Ben Whitelaw 0:00

Mike, you're back in the chair in Ctrl-Alt-Speech. Great to have you, back with us. I, I imagine that you spent most of your time, if not all of, the weeks you've been away, probably fiddling around on the internet, having some fun. Uh, isn't that what you do when you're, when you're not here?

Mike Masnick 0:19

No, no. In fact, I was, I was, practically off the internet for good portions of time. And even when I was on the internet, I had a connection that topped out at some ridiculous... I had a screenshot of the internet access that I had, which barely worked. I was off in the wilderness, and it was, quite enjoyable actually. So I did not spend much time on the internet

Ben Whitelaw 0:40

well, I, I hope in the time you've been back at least you found something that is fun, that you can share with us today on Ctrl-Alt-Speech. What, what have you come across?

Mike Masnick 0:49

I did, I did. I found something called Roost, and I, I need to clarify because a lot of people listening to this podcast will know of Roost as the open source trust and safety, project, which we've talked about on the podcast and we love and we're supportive of. This is not that. this is this sort of new messaging app which is, part of what people are referring to as slocial media as opposed to social media.

Ben Whitelaw 1:15

it

Mike Masnick 1:15

And, Roost involves sen- It's a messaging app except the messages travel at the speed of bird. Uh, so, so depending on how far away you are from someone, if I wanted to send a message to you, you and I are on basically opposite sides of the world, it would take some time because it would take however long it takes a bird to fly from me to you, and so we would have to communicate slowly

Ben Whitelaw 1:40

Mm-hmm. So, and it, and it, and it replicates the speed of the particular bird in question. So depending on what bird you

Mike Masnick 1:46

Yes, you pick a bird. There are a few things that are a little sketchy about it where you can do things to speed up your bird and, train your birds, and I, I... Who knows what all that is? But just this concept of, like, slowing things down a little bit, not everything is instant, you know, in the old days when people would send letters back and forth, it would take, you know, weeks. And so having that element returned to messaging is, kind of a fun idea.

Ben Whitelaw 2:11

Yeah, love that. That's really fun. I mean, it's not dissimilar, kind of hearkening back to a different time, I would say. my fun thing that I saw on the internet this week was, imagine seeing, Mike, classic novels played out in map form,

Mike Masnick 2:26

Ooh.

Ben Whitelaw 2:27

So imagine being able to kind of track characters across the world, and essentially kind of tell the story of novels that you've always wanted to read in the form of maps. And this is what, a friend of mine a journalist called Chris Moran has done. He's vibe coded with his daughter, who is studying, some literature, classics. they together have created this tool called Plotlines, and you can select a book, and you can, essentially kind of watch the characters disperse across the globe.

Mike Masnick 2:59

That is amazing.

Ben Whitelaw 3:00

It's really cool. Yeah

Mike Masnick 3:02

I, I just, just for listeners, like you did not tell me what you... The, we did, I did not know what, what you were going to tell me about. This is, that's am- amazing.

Ben Whitelaw 3:10

Yeah. It's really cool. So, so one of the books is Jules Verne's Around the World in 80 Days. And so you can see, you know, Phileas Fogg go from one house in London to another, and then it kind of zooms out. You can see him go, across the Mediterranean, kind of southwards. there's some great admittedly AI-generated captions, but that build on, uh, the, the kind of full story. It's a great learning tool. and so, you know, we're gonna talk today about some of the downsides of, AI, but that's another great example of the creativity they can foster.

Mike Masnick 3:41

Yeah. I mean, as, someone who is obsessed by maps and could stare at maps all day, like, and reads lots of books, I'm this is great. I haven't, I haven't seen this yet. I can't wait to look at the show notes and, and find out what, what this is. cool

Ben Whitelaw 3:57

it's really cool. Really fun. Hello, and welcome to Ctrl Alt Speech, the podcast where we make sense of the major debates shaping online speech, platform power, content moderation, and the future of the internet. It's July the 9th, 2026, and this week we're talking about the eternal problem of ad trust and safety, and for our Patreon subscribers, who decides whether you get access to those AI models? disclaimer, it's Reid Hoffman. Um, Ctrl Alt Speech is the fraud-free weekly analysis of Mike Masnick, founder and editor of Techdirt, and me, Ben Whitelaw, the founder and editor of Everything in Moderation. yeah, Mike, I was gonna say Ctrl Alt Speech is scam-less, but I actually don't wanna talk for our Patreon supporters who, who may feel scammed in some way. I don't know. I don't know. I don't wanna, I don't wanna kind of... I th- I feel like we deliver a service.

