Crypto Mines in Crawl Spaces + Youtube as Infinite Storage +Voice Biometrics Aren’t Doing So Hot Right Now.
TL;DRA Massachusetts teacher allegedly built a crypto mining operation at his school; researchers encoded binary files into YouTube videos as free storage; and AI voice cloning tools were tested against voice biometrics systems.
A chat episode about the a teacher who allegedly built a crypto mine in a school crawl space, researchers using Stable Diffusion to read MRIs, the state of voice biometrics, a credential stuffing attack at Chick-Fil-A and whether their sauce is gross or just gross looking.
Transcript
Machine-generated transcript; may contain errors.
Speaker 1: Give this a listen. Do
Speaker 2: I listen to this thing?
Speaker 1: Yeah. Give it a listen.
Speaker 3: The quick brown fox jumps over the lazy dog. The quick brown fox jumps over the lazy dog.
Speaker 1: What do you think that is?
Speaker 2: You want you want it's definitely generated. Okay. Listening to it over and over. Give me one sec.
Speaker 1: Mhmm.
Speaker 3: The quick brown fox jumps over the lazy dog. The quick brown fox jumps over the lazy dog.
Speaker 2: It's kind of interesting
Speaker 3: The quick brown fox jumps over the amount
Speaker 2: of background noise. Like, it almost sounds like a shitty microphone, which which is interesting because I'm assuming it's generated because of the artifacting in the ends of the words.
Speaker 1: Okay.
Speaker 2: I'll listen to it again. One sec.
Speaker 1: Give it give it give it one more pass.
Speaker 3: The quick brown fox jumps over the lazy dog.
Speaker 2: K. Wait. What do you what what's the question? Hit me.
Speaker 1: So I wanted to check-in on the state of voice biometrics and see how they're doing, in light of all the big leaps AI has taken over the last little bit.
Speaker 2: K.
Speaker 1: And I wanted to try out some of the tools that have shown up in the last say six months to a year. Last time we did a deep fake voice episode was almost two years ago. What's changed? Right?
Speaker 2: Yeah.
Speaker 1: And in the process of doing that, I ended up asking a very, I think, unsettling question, which is what happens when you start to lie to these systems? Specifically, telling it that you're training it on one voice when you're actually training it on two. I was gonna say,
Speaker 2: it sounds like layered voices. That was actually gonna be the next thing I was gonna say.
Speaker 1: So what if instead of giving it, say, five minutes of me talking, I unmuted your track Give it five minutes of us. And gave it five minutes of hacked podcast talking.
Speaker 2: So, wait, this is a hybrid of us?
Speaker 1: This is this is a robot that thinks it was listening to one person, but was actually listening to us talk, and it produced this voice out of that.
Speaker 2: Give me one second. I gotta listen to it again.
Speaker 3: The quick brown fox jumps over the lazy dog.
Speaker 2: You know what the weird thing is, though? It's like it I can hear bits of my own voice in it, which is why I was You kinda hear it. Weak, but it doesn't really sound like you. You know who it actually sounds like is Moe from our office. Interesting. Have you listened to it? Like, maybe it's just me.
Speaker 1: You hear a little bit. Yeah. It it does sound like someone I've I've heard before. I think I must have, like, listened to a podcast that sounded like this. I I almost called him this guy. It's not a real person. It's it's made up. I I don't know what it is. It's an AI creation based on
Speaker 2: On us.
Speaker 1: It's based on us.
Speaker 2: It's based on us. So if we were to have a voice child, your silky beautiful voice and my terrible nasal voice becomes
Speaker 1: Exactly.
Speaker 2: This voice.
Speaker 1: It is the son of hacked.
Speaker 2: The son of hacked.
Speaker 1: The son of hacked.
Speaker 2: Wow.
Speaker 1: So let Can we
Speaker 2: just feed it a script and have it record an entire episode?
Speaker 1: On a long enough timeline, I think it could take over for us. Yeah. I think so.
Speaker 2: Let's go. Let's go.
Speaker 1: So I think we talk about that, the state of voice biometrics. I think we talk about a a Massachusetts teacher that is facing charges for allegedly building a elaborate cryptocurrency mining operation in the school where he worked. Allegedly. Allegedly. And a new game that I'm calling, where would you hide a clandestine crypto mining operation given your circumstances? It's like a new road trip game.
Speaker 2: No. It's gonna be like the Geo Explorer. Like, it shows you an office building and you have to, like, jump to somewhere where you would hide a crypto op. I mean Yeah. Totally. Pretty good game. Well, because we're gonna talk about teachers, I figured we'd talk about some ransomware stuff in schools. You know, some real depraved stuff where, like, you know, you're holding records on at risk children for ransom. So I'm gonna take the the less fun version of the of the teacher stealing electricity to make free crypto and and go and go dark.
Speaker 1: So School crimes of varying degrees of fun. And then I wanna talk about people so people have taken, stable diffusion and MRIs and slammed them together, and it is pretty remarkable. So let's talk about all of that here on this chatty episode of Hacked. Yeah. How are you how are you healing up? How are you feeling?
Speaker 2: I feel
Speaker 1: Back to January?
Speaker 2: No. I'm like a hard '68. Like a specific I'm at the point Yeah. Like, I'm not quite 70, but I'm not far from it. Like, I can get up and go do things again. Like, I'm I'm resuscitated to the point of functioning. I'm just not functioning at the point of exceptional,
Speaker 1: you know? Maybe by next episode.
Speaker 2: So I think I'm in that
Speaker 1: We're gonna get to watch you heal, essentially.
Speaker 2: Slowly but surely. Slowly but surely. Slowly but surely. So where do you wanna start? I kinda I kinda wanna start with the YouTube hack.
Speaker 1: I mean, I kinda wanna start by thanking just the best dang Patreon on the Internet. You know what I mean? Oh. Oh. I think that's where I personally wanna start.
Speaker 2: You're the real you're the you're the real you're good at this. Like, do you wanna thank Devin and Will and Ben?
Speaker 1: You know I wanna thank Devin and Will and Ben.
Speaker 2: Reed and Jill.
Speaker 1: But don't forget about Namclaw.
Speaker 2: Davey, Brian.
Speaker 1: You're gonna forget about Namclaw? Yeah. That would be like forgetting about Gordon and Sulus.
Speaker 2: I would never forget about Namclaw.
Speaker 1: Which is almost as bad as forgetting about Rob Cass Barberton, whose whose name I had never said out loud until this moment. Rob Cass Barberton. A lot of appreciation for.
Speaker 2: You mean Rob Rob Cass Barberton. And Kelly Fitzgerald, Matt
Speaker 1: The one and only.
Speaker 2: John. We've had a bunch of people kind of flood into the Patreon and into the Discord, and you gotta say, you know, we love you all as much as we love the rest of our patrons and other people we hang out with. I promise I will post a photo of my computer set up at home to go with the rest of the other people's computer setups.
Speaker 1: Rig in there.
Speaker 2: Let's just say that there's a little bit of there's a little bit of chaos on my desk right now that I have to deal with.
Speaker 1: So Sure.
Speaker 2: Mostly cabling chaos. I'm more ashamed. Like, I have because I have two computers on my desk. One's a Mac and one's a PC, and I have them audio bridge and I have them like a bunch of stuff. So I have SPDIF cables running between them. And, like, I just have so much cabling that I'm ashamed to post a photo because I will be judged severely for how much cable disaster my desk is.
