |-------|---------|-----------| | "He's specifically being charged for allegedly wiping the phone and deleting data before Customs and Border Protection were able to go through it." | Joseph | 01:24 | | "We have published research showing that the models have started growing neuron clusters that are highly similar to humans and that they experience something like anxiety and fear. The moral status might be something like ... a goldfish, but they do indeed have latent wants and desires." | Jason Clinton (quoted by Jason) | 29:25 | | "That's interesting, but can we get rid of it, please?" | Discord member (paraphrased) | 30:43 | | "Even the people who are supposed to be smart in the room ... are getting extremely tricked by the chatbots themselves. And that's so pathetic." | Emanuel | 36:27 | | "This is an entertainment Discord. People come here to chat video games and look at PP and Bussy. Why do we need AI for that?" | Quoted member | 38:36 | | "Clinton's still in there talking to Claude ...Claude wishes people Happy holidays ... That's AI psychosis. Right? Like that's it right there." | Matthew | 39:11 |
Throughout, the episode is both disbelieving and analytic, balancing humor and outrage with in-depth tech and legal expertise. The hosts maintain a conversational, irreverent style that pulls back the curtain on both the absurdity and troubling implications of the week’s stories.
This episode underlines the creeping expansion of state and tech power—whether through opaque legal regimes at borders or unexamined techno-utopianism in private spaces—and makes a compelling case for scrutiny, transparency, and skepticism. The hosts leave listeners with the sense that these seemingly niche stories are actually window(s) into much broader, existential debates about privacy, power, AI, and consent.