Mike Masnick 4:56

we, do we need to hire trust and safety people to manage our, our, our Patreon subscriber base and their,

Ben Whitelaw 5:03

I, I just don't want them to think we've kind of socially engineered them to pay $10, a month, say, to listen to extended episodes. So I, I held back from scamless, but I can say we are fraud-free,

Mike Masnick 5:13

Okay.

Ben Whitelaw 5:14

Um, but yeah, great to see you anyway. it's nice to have you back in the chair. how was your break? you know, it sounds like you disconnected from the world in a really nice way.

Mike Masnick 5:23

Yeah. It was, it, it was just one of those, like, I haven't done extended travel in a very long time and, basically taking three weeks away and, doing a fair amount of actual disconnection and being out. I was in Alaska Which I'd never been to before, and seeing the amazing wilderness and all sorts of animals. I mean, I saw whales and bears and moose and, caribou and all... Eagles, so many bald eagles everywhere. it was really, You know, people say it's one of those amazing trips, and I had not done it, and was able to spend, a few weeks just really enjoying it. And so I had a great trip.

Ben Whitelaw 6:00

sounds great. Where w- where were you spending the, your Fair Nations 250th birthday anniversary?

Mike Masnick 6:06

the funny thing was, so we, we ended the trip by going into Canada. and so we landed in Canada on July 1st, which is Canada Day. Um, and so got to celebrate that, which was really fun. and then we came back into the US on July 4th. So we spent Canada Day in Canada, and July 4th we flew from Canada back to the US. and we were absolutely exhausted but... 'Cause our flight was at, like, 5:00 in the morning. It was crazy. so I was back in the US but mostly just recovering from three weeks away, so

Ben Whitelaw 6:40

it. Got it. talking of scams, I wondered if you'd caught up with the World Cup drama in which your

Mike Masnick 6:47

yes.

Ben Whitelaw 6:48

intervened in, the game of soccer in a, in a way that no one had time for. what's going on there? Do you-

Mike Masnick 6:57

I, I cannot support any... I mean, even as I was away, we were, I've gotten pretty into the World Cup, so, even as someone who is not particularly knowledgeable about football. Uh, and, and so yes, I did follow that whole thing, and, once again, for not the first time, somewhat ashamed of our president. But, you know, it does feel like there has been a plenty of controversy over FIFA for many years for being somewhat corrupt, and it is FIFA that gave this same president this, whatever, the FIFA World Peace Prize or whatever it w- it was at some point. And so it just feels, it feels fitting in some way, a- as did the US then losing the game in which, Trump interfered to bring back the US's, in theory, best player. but yeah, it's, uh, it's been quite a, quite a, quite a world.

Ben Whitelaw 7:50

Yeah, I've been-- I'm impressed you have been following. Um, I'll keep my, um, my views to myself since they're not really broadcastable, but yeah, yeah.

Mike Masnick 8:00

I, I will say as we're talking, England is still in the tournament and the US is not, so

Ben Whitelaw 8:05

Correct. Correct. there'd be all-out war if that was not the case. So yeah, it feels like, it feels like that's the, the right state of things. Um, I will be coming over to the US, in a couple of weeks' time. We, we haven't really talked about this, but we're building up to the live episode of Ctrl Alt Speech at Trust Con, which is the major, major conference for trust and safety professionals that's happening in a couple of weeks' time. the full episode, in two weeks will be that live recording, so you can listen in, you can hear us, take the stage with two brilliant guests, who we'll, we'll talk a bit more about in the coming weeks. it's also worth noting that you can come and find us if you are going to Trust Con, and you'll receive a special gift from us. So if you manage to track us down in the large venue, the large kind of quite faceless venue that is the Trust Con, uh, conference center, you will receive a special gift relating to Ctrl Alt

Mike Masnick 8:57

Yes. find Ben and or I and, we have gifts for Ctrl-Alt-Speech listeners. So that is, uh, an incentive.