Speaker 1: My, my desktop tends to look I'm I'm quite proud of it. I think I've laid it out quite nicely. I think it's a good looking desk. And I was kinda, like, fawning over it a little bit the other day, and then my my partner, just, like, dunked on me. She's like, yeah, it looks really, really good. What what about under there? I was like, oh, that's just the bottom of the desk. It doesn't matter. She's, no. How's it going? Right right under that way. And it's just cable hell.
Speaker 2: I have a big desk. I have a lot of cable. Scott. And the underside of my desk is amazing. Like, I have coiled up double sided tape. So, like, there's switches and power bars mounted to the bottom of my desk. Like, every like, I I spent a day sitting on my desk when I first got it. And then I added some more stuff to my desk. And you got a little bit more. Interlink my computers for audio connections. And I wanted to do, like, a few things. And then all of a sudden, there's just a pile of cables that aren't dealt with under the desk. And then you have to rip everything apart and redo it all, and I just don't know if I'm there yet. Like, I've I've I got too much to do, and I'm not bored enough that I want to spend seven hours recabling my desk. So maybe what I'll do is I'll take all of the cables off of my desk. Like, I'll unplug everything, take a photo, and then digitally put on backgrounds on the monitors and stuff. So people can see it, and I'll post that and be like, look at how beautiful this is.
Speaker 1: My shit's impeccable. Exactly. If you if you wanna join all that fun, support us on over at hackpodcast.com, which redirects to our Patreon. And for context, we are recording this pretty early. Recording this on, March 6. So if you have, been kind enough to support us after that point, but before this episode goes up, that is why you're not, included in this list. It's a really well organized legible list, but we'll we'll get you in the next one.
Speaker 2: True, true, true. K. Can we talk about YouTube now?
Speaker 1: Heck, yes. We can talk about YouTube. I'll talk about YouTube all day. What do you wanna talk about?
Speaker 2: I just this this story of the people using YouTube as a storage facility, I love it. Like, it's just it makes like, when you're encoding binary data into video frames and then just, like, dumping them onto a free video platform, granted you're, like, just I guess you could encrypt the data, but, like, it's kinda brilliant.
Speaker 1: It's incredibly it's super fun. So the story here is, a username, Dvorak Dwarf, which is a really, really good that's a fantastic username. And for everyone that knows about dvorak, I know a guy who uses dvorak. I had to Google dvorak to remember what dvorak was. But in any case, dvorak dwarf, posted this, like, proof of concept essentially of how to encode binary files into video on YouTube. Interestingly, it it it does make the file size bigger.
Speaker 2: Mhmm. I think
Speaker 1: it makes it four times larger than the original file, but it does allow you to post data for free on YouTube.
Speaker 2: You just have a monstrous like, instead of having a paying a Google Drive subscription, you just need a piece piece of software that can retrieve because you're getting I don't know. It's like a it's like a like a fun hack.
Speaker 1: It's a super fun hack.
Speaker 2: No. It's not. Anyway, I just wanna remember, Dvorak is the weird keyboard layout. Right?
Speaker 1: Dvorak is the weird keyboard layout. A mutual a good mutual friend of ours uses it or used it for years. And every time I would try and
Speaker 2: Let's give him a shout out. Logan uses it. Right?
Speaker 1: Our friend Logan uses it. Our friend Logan uses it. And every time he would, Yeah. Any of us would go to type on his computer, type anything.
Speaker 2: Be chaos.
Speaker 1: It's just just madness. Like, the button has a letter written on it, and when you hit the button, not that letter.
Speaker 2: Yeah. I feel like I feel like I feel like Logan's at the point now, though, that with Dvorak I imagine most of our users are probably also mechanical keyboard nerds. Yep. And they probably reconfigure their keys
Speaker 1: like I would. Sure. Reconfigure configure it to Dvorak.
Speaker 2: Like, at least you know when you hit you hit the a button, you're getting an a button. Exactly. Exactly.
Speaker 1: Yeah. That's pretty good. Yeah. That's a really good call.
Speaker 2: I wonder I wonder I wonder if you can get Dvorak They must. Caps. He must be able to.
Speaker 1: I mean, it's the same key set. It's just have they shuffled it around for you.
Speaker 2: Well, but but the are the keys have different the keys have different slopes. Right? Like, that's the big thing with key caps. If you're a mechanical keyboard nerd, which I'm which I make sense.
Speaker 1: That's true. Your j needs that little bump. The f needs that
Speaker 2: little bump. But it's not just that. It's actually if you take all the keys off and lay them, there's a profile to your keyboard for how how the keys I didn't know that. Function with the layout of the keyboard for reaching and sloping and angles and stuff. Like, if you're you're a keyboard nerd, which I've now exposed myself as being,
Speaker 1: Yeah. Gotcha.
Speaker 2: You'll know this. So so so there must be Dvorak key caps. And if there and if Logan doesn't have a Dvorak keyboard, I know what I'm getting him for a wedding party.
Speaker 1: So in order for Dvorak Dwarfs hack to work, and I was curious about this when I read the headline, is that I know that YouTube's compression is not only pretty severe, but kind of constantly changing.
Speaker 2: Mhmm.
Speaker 1: So this is really interesting question of how do you do this in a way that YouTube's compression isn't going to, like, just result in a bunch of data loss. The quick sort of early solution at this point was just make sure that you're not using any RGB. This has to be entirely black and white pixels. The blocks have to be of a certain amount of, like, certain kind of size. It's not totally clear at this point if this method is gonna violate YouTube's terms of service. So there could be a situation where if this reached, like, became a thing people were actually doing where YouTube could basically just apply stronger compression to corrupt the data stored inside of these staticky Sure. Data storage videos. It's really, really cool and interesting.
Speaker 2: Well, the the way that they've done it is I would say, is very what's the best way to describe this? It's it's very low Yeah. Efficiency. Yeah. Like, the size of the size of the blocks and stuff, like, to to ensure that it doesn't get corrupted away. Mhmm. Like, you you could do this so much better with actually just saving the data, but but that's
Speaker 1: That's missing the point entirely.
Speaker 2: Exactly. Exactly. It is what it is. It's just like black and white squares that represent binary things, like, laid into frames per second, laid into a video. And you could probably build in some form of parity check-in it and things like that to rebuild corrupted data if it were possible. Like, almost use like a like a RAID algorithm across different YouTube videos. But, anyway, does it does it doesn't matter. Still a cool thing.
Speaker 1: Super.
Speaker 2: Like a happy hack, something neat and fun that people do. And I, you know, I think it's a good story to kick us off before we get into holding
Speaker 1: Yeah.
Speaker 2: At risk child information hostage.
Speaker 1: Hackaday when they posted it, they they brought up an episode of a podcast they did where they they experimented with baking information into the other audio file, which sent me down a reading rabbit hole of the technique that they use there, which is something I had never heard about, which is called the Kansas City standard or the byte standard. Yep. I never heard about this. Goes back to 1975 and is like the foundation of, like, storing data on tape. Incredibly interesting.
Speaker 2: Mhmm. There's there's, like, what we're talking about weird hacks is actually, like, a for profit hack that's like this. Mhmm. Like, you're an audio guy. Have you ever heard the term MQA? MQ. Does this ring a bell?
Speaker 1: No. It's not.
Speaker 2: Masters quality audio? Okay. Anyway, so so, like, titles titles, HiFi, subscriptions, stuff like this. Like, you know, there's, like, different file types and
Speaker 1: bit rates
Speaker 2: and all the rest of this stuff. MQA, what they did is they they took the inaudible space inside of, like, the way that, like, there's all this frequency that humans can't hear. Right? So so our audio files still encoded. So there's, like, most people can hear above 17,000 kilohertz. Like, dogs can and, like, bats can and, Yeah. Sure. But most humans, like, over over 20 k is, like, you'd have to have super ears to hear it. But the file is still in code to, like, 48,000 k.
Speaker 1: So there's empty room and everything.
Speaker 2: All of this. Exactly. So what MQA did is said, you know what? Let's just and I'm, I'm by no means an expert in this, but this is the way that I understand it. So if I'm incorrect here, please read me on Twitter, and I will apologize. But they've taken inaudible space and they've used it to essentially boost up the resolution of the audible space without changing the file size. And that's like the big it so when we're talking about like compression types and and using things kind of in a non traditional way, that's that's kind of what this dynamic is. There's a big argument about whether it actually increases the quality or decreases the quality and a bunch of people that hate it because it's created by a company and they charge licensing fees.
Speaker 1: Right.
Speaker 2: There's a whole political aspect around it that I don't wanna get into and don't rib me on Twitter about. But, anyway, just a inter interesting way to to to look at a problem and solve a problem by just looking at outcomes rather than than the inputs. So
Speaker 1: Is it better or worse than the 300 bits per second I can store on tape using the Kansas City standard? Because I'm a traditionalist. I like that that old school, Kansas City standard sound, you know? So it can be that.
Speaker 2: I I feel like the Kansas City sounds is more of like the d s d in audio, Direct Stream Digital. I don't know if you know what it is, but it's essentially a format used to preserve and encode, like audio files. And DSD files are monstrous. Granted, they're essentially a bit perfect or they're supposed to be, like, pretty close to it, I think, but they're they're massive. Like, you're talking about, like, I think I've got a DSD album and it is multiple gigs for, like, one album.
Speaker 1: Interesting. Well, find this episode of Hacked stored as thousands of static frames on YouTube. There was a a comment there was a comment that I really liked, where someone said, like, this looks so much they were talking about the the data that had been encoded using the system on YouTube. And they said, this looks so much like TV static. It makes me wonder, what if there are clandestine analog TV stations broadcasting quote static that's actually data? And I love that as like a spooky sci fi concept that all the static just that you can pick up on a TV is actually some data being broadcast from somewhere. Like that's some some quality, what a spooky past or whatever those are called. R slash no sleep stuff. I'm here for it.
Speaker 2: I I was gonna say that's like an r slash writing prompts.
Speaker 1: Totally. That's what
Speaker 2: it feels like to me. Yeah.
Speaker 1: It's great. And you you finally crack how it's been, how it's been encrypted, and it it says some stuff.
Speaker 2: So where do we where do we go from here? We could talk about chick Chick fil A. You know, we had a had a long lengthy conversation with Chick fil A before we pressed record. I tried it at Christmas because we're Canadian. We don't have it, but I tried it at Christmas. And, I the obsession with the Chick fil A sauce
Speaker 1: But you said you love it. I mean, you said it
Speaker 2: I'm gonna I'm gonna mix it at any of these when I say this, but it disgusted me. You're not here for it? The Chick fil A sauce, it just it it just looks so unhealthy. Like, it just the the glimmer to it and the shine, the like, just the way it moves.
Speaker 1: Like, it doesn't move
Speaker 2: like natural sauce. Yeah. Literally literally, it's just of
Speaker 1: its own accord. I just
Speaker 2: I'm not there for it. I'm sorry. I'm gonna get roasted because I know Chick fil A is like a loved institution in The States.
Speaker 1: But it's not this sauce because we talked about it for for a minute before we started recording. It's a shade of yellow. Yeah. I wouldn't say it's a shade of yellow. I you see often in nature or food. But it's certainly a shade of yellow.
Speaker 2: Yes. Soy soy the soybean oil base really gives it a texture and a dynamic that is feels unnatural to me, which therefore makes me not wanna eat it. It does I will say this, when you do put it in your mouth, it does taste good. I mean, you buried that leach
Speaker 1: and that it's actually delicious. This is this is all aesthetic complaints. You you don't like the way it catches the light. But in terms of putting it on a chicken sandwich, you're like, I'm here
Speaker 2: for it. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Well, like but, like, I'm a I think who doesn't love a good honey mustard? Sure. Sure. Like, I even make my own honey mustard with just honey and mustard, and it's delicious. The Chick fil A sauce is like essentially a kind of a honey mustard and it is it does have its own sweet salty insane taste, but I just aesthetically, I'm a texture guy, and I just can't do it, you know? Like, textural foods. Like, if it's I can't eat oatmeal. Like, I'm telling you guys all something about myself right now. I cannot eat oatmeal. The texture of it's just too too slop like to me. I can't do it.
Speaker 1: I I rock with some oatmeal. Yeah. I really rock with oatmeal. So, presumably now your information regarding the this this this sauce order was part of the compromise that occurred on December 2, 2022
Speaker 2: Yeah. Yeah.
Speaker 1: Yeah. In in which Chick fil A users accounts were stolen in a credential stuffing attack. I had to kind of reacquaint myself with this. Credential stuffing, blooping computer reported on this. It's essentially just we're gonna go find a bunch of credentials in some other data like something entirely unrelated to Chick fil a. We're gonna buy a bunch of credentials, emails and passwords, and then we're gonna automate a system over here. Trying. Exactly. We're just gonna maybe you use the same email and password combo on Chick fil A's, user profile, whatever it's called. I don't know what their point system is named. Maybe you use that same combination there that you used over on this site over here that was compromised in a breach.
Speaker 2: Which is maybe the single largest reason to have a password manager. You know, let's leave let's leave LastPass out of it. I know we talked about talking about it more because it seems like it's just a never ending story. But, but, like, that's the I would say is one of the largest reasons to have a password manager. Like, in the problems with password episode that we did, like, five years ago, that's, like, the big thing that's changed now is, like, people are hacking your like, I got a a notification from, like, do you remember Evernote? Do you remember that service?
Speaker 1: Oh, yeah. Yeah. I do. I used to
Speaker 2: use Evernote a ton in, like, 2010. And I got an I got a I got an email the other day being like, hey, your password request is in. And I was like, Excuse me? Like, I don't think I've logged into my Evernote account in ten years. And so I logged in, changed my password, set a two factor authentication and just moved on with my life. But but somebody clearly credential stuff me probably off of some hack to some old system, and I'd had to use the same password. And, anyway, so it's just it's yeah.
Speaker 1: Can you imagine what would have happened if you'd used those same credentials for your Chick fil A
Speaker 2: account? Oh my god. I will say, let's just talk about Chick fil A here again. Is so the lineup was ridiculous.
Speaker 1: Not about the hack. Just about the the shopping experience at Chick fil
Speaker 2: a. I I we gotta talk about this. This is there was, like, four lanes of drive through. Wow. They have, like, an Uber Eats and, like, skip the dishes, kind of like
Speaker 1: flying out the door
Speaker 2: in their own lane in the drive through just to pick up takeout orders for delivery. Anyway, it was like, I wish I owned Chick fil A after seeing the operation. And, like, the school buses of, like, sports teams showing up and, like, waiting in line for food.