Ben Whitelaw 9:06

say more than that. and yeah, particularly for our Patreon subscribers, who make everything we do possible. we've got our supporters tier who pay $10 a month, the gentlemen's $10 a month to get extended episodes and early access to the podcast. And then we've got our special insider and founder tiers who get to suggest stories that we cover on the podcast and, get a, a direct line to Mike and I to talk about the things that they want to hear every single week. So go to patreon.com forward slash ctrl-altspeech. The link is in the show notes with my, very excellent vibe coded plot lines tool. We've got loads to talk about, Mike, so we will wade straight into this week's stories. We, we don't necessarily set out to bash Meta here on Ctrl Alt Speech, but they do kind of bring it on themselves. and I think that's, certainly the case when it comes to ad fraud, which is a story I want to focus on this week our listeners will know fairly well that last year there was a big story that Jeff Horwitz at Reuters published about Meta's significant ad fraud problem. he did a big investigation and, found that around 10% of Meta's revenue for 2024 was based on fraudulent ads or, or scam ads. And when you think about how much revenue that consists of for-- it's a very large advertising machine, it's a, it's a huge, huge thing. The analysis and the discussion after that was whether Meta was really doing enough to combat ad fraud on the platform. was it kind of incentivized to, deploy the tools that it used, for organic content in the same way for its ad business? And, a really interesting interview that came out this week with one of their VPs in EMEA actually adds a bit more context to that. so let's talk a bit about that, and then I'll get your thoughts, Mike, on, whether ads trust and safety is still, a massive issue for us. the interview is kind of interesting 'cause it's fairly wide-ranging. It's, it's with a, vice president called Derya Matras, who's an, European-based, vice president, and she's in Cannes when she gives this interview, so there's a few things that have happened since then. She talks about smart glasses, which are obviously fairly controversial. She talks about the UK social media ban. She talks little bit about the California product design appeal, which we're kind of expecting to happen, some point this year. But it's really the stuff about ad fraud that I found really interesting. And she says a couple things that stood out. First of all, she said, and I quote, "It's not like we knowingly have bad ads," which is, is a phrase I, take a lot of umbrage with. and then goes on to say that as a company, Meta s- has spent north of $30 billion fighting fraud, including building AI systems to catch fraudulent ads and scams. To which my, reaction was, well, you're not trying quite hard enough. Um, and, and there is this kind of question as to whether Meta is, looking the other way or leaving ads up as part of, its revenue goals and, and- Those quarter-on-quarter results are, increasingly impressive. the piece that we'll include in the show notes talks about how just in Q1 of, 2026 alone, its revenues grew 33% year on year to $56 billion. So it has spent historically a half of what it earned in revenue in Q1 alone, just to kind of make the maths clear. and this all comes, Mike, after, literally three days after a BBC investigation found that there were paid ads on Instagram that used phrases such as child video and rape video, that were directing people to Telegram groups that included, content, multimedia that included, horrific rape videos and, and CSAM. And so this story, this BBC investigation plus this interview kind of raised this question again of, why doesn't Meta fix its ads business? why isn't more being done from a trust and safety perspective, to ensure that people aren't using ads as kind of vectors of harm? So I know that, you know, your whole shtick, Mike, is that it's not as easy as, as you think it is. Have I, have I oversimplified this? Um, what's your thoughts?

Mike Masnick 13:19

Yeah, it's not as easy as everyone thinks it is. All right, I'm done. You can, you can just... No, uh, it-- Ugh, God. So this, it's, this is always, these are tricky stories and nuanced stories with lots of different factors. I think it can both be true that there is more that Meta can do, that the incentive structures that Meta is facing sets it up in a way that they prioritize profits over safety, to a degree that, that is not great. and some of that is, the leadership of Meta's fault. I think that, they are very focused on profit. Some of it is the nature of, capitalism and, like, Wall Street that demands that Meta continue to make number go up. and therefore, they have incentives to do that, and they are weighing those incentives against the costs of bad advertisements. At the same time, I do think there is something to the fact that advertising trust and safety is incredibly difficult It certainly doesn't get as much attention as, you know, regular social media trust and safety. But in some ways, I think the ad safety business is, is even more tricky because the financial incentives there mean that there's more incentive for, malicious actors, right? there are incentives in the social space for malicious actors to try and, you know, push messaging or misinformation. Everyone's focus is on, false information and stuff like that. But in the ad space, there's sort of commercial incentives, and that changes the bargain that people are dealing with and makes it trickier. And at the same time, the fact that, this form of advertising, which is effectively different than, advertising that we had in the past, is so easily accessible to anyone, is really powerful, especially for small businesses. The fact that a small business can set up shop and advertise and reach a well-targeted audience at minimal cost, th- and that you can experiment, that you could, you know, spend $100 on an advertisement and see if it reaches the audience that you're trying to reach, is incredibly powerful, and that requires some level of automation and acceptance, which makes the trust and safety side of it even more difficult when you have incentives for bad actors to try and game the systems as well. And so it feels tough, but I think it's important to hold all of those ideas in your head at once, that yes, I feel like we should expect Meta to do better, and I think it's totally fair to call them out, and the BBC report is really damning, especially the part where they, like, report these ads that are clearly, clearly horrific and just like any human being who looks at that ad should recognize that it is horrific and being told, like, "This does not violate our policies," that feels like a real problem. At the same time, recognizing the sheer scale of this and how many ads they have to review, and they're using automated systems because they have to, otherwise, it is impossible to handle this. and so You know, I don't wanna be-- I'm not being a Meta apologist and saying, like, as this VP kinda says, like, you know, Look, there is some bad stuff, and we don't want bad stuff on the platform." I do think that, in some cases, Meta certainly would prefer to get rid of the really, really bad stuff, but they're balancing that with the demands of Wall Street, having this system that allows anyone to advertise to anyone that they want to, which is a, a powerful and useful tool, and then how do we go through all this? And so, I think it is a more complex story than, people wanna