Speaker 1: When they're moving units of sauce too, and it's clearly just soybean oil. I mean, profit margins must be high.
Speaker 2: You know, I'm gonna make an economic argument for why you should have a Chick fil A account even after the hack, because the lineup was massive. But then I was noticing that, like, people would, like, walk in, walk up to a little desk, and grab their bag, and leave. And if you order from the app, you could skip the line entirely. So there were people that were in line with me, who were downloading the app, would buy and pick up their food before we'd move like two people. And I was like, this is why you have a Chick fil A account, even if it gets hacked. There's the value right there.
Speaker 1: Well, that about covers this this hack of this account and why this account rules.
Speaker 2: That about covers it for this month. Take care.
Speaker 1: And that's catch you on the next one. It's a pretty quick turnaround on, so Christmas twenty twenty two, bleeping computers bleeping computer emails Chick fil A about reports of an account breach. January 2023, they say, we're looking into it. Meanwhile, there are Telegram channels showing people purchasing these accounts and sharing their purchases made through these purchased, these stolen Chick fil A accounts. And then in 02/12/2023, Chick fil A confirms there was a credential stuffing attack that happened over the holidays. 71,473 Chick fil a accounts hacked, personal information including name, email address, your Chick fil a a one membership, which apparently is the name of this program, your mobile pay number, your q QR code, a masked version of your credit card, and the amount of Chick fil A credit on your account, if anything, may have been accessed by these actors.
Speaker 2: So so the in the screenshots they provide the you can see the different accounts. And, like, it's like, buy this account. It has a $2,000 balance on it. Pay $200 for it.
Speaker 1: And Chick fil A points, essentially.
Speaker 2: It's 10¢ on the dollar. But, like, I could you imagine I don't know. I don't know if it's well, let's that's a whole different conversation about if you had $2,000 worth of Chick fil A points. But, like, let's let's not get into that. But with the but a $2,000 balance inside your they must have, like, charged the Visa to add gift certificates, I would assume, or something. And then before the Visa got canceled and the account got canceled or something, but, you know?
Speaker 1: Right. I thought you were saying that there were $2,000 worth of whatever their points were, which would then evoke the chilling math of how much money do you have to spend at Chick fil A to get $2,000 worth of their points. And I was gonna crunch that math.
Speaker 2: Potentially Yeah.
Speaker 1: Okay. I hear what you're saying.
Speaker 2: We're intentionally we're we're intentionally avoiding that. Let's not we're not having that conversation.
Speaker 1: Need to know that number. We already heard what happens when an AI slams our voices together. Enough love crafty and horror for one episode. Thank you very much.
Speaker 2: Oh, man. Could you I don't know. Whatever. I assume it's gotta be 1%. So that means you're, like, one or 10%. Let's hope 10%. So to get $2,000 in points, that means you've spent $20.
Speaker 1: Conservatively $20.
Speaker 2: At 1%. No. At 1%, it would be $200. And if you've spent $200 at Chick fil A, wow.
Speaker 1: Wow. You really do love that yellow, yellow sauce. Chick fil A forced a password reset. They froze all the funds, did what you'd expect. But it is a good reminder from that problem with passwords episode. Hey, even in, like, seemingly very, I would say low concern accounts, Not to editorialize the value of a Chick fil A account, but even in those do not reuse your passwords. Try and have something unique for each individual one because that eliminates the concern of credential stuffing, which is still a very viable tactic that people are using.
Speaker 2: Oh, a million. Like, the yeah. A million percent. Let's stay on the light side. Let's stay on the light side of life here. The I was creeping on Krebs the other day, you know, just to give them another shout out. Krebs on security. And apparently, Hulu has commissioned and is soon to releasing a documentary series about the Ashley Madison hack. Will watch. Which they've called the Ashley Madison Affair,
Speaker 1: which I
Speaker 2: think is an appropriate name given the fact that, hey yo, Ashley Madison was there to have affairs. But, but yeah. So I I think that's just, like, an interesting pop culture rip on I feel like that could be a hacked production. Like, hacked makes a like, we make documentaries. I don't know if people know that, but Jordan and I have made some some videos together with some friends of ours and colleagues. I feel like that's something that we could have done very well. Maybe we should maybe we should have done a dub when we're not making video games, our video game podcast, talking about hacked and working our day to day jobs, we should be making Hulu series for the Ashley Madison affair.
Speaker 1: Apparently. Yeah. I feel like there's a there's there's a lot of hunger for hacky scammy grifty content right now. It's kinda in the air. A bunch of the stories we've talked about over the last, like, six months are in some stage of active development. There's a real hunger for it. And I'm just really happy to see Brian Krebs in everything. I put him in all of your movies about cybercrime. He's great. That yes. Absolutely. Yeah. Yeah. You love to see it the,
Speaker 2: Yeah. Absolutely.
Speaker 1: Razzlekhan and Dutch episode we did. They're making that into, like, five different shows.
Speaker 2: Beautiful. Beautiful. Did you know what I've do you know what I've seen in, remember the episode we did about the, Crypto Queen? Yes.
Speaker 1: The wait. The the the OneCoin episode or the Razzlecon and Dutch episode?
Speaker 2: Yeah. Yeah. Yeah. No. The OneCoin episode. I I I Are they making that one into something? I have a memory that I read an article, like, a month ago that they are making that into something.
Speaker 1: Oh, no. They are making into something. I think we talked about that in that episode, actually, but it was early.
Speaker 2: Oh, did they? I think
Speaker 1: we might have. I
Speaker 2: Mhmm. I feel like it was a dramatization, though. Like, it wasn't actually,
Speaker 1: Oh, a docuser.
Speaker 2: Not a docuseries. I feel I feel I feel like I've read that it was actually being made into a real show.
Speaker 1: I mean, it makes a lot of sense. They're interesting stories. Like, that one oh, yeah. The double whammy.
Speaker 2: We can leave it there. I I'm not sure if I'm making that up or if I read that late at night one night laying in bed trying to fall asleep. So Why don't we kick it
Speaker 1: over to the, the old commercial break? And when we come back, we'll talk about talk about crypto crypto mines in crawl spaces, right after this. Starting something new isn't just hard. It can be downright terrifying. You put a lot of work into a thing. You're not entirely sure it's going to work out. You're taking a huge leap of faith. I've started a few things. Now I know I was right for believing in, you know, the idea, the product, despite all of those fears and hesitations. But boy, does it sure help when you have a partner like Shopify on your side. Shopify is the commerce platform behind millions of businesses around the world and 10% of all e commerce in The US. From household names like, well, hacked podcasts merch, to brands just getting started, you can get started with your own design studio with hundreds of ready to use templates. Shopify helps you build a beautiful online store that matches your brand style. Did I mention that that iconic purple shop pay button that's used by millions of businesses around the world? I don't know why I wouldn't. I should. It's why Shopify has the best converting checkout on the planet. It also helps boost conversions, meaning less carts, sort of getting abandoned in the parking lot, and more sales for you. It's time to turn those what ifs into sign up for your $1 per month trial at shopify.com/hacked. Go to shopify.com/hacked. One more time, that's shopify.com/hacked.
Speaker 4: When you need to build up your team to handle the growing chaos at work, use Indeed sponsored jobs. It gives your job post the boost it needs to be seen and helps reach people with the right skills, certifications, and more. Spend less time searching and more time actually interviewing candidates who check all your boxes. Listeners of this show will get a $75 sponsored job credit at indeed.com/podcast. That's indeed.com/podcast. Terms and conditions apply. Need a hiring hero? This is a job for Indeed sponsored jobs.
Speaker 5: Sierra has all the best active and outdoor brands for the super athletic stuff, like running gear for cruising up the trail. Whoo. And the super athletic ish stuff, like fishing gear for chilling by the creek. Nice cast. Fitness apparel to push for higher reps. You got this. And golf balls priced so you could afford to lose one. Or a few. Head to Sierra or sierra.com for the brands you want at the prices that let you do it all. From athletic to athletic ish, Sierra's got it.
Speaker 6: Thinking about refreshing the carpet in your home? Now is the time to do it. For a limited time at The Home Depot, get 10% off installed carpet projects on trusted brands like Lifeproof, Lifeproof with PetProof Technology, Home Decorators Collection, and Traffic Master. Plus, with installation starting at just 49¢ per square foot, upgrading your space is more affordable than ever at The Home Depot. Offer valid, 06/11/2026 through 06/28/2026. Exclusions apply for licenses. See homedepot.com/license numbers.
Speaker 2: So I don't think you understand the gold that you just outroed us on there. Crypto mines in crawl spaces, a new docuseries by hacked podcast Pretty good. Getting to Netflix this fall. How pretty
Speaker 1: I'm, where do I sign? Where do I this this shark is interested in investing. Yes. I'm here for it. I feel like that's
Speaker 2: Listen, Netflix. Yeah. We'll take the meeting if you ask, but, you know, only if you ask. We're not gonna come we're not gonna we're not gonna come for you.
Speaker 1: Disagreed. I'm
Speaker 2: just joking. Just joking.
Speaker 1: Please let us in. I'll I'll camp in front of your building. Let us inside. Please.
Speaker 2: We'll buy you lovely dinners.
Speaker 1: All of the Chick fil A sauce that your heart desires, we will show up with it in a big pillowcase. Yeah. Crypto mines and crawl spaces. I I feel like that's probably at this point Crypto
Speaker 2: mines and crawl spaces.
Speaker 1: A pretty, like, thriving genre of story. And there was, a very interesting one that kinda sort of came to a little bit of a head this month. Massachusetts authorities first discovered a very unusual electrical something going on somewhere in the Cohasset High School. And it was announced this month that Nadim Nahas, a 39 year old teacher, has been arrested in relation to it. It's a very weird one. You'd read about this before the arrest and the announcement of who'd kind of been charged with this when it was just a school found some computers that were being used to mine cryptocurrency.
Speaker 2: Yeah.
Speaker 1: You you learned about this before we knew kinda where this was all going. Yeah.
Speaker 2: The first time I I saw this story, kind of flow through my timeline, It was like they'd found this. Like they'd somebody had gone into a thing, saw something. The IT person was like, oh, shit. I know what this is. They brought in external investigators, but I don't think they had figured out who it was. So it's interesting that it turns out to be a teacher. I just assumed it would have been like one of the IT staff or something, but but essentially
Speaker 1: Oh, interesting.
Speaker 2: Somebody had figured out how to leverage public power and, turn that into money. Or I guess into crypto, which, you know, arguably can be turned into money. So
Speaker 1: It can be turned into money depending on what day of the week it is. Yeah. Yeah. So the the the the town's IT director calls the cops saying there's a weird computer rig doing something or other in the school. They contact the US Coast Guard investigative service and the Department of Homeland Security, which is quite the escalation. But they swoop in and they figure out what, allegedly this teacher kinda was was doing. And this really just comes down to so, the the coverage of this spoke with a guy, a crypto expert, who said that the reason something like this might happen is because the cost of electricity in New England is prohibitively high, and so this can't be done efficiently at home, which is a a wild way to understand what occurred here. It did lead me to go down the rabbit hole of researching whether or not, Massachusetts Massachusetts does have expensive power and whether it would be a good place for crypto mining and crawl spaces. And? And the power there is very, very expensive. Yeah. Fifth from the top. It's sandwiched between Connecticut and Oklahoma. A very expensive place to try and mine cryptocurrency. Not a good place to do it, if there's a good place to do it at all.
Speaker 2: A little bit of petty electricity thievery. And next thing you know, you've got a semi profitable business off the cost of the public tax dollar completely. This is I always find this style of, like, crime interesting fraud because it's like,
Speaker 1: sure,
Speaker 2: like physical access to anything is insecurity. You know? Like the like this person just has access
Speaker 1: to the school. Easy way to get caught.
Speaker 2: Exactly. Like, they just have to run extension cords into a essentially, a completely vacant space that no one really probably like, who knows how long it had been there? Totally. Like, he probably started with one or two rigs, and then I think in the photo, you can see that there's, like, ten, twelve, 15 rigs.
Speaker 1: Yeah.
Speaker 2: But it's like Talking
Speaker 1: about cable rats nests earlier, and it's really it's quite the operation you built out in in there. Yeah.
Speaker 2: But it's it's like I remember when I was younger, like, before Wi Fi security got really high and stuff, like, you can build a like a wireless packet sniffing device and just, like, walk into a public place and, like, plug it in. And most people won't know what it is, and it'll stay there forever. Like, you can like places where the public has access to building things that look nondescripting can be like just set aside like these things look like crypto mining rigs. But like you can imagine, like if you go to build a small enough
Speaker 1: It looks like a LAN party in a crawl space. It's It's not inconspicuous.
Speaker 2: But like you can imagine if you if you made these things look more like office equipment, and you just threw one
Speaker 1: Sure.
Speaker 2: Sure. If you just plug them in around, like most people would walk by them a thousand times and not even look at them twice. And it's like there's there's such an interesting it's not social engineering, but it's like social norm engineering where it's like, you know, I expect I expect to see office equipment when I'm in an office. So when I see office equipment in an office, it doesn't trigger my senses. Yeah. Sure. So it's like if he if he if if they didn't put these or allegedly, if they didn't put these allegedly in the in the crawl space, and if they were made to look more realistic Mhmm. It probably could hid them in plain sight.
Speaker 1: What raises an interesting question of how that IT person caught him because it could have to do with power demand. It could also have to do with whether or not he accessed the network infrastructure of the schools. Like, was this thing lighting anything up on the network side saying, like, hey, there's a crypto crypto mine.
Speaker 2: I know he knows called crypto mine one two three four five.
Speaker 1: But then on this on the social side, like, hiding it in plain sight, you wouldn't have what must have occurred here, which is this IT director at some point peeked his head into a hole in the floor and was greeted by a bunch of computers.
Speaker 2: Totally.
Speaker 1: Like that very weird, weird moment where he's like, I think I need to call the coast guard, apparently.
Speaker 2: I'm in the crawl space of my on my like, I'm on land. I'm in a school. I'm in a crawl space. I need the coast guard. It's very strange. I don't understand American legal system. I don't know if you're saying about the coast guard. Is your first call there?
Speaker 1: I wonder if the coast guard, they got that call and they're like, I know we're the coast guard, but we did just get a call about an illicit crypto mine in a school somewhere. But the guy sounds nervous, so maybe we should check it out. Yeah. This this teacher who was allegedly responsible resigned from their position at the school in early twenty two. A criminal complaint was issued against him for fraudulent use of electricity and for school vandalization. The vandalization charge is very interesting. He was scheduled to appear in Quincy District Court for his arraignment this past month. The reason we're hearing about it now is that he never showed up, forcing the court judge to issue a warrant for his arrest. And that's kind of where the story is up until this point.