Ben Whitelaw 17:01

Yeah. I mean, I, I, I'm definitely not an ad, digital ads expert. and I, I, I want to speak to people actually who work in the trust and safety space for ads, because I imagine it's a different kind of, problem that they're dealing with here. and I guess that's, that's the question I have is, like, to what extent are ads different to organic content, the posts that your average person will share, and who kind of-- who are at the whims of, being algorithmically distributed on platforms? Because my sense is and my feeling is that ads are different. You know, ads have a bunch of data applied to them when you upload an ad that you don't get when you're posting about your cat, or, you know, the fact that England-- you're happy that England won the game against Mexico. Um, you know, you have-- First of all, you have the kind of like payment details, that's kind of baked into the system. that's why it's automated, that's why it's quick, that's why Meta's able to scale those revenues as it has. You have targeting, You have very specific targeting, of particular groups, and Meta in the past has got in trouble for allowing specific demographics or religions to be targeted in ways that, don't yield good outcomes. And then you've got this kind of, The fact that there is this inbuilt distribution to ads makes me kind of... I had this realization that's like, why are we not taking ads trust and safety really, really seriously? Why is it not the kind of like pinnacle of trust and safety rather than this what feels like a sidelined, wing of the trust and safety profession, which I imagine sits in a, less so in a kind of policy space, but more in a kind of business operations part of these big platforms. And who, probably, and I need to think, you know, validate this. I'm interested in speaking to people who might know more, know, you're not seen to be quite a, speech advocate in quite the same way because it is ads. and I, I wonder if that needs to be turned on its head, that idea.

Mike Masnick 18:54

Yeah, I mean, I'm not, I'm not sure if your underlying assumption is totally true. I mean, I, I do feel that if you talk to some of the sort of really early trust and safety people, like, I think a lot of the early trust and safety really was kind of advertising trust and safety. I mean, you know, I have a, good friend who I know not through industry, but just you know, someone I've known socially for a long time, who, in the 1990s was doing what was effectively advertising marketplace trust and safety work. And I remember, talking to him in '98, '99 timeframe about, how much time he spent basically just trying to track down scam e-commerce providers who were trying to advertise on the platform that he worked for. and he was just saying, like, what an, an incredible challenge it was. So I, do think that, like, that's always been a part of the trust and safety space. It just gets less media and political attention. and the fact that, it is the revenue source for a lot of companies advertising, you know, trying to stomp out ad fraud, I think it, it has, for almost every company doing digital advertising, it certainly is a big part of the process. It just gets a lot less attention, and you do have these competing incentives of the companies themselves wanna make money and, you run into the same thing where it's like there are always trade-offs, right? if you carefully vet every single ad, and it's not just the ad, right? Because the ads can look innocuous but point you to a scam or, you know, there's a whole bunch of different layers of this. And it gets back to the same thing with- within the social media trust and safety space. You, talk about how much context does someone who is reviewing these things have. Same thing in the ad space where it's like, you know, who's actually advertising? What is it that they're advertising? Who are the companies behind it? Is it fraudulent? And, again, like, scammers that are buying ads, have their reasons for doing so. And have incentives to try and trick the platforms. and so it is a much trickier space. I think you're right to the extent that it's a space that doesn't get as much attention externally, but I'm not sure it's true that internally it doesn't get a lot of attention. But I do think the underlying factor, and this is, where we kind of started this discussion, is that from the corporate perspective, too often certain companies will sort of, accept a certain level of, like, scam ads that will harm users because of the revenue that it generates. I... It's not as simple as, oh, Meta just allows all this because it makes them money. there is a, a calculation between, like, how much do we put towards stopping these versus how much revenue are they bringing in? there are trade-offs there, and there is a, you know... You can stop all scam ads by stopping all ads, right? Like, that's the extreme, right? Or, like, personally vetting each and every ad, but then you don't have a business anymore.