Speaker 2: Interesting. I wonder if he went, went full CryptoCon and he's, like, disappeared into the the places with non extradition treaties.
Speaker 1: Well, the one the one data point we are missing here right now, is how much of it he mined. So maybe he had the funds to do that.
Speaker 2: Yeah. Maybe he's a quiet billionaire.
Speaker 1: And Netflix, if you want someone to make a doc about it.
Speaker 2: Yeah. I'm talking. The I let's just talk about how they probably found it aside from looking at the hole, like, I assume
Speaker 1: Yeah. Right.
Speaker 2: Power metering at the level that you know what's using it, especially in an infrastructure and and at the scale of a school would be very, very complicated.
Speaker 1: Yeah. That that tracks.
Speaker 2: Network access, I can definitely see. Like, because you would see those nodes on the network, and you would see their network names and their MAC addresses, and you would wonder what they were. That would also, in an infrastructure the size of a school with Chromebooks and IoT devices, like, out to yin yang, like, you can only imagine the level of detail that it would take to be unless they were named something like CryptoMine one, CryptoMine two, Nahas' retirement crypto plan one. But
Speaker 1: Yeah. Sure. Sure.
Speaker 2: But it would be it would be it'd be it'd be tough. I assume truthfully, I assume it's just I shoved my head in a hole, and I saw a bunch of computers whirring away and was like, those aren't supposed to be here and called the police.
Speaker 1: Yeah. It honestly could have been that that he just found computers. It could be that it wasn't even the IT director. Someone finds computers in a crawl space. The person they're probably gonna go to is either the principal or the IT director. Or one leads to the other. And suddenly you're trying to figure out what these computers are doing. This is an aside. For years, I've lived in an apartment building. You you you you know the apartment that I was living in. And,
Speaker 2: I do.
Speaker 1: After years of of living there, I was when we were moving out, I had a conversation with, like, the building manager. And we were talking about just the building and the experience of having lived there and the rent, and we started talking about how, I brought up how, power was not included in my bill. It was factored into my rent. And he said, oh, that's not the landlord's choice. We can't tell there's no individual power meters on any of these units. We don't know how much any one unit is actually drawing. The building's average is amortized out and it is charged as part of the condo fees. And I had this, like, brief moment where I was like, so you're telling me
Speaker 2: I could have been mining
Speaker 1: crypto been running a crypto mine in the crawl space of this building this entire time. I was livid.
Speaker 2: Not even in the crawl space. I rent apartments in that building.
Speaker 1: It's in my house.
Speaker 2: And they're just not even nobody lives in them. They're just Nope. Server rooms full of crypto mines. Yep. Because you had a lot of windows too, so you could just easily create, like, an air like, put some fans, intake fans on some, and exhaust on the other.
Speaker 1: The thermals would have been incredible.
Speaker 2: Especially in in Canadian winters, you could have just Yep. Just hummed power into money.
Speaker 1: Woulda, coulda, shoulda, man. Woulda, coulda, shoulda.
Speaker 2: We've missed opportunities.
Speaker 1: Yeah. Missed
Speaker 2: and I can almost guarantee too that your lease agreement didn't stipulate that you weren't allowed to put a server farm and crypto mine in there.
Speaker 1: Yeah. I don't think
Speaker 2: I imagine we're gonna start seeing that in residential and and commercial rent agreement soon, especially ones that include power. But
Speaker 1: Well, that was such an interesting story. Like, the province that we're both from has relatively affordable in the context of North America, relatively affordable energy. Yeah. And that was why about four four years ago, roughly speaking, there was a huge influx of entrepreneurs from all over the world. And they were their whole thing was we're gonna identify jurisdictions with relatively affordable power, and we're going to, like, airdrop in shipping containers full of servers, basically. And we're gonna boot up these little businesses. And you ended up with this very strange recurring story of very rural communities in Northern Canada, who suddenly there was like a weird worrying off in the distance. And it was that a giant server farm had just been, like, plopped into some farmland somewhere and was mining cryptocurrency. And the second, you know, that math changed, they would vanish and they would go off somewhere else. But for a while, that that hit where we were from, like, reasonably hard.
Speaker 2: Well, the funny enough, like, south of us, Medicine Hat. Because they have their Medicine Hat's like a unique little town. They have their own, like, electrical utility. They they got hit with that and as well as the marijuana industry saw that as an opportunity. Cheap cheap utilities. Yes. So it's like crypto and and and Well grove farms. We're like, yeah. Cheap cheap cheap power. Like, we're in cheap water and power.
Speaker 1: It's like, good.
Speaker 2: Let's go.
Speaker 1: There's a fun parallel there in that the exact same story of a a teacher was running an illicit operation in the crawl space of a school. I feel like in 1975, that would have been a little grow up. They would have found some little fun way
Speaker 2: to to
Speaker 1: grow weed in some far off corner of the school and no one had to know. Exactly. Maybe a crawl space wouldn't work because plants, but it feels like it's got the same spirit about it.
Speaker 2: You know, I I remember when I was, like, really young, I remember hearing about grow ops being busted, and the way that they caught them was power usage.
Speaker 1: Power usage. Yeah. So I
Speaker 2: wonder if, like, I guess weed's legal in Canada now, but, like, I wonder if that really, like when crypto farms were getting big, but weed wasn't legal, I wonder if there was a cross hold where they're like, hey, guys. We got another grow op popping up, and then they go raid it. And it just turns out it's like some crypto farm. And they're like, oh, okay. Never mind. And then they like the next one. And it's like, okay. We got another grow op, and they go raid it. And it's like, oh, it's another one of those crypto things just like consuming so much power.
Speaker 1: Lot teams walking out of buildings with their heads, like, between, like, heads down all forlorn, like, just a bunch of dorks again. Making that weed in there.
Speaker 2: Bad stuff. Teaching bad stuff, school stuff.
Speaker 1: Bad stuff.
Speaker 2: Yeah. So in The UK
Speaker 1: Bad stuff in schools.
Speaker 2: Yeah. Classic ransomware, the Vice Society ransomware group. Go to British school and, they were trying to demand a ransom. And in doing so, they started, posting files about children on the Internet as, like, a proof that they had it, saying, you know, pay our pay our bills. So not a positive story, but still in the education space. Not that having an illegal crypto farm is a positive story, but way less way less chill than than than having a legal crypto farm to, like, you know Yeah. Outing kids.
Speaker 1: Yeah. It's like the collision of two things that we've talked about before, which is the weird part of a lot of ransomware operations where they have to kind of I think we've likened it to the part of a traditional kidnapping where they they they mail someone a finger type thing.
Speaker 2: Exactly. It's the part of
Speaker 1: the whole process where they have to show, here's what we could do. A lot of the time it's not quite as sinister as the information that you were storing on minors, But, it then raises the another question we talk about a lot, which is, like, be really, really thoughtful about the amount of information you just maintain ever at when you're that level of, an institution. It's like, how much info on kids do you really need to be keeping?
Speaker 2: I think it was from a panel of 14 schools. Highly confidential is what they're saying. So I assume it's I assume it's childhood behavioral information, and and the news reports and stuff clearly stipulate that it was about at risk children. So it's probably kids
Speaker 1: Woof.
Speaker 2: Yeah. Exactly. Kids experiencing risk issues. So it's definitely not good. Highly sensitive, highly confidential information about at risk youth. So not not not chill.