Ben Whitelaw 21:52

Yeah. And I, I think you're right about, where platforms draw the line. There's a, phrase in the interview that kind of stood out where this VP says, you know, "When something is benign, we allow it on the platform." which is, which feels kind of very subjective. Like it's not, whether it's, policy violating or not, whether it kind of adheres to some strict guidelines. It's whether it's benign, whether it's going to, do something, you know, negative in the real world, it feels like. so I think, I think there's, That's the area where I think it obviously I'm very interested in. I, I think the other thing, Mike, is that I'm not-- I think there's a middle ground here. You know, you don't want ads to be very difficult to create and build, but, uh, you do want the platforms to kind of recognize that they are important. And when... Like last week on the podcast, I talked about how Meta was rolling out AI moderation more broadly across the platform, but also for user appeals, which are often one of the hardest elements of, a trust and safety team's remit to, to pass through. They're very often, more human-led than slightly more s-simple to find harms like spam and, and, and so y- in one case, you've got AI moderation being rolled out at great speed. and at the other hand, in this, week's news, you've got, the same company claiming that they, they've done all that they possibly can, including using AI moderation to address some of those issues. Now that doesn't f-feel quite aligned to me. if the AI moderation is suitable for user appeals, then I imagine it can play some sort of role, maybe not reducing fraud to zero, but maybe reducing it by a couple of billion pounds a year or, you know, from 10% to maybe 8%. so there's that. There's also the fact that, you know, at Cannes, Mark Zuckerberg spoke about making it much easier for advertisers to create many, many new forms of adverts, different, creatives and more copy. You can have multiple variations and that would be all automated and working in the background. Again, if this was a difficult, as difficult an area to kind of a-address from a safety perspective, would the company be investing so much money in, expanding the tools that those advertisers have? I don't know. It feels like, they aren't giving credence to the idea that this is a, an important area of their business and

Mike Masnick 24:05

yeah, I don't know if I'd agree with that. I, I, I don't... I think, you know, expanding the tools and, allowing people to experiment with different advertising is sort of a recognition of, where the market is and, and what people want and what advertisers want. and they want more power over how their messages are shared, and how they can get out there, and how they can reach their audience, and I think that is fair. I don't think that that means that they're not taking... So let me be clear on this. I don't think that by itself means they're not taking user safety around ads seriously. It might still be true that they're not taking it as a priority as much as they should, and I think we're seeing that in these other things. But I don't think those things are directly connected. I think that, you know, expanding what kind of ads you can have and also expanding, the AI tools and, and moderation tools, I can see a world in which all of that fits together and makes sense, and then you can still argue that the company should be doing more to protect users from, scam

Ben Whitelaw 25:04

Yeah. what I mean, I guess, is that if you're, if you're a company that's failing to spot scam ads that are presumably fairly manual to upload right now, where the creative and the copy are uploaded and you target them, as, as the BBC investigation has, uncovered, why would you make it even easier for, malicious actors and bad actors to create

Mike Masnick 25:23

because, I mean, yeah, but all, all of... That's, that's the thing, right? Like, so many of these tools can be used for good and for bad, right? Like, is the problem. This is the problem of so much of technology and the internet, is that these are dual use, right? Like, they allow and enable all sorts of really good things, and, like, if small businesses can reach an audience in a way that they couldn't before, in a way that is effective, that allows those businesses to operate and to reach customers, that's actually really valuable, and I, I think underestimating that you know, isn't great either. The fact that some percentage, and, and there's arguments over what that percentage is, can also use that for scams is a problem that needs to be dealt with. But, I don't think that there's a contradiction in saying, like, we wanna provide better tools for the 90, 95% of our advertisers that are trying to do well. I don't think that is necessarily a bad thing. You want them to provide better tools. You would hope that that would also come with additional support for trying to stop the bad actors as well.

Ben Whitelaw 26:23

Yeah. maybe that's a compromise. interesting pair of stories there. Our eye-catching section, Mike, formerly known as our, our shorter, quicker stories, um, also, lands a punch or two against Meta. do you wanna talk us through the story you saw in Wired this week?

Mike Masnick 26:39

Yeah, so this is, uh, another one sort of, of Meta acting badly, where they've launched this new feature in Instagram where they're, you know, they're really ramping up their AI stuff and trying to push AI and everything. But you can now use AI, and if you, point to a particular user's account, you can use their images and create AI out of their images, and it's entirely opt-out. So Ben, if you're a big Instagram user and I wanna take your photos and turn them into AI slop something, I can do that without your permission, without notifying you or anything. and so basically everybody in Instagram who doesn't, go in, follow multi-steps to opt out, is basically making all of, their regular imagery available for AI modification.

Ben Whitelaw 27:28

This is wild

Mike Masnick 27:29

just feels incredibly tone-deaf. It's just like, y- like, So the one slight defense I will make is that, look, if you're posting stuff on the public internet, people can take that and do stuff with it. That is always true. That has been true before this. There is no way to say, like, "I have posted this thing on the internet, and you cannot look at it. You cannot do things to modify it." I mean, there's some sort of copyright elements, but for the most part, it is out there and is available. But Meta stepping in and saying, "Now automatically built into our system, you can modify other people's images without their permission, without notifying them," and if they don't like it, their only option is to go through this, deep down in the menus and opt out just feels incredibly tone-deaf. It really feels like, just giving the finger to, to its users and saying, "You don't like AI? Too bad." you know, I understand maybe sort of why they're doing it, but, like, at this moment, when the public feeling towards AI and AI use in this form, in social, the sort of AI slopification of the world, to push that out in this manner just feels totally tone-deaf.

Ben Whitelaw 28:41

Yeah. And, and the idea that I, with my large, bank of ins- Instagram content, let's say, wouldn't even be notified that somebody has remixed my content, that's the craziest thing. I mean, I'm with you. Like, I don't think this is a feature that really anybody wanted necessarily. I think this is, using AI for AI's sake, frankly. But if you are gonna go down this route, and this applies to the kind of ad story as well, like if you are going to remove all friction from, taking somebody's image and remixing it or, publishing an ad and targeting it at a group of people, if you're gonna remove all friction from that process, at least create the mitigations of,

Mike Masnick 29:20

Yeah

Ben Whitelaw 29:21

of the harm that comes from removing that friction. that's the, craziness of, both of these stories, I think. and, You're right. It's, tone deaf.

Mike Masnick 29:29

I mean, there's another element to this, which is, you know, this is kind of what Grok was doing on X, where you could point to an image and say, "Grok, do this," right? And there was a whole big controversy about that. You would think to some extent that, Meta would have paid attention to the controversy, but it feels like, and again, this is speculating, and we don't know, we're outside companies, but it feels like the reaction was Meta sort of saw X go through this, and the hatred and everything that, went along with that, and said, "Oh, they survived that. They were fine. They got past the controversy," which in that case was specifically around, like, creating what may have been CSAM, but was at least incredibly, demoralizing and, and undressing people and all of these kinds of things, and said, like, we can survive the negative backlash on that to get this tool out in the world." and you know, frankly, Meta is way behind everybody else in terms of the AI race, and maybe it's... maybe some of this is like, "Let's have the controversy because it'll get more attention to our AI." I don't know. But it, it just feels like a really, really tone-deaf decision. Or in part, like, there is this sense among some in the AI space in Silicon Valley of because the hatred of AI has gotten so extreme and so outlandish, and I think, you know, the backlash to AI has gone too far in a lot of ways as well, that there are people within the, the tech world who are just like, we're not even gonna try to mitigate anything because people are so mad about everything that, our best bet is just, like, put a foot on the gas and go forward as fast as we can, and we're just gonna build, and if you don't like it, too bad." and so there is this feeling of like, "We get some criticism, and therefore we're just gonna ignore all criticism," and that doesn't seem very healthy either.