Speaker 1: Not not chill and good. Vice Society has carved out a little bit of a reputation in this space apparently. Largely going after health care and educational organizations.
Speaker 2: Woo hoo.
Speaker 1: Agnostic to kind of where they've done it in Europe, they've done it done it in North America. As targets go, again, we've talked about this before. It's like I I get why folks would go after healthcare. It's very high stakes. There's a huge incentive for people to pay. There's relatively, relatively large coffers, but it contains such an obvious ethical dimension that it's it's pretty pretty pretty not cool. When you've drifted over into educational organizations, you've kind of, I've lost the thread on what it is you're looking for here because that's not you're not flushed with cash. There isn't that urgency. It just seems like you're kind of really picking the worst things to go after at that point. Be like hospitals and schools. It's like, wow. Okay.
Speaker 2: So ransomware.live, a a site that tracks ransomware attacks, has Vice Society is the number one education attacker Oh, wow. In the field. So they were like and by, like, a five x multiple. So they're like they're like one of the biggest ones. So the, I'm also seeing I wonder if this is still up, but they they have a website, apparently, and it is based off of the Grand Theft Auto Vice City design.
Speaker 1: Oh my god. Is that the name? Vice Society?
Speaker 2: Vice Society. So it's,
Speaker 1: Wow.
Speaker 2: It's, I don't know. They they are what they are, but
Speaker 1: k. Well, let's let's balance out one or two more things we could chat about here. Let's balance out some of that those bad vibes from that Vice Society story. With the story that actually came out of our Discord user, BiteMantis, who I think also gave us the YouTube encryption story, shared some stuff about Trace Labs. In a previous episode, Scott, we talked about I can't remember what the name of the group was, but they essentially functioned as like a like a Hackers Without Borders type thing.
Speaker 2: Well, it's funny enough. That's what their name is. Hackers Without Borders. Is Literally hackers. I wonder why that was in my brain
Speaker 1: for some reason.
Speaker 2: Nailed the memory recall there, or you just didn't trust yourself?
Speaker 1: Wouldn't be the first time. So Trace Labs is this, like, donor funded nonprofit. They do open source intelligence, but their whole kind of purpose for existing is to help find missing people. Never seen anything quite like this. It's super interesting. And it's basically if you're if you are interested in open source intelligence and you're interested in participating in some kind of big decentralized project on the Internet to help do a little bit of good, it seems like a very, very interesting, initiative. And we just wanted to kinda put it out to people. We don't know a ton about it, but it seems pretty cool. They're not going after bad guys. They're not doing that, like, no, we're gonna find them and take them down. Like, they're just they're just trying to find people that are missing, and then they report it on to the police, and then they move on to the next one.
Speaker 2: Love to see it. It does it does it does. I think this is gonna be something we see more and more, you know, now that I think government well, like, we talked about this in the Christmas chatty chat, but, like, governments governments need to accept that cybersecurity is a thing, and it's here, and it's not going anywhere. And some of the most harmful thing that some of the most harmful behavior and and policy that we're gonna create is stuff that limits knowledge development in it because
Speaker 1: Yeah, right.
Speaker 2: It's kinda like, Yeah, I don't know. But I say that I was about to make a paradigm to gun control there, and I don't know if I want to go there because it's like it's like is is the best way to to fight guns with more guns. And maybe I've just caught myself in a in a moral quandary that I'm not not okay with. But
Speaker 1: And out no. No, Scott. An hour and one minutes into this podcast, I say take on gun control.
Speaker 2: It's like, yeah. Yeah. Yeah. Yeah. We're gonna have cybercrime, so we need to develop better cyber criminals to fight the cyber criminals. Whoops. I I The
Speaker 1: only way to stop a bad guy with a computer is a good guy with a computer.
Speaker 2: Oh, no.
Speaker 1: Oh, no.
Speaker 2: Anyway, I've got a quick shift.
Speaker 1: Quick shift. Pivot. Pivot. Let's go over here. The last thing I wanted to talk about there's not really even a big story here. It's a little outside of scope, but it was wild. Or I think it's incredibly interesting. Researchers at the Graduate School of Frontier Biosciences at Osaka University published a paper called Stable Diffusion to Read fMRI. What they've essentially figured out how to do, for anyone that doesn't know Stable Diffusion, is like MidJourney or DALL E, a commonly used practice for AI image synthesis. It is one of the more kind of open source y ones. There's an official version of it with some controls built into it, but it is open source. So there's people creating new versions and forks of stable diffusion all of the time. It's becoming kind of the workhorse of the AI image synthesis world. Very interesting stuff. These researchers went full mad science and figured out how to wire in images produced by an fMRI. So like essentially a brain scan and they piped it into the stable diffusion process to the point where they were able to show an image to a person who was inside of an MRI, scan that person's brain, take that brain scan, a two d image that was produced, wire it back through a stable diffusion based process and output an image. That image that outputted looked a great deal like the image the human user was shown. So, what they essentially managed to do was reverse engineer what that person was seeing without seeing what they were seeing. Not to be kind of hyperbolic about it, but they kind of read the person's mind in a weird way if you think about it like that.
Speaker 2: This is very dystopian sci fi. I feel like we're knocking on that door. Very sci fi.
Speaker 1: We're this one felt weird to me. Like, it, I think in Prometheus, the Ridley Scott alien prequel, There's a there's a big plot point that involves a technology that is used to scan essentially be able to view people's dreams.
Speaker 2: Oh, this is
Speaker 1: And this this feels kind of like that where it's like, oh, you're you're you're seeing what I'm seeing without seeing it just looking at my brain.
Speaker 2: Maybe
Speaker 1: it's very interesting tech.
Speaker 2: Maybe you maybe our age difference is showing here, but this is minority report, like, nailed. Yeah. Like. Yeah. Like.
Speaker 1: Oh, yeah. Minority reports a good parallel
Speaker 2: to it. But it's like like minority report was based on a story out of, like, 1950s. Like, this is like some some real true scyfall, like total recall. Like, like, I could I don't know. That's that's crazy. Like, imagine imagine a world where again, this is just me synthesizing this into, like, weird talking points and potential dystopian futures that we can turn into future Netflix series. Hit us up. The like, imagine a world where, like, you know, we all wear these biomechanic computers now, Apple watches, Whoops, Garmins, you know, you name it. We have things on us all the time measuring our heart rates and our breathing and respiratory conditions and you name it. Imagine one that was literally constantly monitoring our what we were seeing, recording and and and monitoring what we're perceiving. Like, how far away are we from that dystopian future where it's like when you do something or see something, bang, it shows up in the, like, it shows up in a a global database and is auto analyzed by a computer to be like, oh, this person just committed a crime.
Speaker 1: There was, there was a Black Mirror episode. It's pretty hacky to to to bring a Black Mirror in the concept of a dystopian sci fi premise, but here we are. There was a Black Mirror episode where the the sort of core mechanic of it was that they had the ability to plug a person into a thing that could visualize what they were remembering. Mhmm. The story was following a person essentially trying to solve a crime. They could interview people but they could put this little thing on their head and then watch on a little screen what they were remembering in real time. Like that feels very very analogous to this because it's not just about seeing what you're actually perceiving, it's about seeing what you're imagining in your head as a rendered two d image. For now, it's a single two d frame, but the speed at which we went from, okay, stable diffusion and mid journey can generate one static frame, Now they can generate 24 static frames in a sequence and you have animation. This is you're starting to get a bit of a pipeline for being able to just sort of see what a person is imagining in their head and the, massive implications of that are certainly not unpacking this paper, but are really interesting to think about.