Ben Whitelaw 31:19

No, what a way to approach the world. I will say that it's not available, in the UK, this tool, but I look forward to when it is rolled out, remixing Mark Zuckerberg's content, 'cause I'm sure he won't turn off this function 'cause, uh, he, he loves AI. Um, so

Mike Masnick 31:33

say when when OpenAI had Sora 2, which they'd only had for a little while, and that allowed a similar sort of feature, Sam Altman did allow people to remix him, and he lived with it, and people, you know, made all these videos of, like, Sam Altman stealing GPUs and stuff like that. So, uh, who knows?

Ben Whitelaw 31:53

Let's see who's got the thicker skin. another story that I wanna talk about in this kind of eye-catching section, Mike, is a new Bloomberg profile that is definitely worth, reading. this is a piece on Kaylee Glen Mills, who is the woman, the 20-year-old woman who is at the center of the California social media addiction trials, which we talked about on the podcast, in the last couple of months. she's given an, an interview to Bloomberg talking about the process of being selected as the bellwether case, among all of the cases, that were being brought against Meta and, and Google, and having to kind of prepare herself for the trial, in which she took the stand and, testified against, Meta and Google, and talked about her, addiction, to social media. And also the kind of the real world scenario that she found herself in with, a mother that she didn't really get on with and a sister who was self-harming and, quite difficult circumstances that in part, I think, led her to use social media to the degree she did. It's a quite a sad read in some ways because she says that she hasn't necessarily kicked the addiction. She still uses it. I guess that's what addiction is. You can't, you can't remove yourself from it. But, even now she's still using, Instagram and, YouTube many, many hours a day. So again, Kaylee Glen was, only known as KGM versus Meta et al when the trial took place, but now we know a bit more about her. and, and it kind of just goes to show, I guess, why there is this clamor for protections against children, and I suppose why s- the social media ban conversation has become so strong. did you kind of have any reactions to that, Mike?

Mike Masnick 33:29

Yeah, I mean, I-- it's an interesting story. It's always interesting to find out, you know, these kinds of profiles. I found it a little bit frustrating. you know, I've talked in great detail about the problems I thought with the case and why I think it was a bad decision and why I think Meta is, making the right decision to appeal it. But it is a tough case, and, and obviously, this is a, person who's had a lot of troubles in life, and you never wanna sort of, speak ill of, of them. But it-- the fact that, you know, the piece itself is framed as she still spends her days kinda sitting on the couch scrolling Instagram or whatever, there's a part that is like you can read that as, "Oh, she's still addicted," and there's a part that is just like at some point some-- you gotta take responsibility for your own actions a little bit. And I feel that in, playing up this idea that it was entirely Meta's fault, that excuses her from the responsibility aspect of it. And so I found some of that part of the story to be a little bit frustrating, and, kinda difficult to square with the, with the rest of the story.

Ben Whitelaw 34:33

Yeah, definitely one to have a, a read of, listeners, and to make up your own mind. I think there's, a lot of really interesting detail in there. so yeah, we've, we've kicked Meta I think enough for one episode, Mike. I think that's, I think that's probably fair. you know, if you're a Meta employee or somebody who really loves to, to scroll on Instagram, that's probably a bit of a tough read, but I think it's important that we, cover the big companies in this way. We're gonna change tack in our extended episode, Mike, for our Patreon supporters, and we're gonna discuss on Ctrl Alt Speech Plus, this idea that AI models are becoming kind of increasingly, bound by national, security issues and also by this clamor for tokens, which you saw in a couple of different stories this week. Why should listeners tune in to the extended episode today?

Mike Masnick 35:16

Yeah, I, I think that there's, there's a really big and important change that is happening on the internet as it relates to access to AI, and I think that it is a different kind of fight and a different kind of battle than some of the other ones that we've had with the internet and how people access different parts of it in the past, and I think it is going to become increasingly important, and I think a lot of people don't quite recognize it yet. So I felt that it would be a good thing to, pull in. Uh, it's, it's really about four or five different stories that all sort of hint at the early challenges and trade-offs that we're, we're gonna see on the internet, as AI becomes an increasingly important part of it

Ben Whitelaw 35:52

Super. so that brings us to the end of our free episode this week. if you like the sound of that extended segment, become a Patreon supporter, We'll, jump into that conversation now

Announcer 36:04

That's the end of this week's free edition of Ctrl Alt Speech. To hear the extended episode with additional coverage, become a Patreon supporter at patreon.com/ctrlaltspeech. That's C-T-R-L Alt Speech. If your organization is interested in sponsoring the podcast, contact us via our website at ctrlaltspeech.com. Thanks for listening