Speaker 2: If they're releasing this paper and talking about this research, that means that it was done probably long enough ago that they are probably actively working on pulling memories out as animated sequences. Like, you've got to imagine it's the same parts of the brain that are creating. Like, if I'm looking at an image, it's or maybe it's not. Maybe memory center is different. It'd be very interesting, but you've got to imagine that that that's something that they're working on. Like, if they have the ability to pull information out of your brain and turn it into a a generated image.
Speaker 1: I think the okay. So Lord Lord knows I don't know anything about this, but I think that the bottleneck would be the speed at which you could capture these fMRI images of the brain. Because that's the input that you're using, to get the final image. So unless you could capture those images at an animatable frame rate, like unless you could capture 12 of those bad boys a second, you're gonna hit a bit of a bottleneck using this technique to be able to create, like, a stream of video of what a person is picturing.
Speaker 2: K. Well, I've got I've got news for you. F m r i's run at a frequency of
Speaker 1: Yeah. What's your frame rate there?
Speaker 2: Oh, my God. I'm finding conflicting data. So I I shouldn't say anything. One says a 100 hertz, which is like a 100 frames a second. Woah. So but the other one says point o one to point two five hertz. So I'm not sure which one is right, but a 100 hertz is definitely enough to pull the video out of.
Speaker 1: Wild stuff.
Speaker 2: Anyway, we that's a
Speaker 1: What a time to be
Speaker 2: alive. What a time to be alive. So lessons learned from this episode, find free public power, set up crypto farms. Your voice is no longer a part of you. It's now property of the Internet, especially when you put it on the Internet like we do.
Speaker 1: We never actually talked about the voice story. I only introduced you to Son of Hacked and never where that came from. Let's let's wrap up there. So Joseph Cox over at Motherboard posts this really cool story and video about him using, some of the newer voice, free speech synthesis programs that are out there to circumvent Lloyd Banks' voice ID biometric service. So basically, he used this service to hack into his own bank account. Great great story. Great video. Everyone should go check it out. Lloyd's Bank's voice biometric security thing uses all of the standard language that a lot of these do. Talks about how your voice is like your fingerprint. We use all these unique characteristics, and it can even recognize when you have a cold or a sore throat or apparently we have learned if you use a deepfake version of yourself. So what Joseph Cox did was he recorded a five minute sample of himself, uploaded it to a product made by a company called Eleven Labs, and it generated this, synthetic version of his own voice. Eleven Labs has been implicated in a couple of, like, kinda not great things that have happened online. Nothing that they did, but people on four chain using to create videos of celebrities saying all kinds of messed up stuff. Really the classics.
Speaker 2: Important Classics. I like that.
Speaker 1: At the time that motherboard all the classics.
Speaker 2: You've seen them on Twitter. You know which ones are talking about. Yeah. Completely.
Speaker 1: I think part of the reason this product was at the time really popular for those things, and this was still true when motherboard published it, though not when I went to use it, was that for a long time, the entry level version of this was free. So basically with a burner email, you could do voice synthesis. And again importantly, it only requires five minutes of speech for this to work. When we did our deep fake episode like a year and a half, two years ago, In order to get that fake voice, I had to upload forty minutes of audio of myself talking. It took, I think, close to an hour for it to, like, just to synthesize and process it into a usable voice. And then the amount of time it took between me typing, hitting enter, and it producing an audio file was on the scale of minutes. The, I wanted to try Levin Labs product out to see how you could use it for something like this. Yep. And it now requires five minutes. It was able to generate voice within less than a minute, almost instantaneously. And then between typing and hitting enter, it's not quite real time, but it's a delay of less than five seconds.
Speaker 2: That's crazy.
Speaker 1: Which is how, Motherboard was able to basically hold a real time conversation with this voice authentication service and break into this bank account. You would need the, the victim's birth date, but with the amount of breaches where that information's gotten out, for a lot of people, that's pretty, that's a pretty low barrier. Lloyds Bank is not the only place that uses something like this. TD Bank up here in Canada uses voiceprint, use that same language. Your voiceprint, like your fingerprint, is unique to you. Chase Bank has voice ID, Wells Fargo has voice verification. I think maybe for a
Speaker 2: little bit maybe we put a
Speaker 1: pause on these. Maybe until we we figure out this speech synthesis thing, maybe we don't consider voice biometrics as secure for a little bit. Let's just let's just pause that as as a way of locking down an account. I'm tossing that out into the world.
Speaker 2: Especially for people like you and I, Jordan, who have hour hours and hours and hours of our voices on the Internet. Mhmm. Maybe maybe I am not a TV account holder, so let's be thankful for that. Because I'm sure Yeah. It wouldn't take much to generate my voice.
Speaker 1: No. I did it by accident. Like, it wasn't your voice. It was weird.
Speaker 2: But you could've made my voice.
Speaker 1: Of both of our voices, but, like, I could've. I actually I thought that. I was, like, I I got I was, like, no. It feels like if we're gonna start developing cultural rules and mores around this, don't deep fake someone without their consent probably seems like the very first one. So don't accidentally do that for your podcast, Jordan. I thought
Speaker 2: to myself. As I created my new twenty twenty three morality.
Speaker 1: Kinda. Like, try trying to write some rules at a dead sprint here. But I I did feel like I could cross the line of of, you know, playing around with the the both voices version and Yeah.
Speaker 2: Making it a child a child of us. It's cute.
Speaker 1: A child voice, Son of Hacked. Yeah. No. It's good fun. It'll it'll it'll inherit this show from us at some point in the future. We'll have Chat GPT write it. We'll have Son of Hacked say it out loud. We'll get MusicLM scoring it with all the music we've already written for the show. And this thing will just be on rails, baby.
Speaker 2: Let's go.
Speaker 1: Let's go. Okay. I think that's it. YouTube is infinite storage, crawl space, crypto mines. Robots can read
Speaker 2: your brain. Space, crypto mines on Netflix 2024.
Speaker 1: Get at us. Chick fil A sauce.
Speaker 2: Get at
Speaker 1: Thank you for listening.
Speaker 2: Okay. Thank you. We appreciate your time. We hope this wasn't too chaotic. And, we look forward to seeing you in the next episode.
Speaker 1: Catch you in the next one. Ciao.
Speaker 7: The biggest tournament in soccer is finally here, and I've already started planning my watch parties. My go to move before kickoff is stopping at Total Wine and More to grab drinks for the whole crew. Wine, beer, seltzers, maybe a few ready to drink options. Everything we need for a full day of matches. With this many games, it definitely helps knowing you're getting the lowest prices. Total Wine makes it so easy because I can grab everything I need in one stop. Get match day ready with Total Wine and More today so you're set from kickoff to the final whistle. Spirits are not sold in Virginia and North Carolina. Drink responsibly. Must be 21.
Speaker 8: Shipping, billing, admin, payroll, marketing. You're managing all the things, so why waste time sending important documents the old fashioned way? Mail and ship when you want, how you want with stamps.com. Print postage on demand twenty four seven and schedule pickups from your office or home. Save up to 90% with automated rate shopping. That's why over 1,000,000 small businesses trust stamps.com. Go to stamps.com and use code podcast to try stampsstamps.com risk free for sixty days.