Identity Slop: The Rise of Synthetic Media Experts

Jessie Chambers is a travel expert quoted across British newspapers. She gives advice on everything from cruise ships to long-haul flights. There's just one problem. Jessie Chambers doesn't exist. Welcome to The Machine Dialogues, conversations about AI systems and the self, exploring what happens to identity in an increasingly synthetic world as we evolve into machineable and unmachineable selves. Today's guest is technology journalist Rob Moore, a prominent British tech journalist. Rob has spent years reporting on the evolution of the internet and emerging technologies. He's written for titles including The Daily Telegraph, Daily Mail, and Metro, and previously served as science editor at the Mail Online. He also launched the podcast network at Associated Newspapers, overseeing shows across news, politics, and culture. Recently, Rob stumbled onto something unusual. A series of media experts appearing in newspapers across Britain— travel specialists, palace insiders, lifestyle commentators— turned out to be synthetic biographies. These were experts manufactured to give quotes to journalists, quietly inserted into the news ecosystem. In this episode, we talk about the rise of fake experts, how synthetic authority is slipping into mainstream media, and what happens when even expertise itself becomes machine-generated. Because if expertise can be generated, identity can be fabricated, and then authority can be automated. That's what we're exploring this week on Machine Dialogues. Rob Warr, welcome. So Machine Dialogues, you are the first, and I really wanted to talk to you because obviously the subject of Machine Dialogues is around the The Machinable and the Unmachinable Parts of the Self. And you in particular have been doing this groundbreaking work on what I would call identity slop.

There's a lot of it about.

Yeah, I wonder, especially for the listeners, if you could take us back to the beginning and tell us what you uncovered when you were looking at these experts that were appearing in really quite, you know, quality press outlets with their varied opinions Sure.

Well, I mean, it was a weird one because the first time it happened, it happened a year before the piece, first piece I wrote for Press Gazette. I was doing some freelance work for a marketing agency and we got approached by somebody to do a guest blog. And I looked at the person and I thought, this is really strange. This person doesn't exist. And the other people on the site were just going, no, what do you mean? Why is the— and I said, Really, take it from me, this person doesn't exist. I forgot all about it. And then a year later I was using ResponseSource, which is a response service for journalists where if you need somebody to talk about something, you send a request for a specific kind of expertise. And I needed a psychiatrist to talk about the psychological impact of identity fraud. And I sent out a request and I got a response from this psychiatrist. An Oxford-educated psychiatrist no less, who called herself Barbara Santini. And the thing was that the copy seemed a little strange. It had things that I think we're all very familiar with now as a result of ChatGPT, weird sort of capitalized subheads inserted into the copy for no reason and stuff, which often are quite hallmarks of LLM writing. And I was like, this is a bit odd. But really the reason I sort of looked into it was because I was writing for The Telegraph and she worked for a sex shop and I thought, I can't put a sex shop in the Telegraph, you know, some old Tory colonel will have a heart attack into his cornflakes. And so I sort of thought, oh well, I'll just look up what her other jobs are and I'll just list something else. And then when I looked, I realized that she wasn't real. There was just no other jobs. The only thing that she existed and did was speaking to newspapers. But she'd been hugely successful. She, in fact, she'd been in newspapers 120 times and that included all the big sort popular ones, you know, Mail, Express, Mirror, Sun, dozens of times. And she'd even penetrated into the BBC, The Guardian, international press like the New York Post. And when we confronted the organization behind her, we got a series of extremely convincing-sounding legal threats because I won't say his name, but the guy who's actually behind this is, he's a qualified solicitor. So I was like, the legal threats were—

Is he though? They were.

They were highly convincing legal threats 'cause you get legal threats a lot of the time, but then occasionally you get ones where you're like, okay, this is serious. And anyway, we published the piece and the legal threat said that they would sue immediately if we said that this person didn't exist. We said it and then the legal suit didn't happen. And over the coming months, I began to realize that she was far from a low There are a huge number of fake experts. Last time we counted up, there was more than 50 and more than 1,000 articles, but I think both of those are gross underestimates. Uh, everybody from gardeners to former police officers to pilots to sort of traffic control experts. It featured dozens of times in the British press, and the people don't exist. It's often AI-generated image. AI-generated copy and yeah, the person's just not there at all.

I mean, I've been following this since the beginning when you started it. I think it's about, is it 18 months, 2 years you've been looking at this?

It's just coming up to a year since we published the—

And we've had, well, as you say, Barbara Santini, Rebecca Lee. She's a good one, isn't she? Because one of the interesting things about this is the way that they're sort of credentialised online. So, Correct me if I'm wrong, but wasn't she an expert on the environment or she was opining on the carbon footprint of avocados?

Yeah, she, the thing was that with that one, I think the people really weren't making that much effort because she offered opinions on absolutely everything and had been quoted in really quite high-profile publications. The thing was that it was on that avocado story, the reason I looked at it was it was actually quite hard to find an expert I mean, I got one in the end, but then when I saw this one, I was like, okay, this person, they'd made mistakes with her because they had a different profile picture on some of the pieces. And, uh, on other ones, it was that same profile picture, but with a different name. And I mean, the reason people are doing this is because backlinks from sites with high reputation, which most newspapers have a sort of high domain reputation, backlinks from sites with a high domain ranking are extremely valuable. So what these people want to do is they're sending out hundreds or thousands of responses in an effort to secure links on newspaper websites, which will in turn boost the search engine rankings of sites. It's pretty trashy and it's very hard to get links in websites with good reputations. So casino sites, vape shops, sex shops, et cetera, et cetera. But as we sort of, as we went into the sort of second phase of it, we realized that some of these guys are clearly selling their services to other businesses like plumbing websites, travel agents, et cetera, et cetera. And I mean, what's never been clear is whether any of these sites knew what they were doing or whether they bought services from an SEO agency who just went, we'll improve your ranking, and then didn't tell them this is what we're going to do.

Well, that's a very interesting question because, correct me if I'm wrong, but I think on the case of Jesse Chambers, who I think was a prolific a travel writer, not?

Yeah.

I think you went to, if you can talk about this, to the people who were kind of running her, and they said, oh, it's just a representative voice. She represents a collective of our expertise. So was there, I mean, I don't know if you can tell our listeners a bit more to the background of that story, because it seemed a little bit of a mea culpa, although also still quite murky.

Totally. And I mean, I have to say, if I had been offering Global Work and Travel PR advice, my main piece of PR advice would be ignore him. Just don't say anything. It will make it worse. But they went ahead and talked to me and they did make it worse because I mean, it's just mind-bogglingly stupid argument that it's just okay to synthesize the opinions of everybody. And they also admitted that they were using AI to write it. So they're synthesizing the opinions of all the real people in their marketing department, but also using AI. And it was just, you know, it was classic because it's just a really bad AI-generated profile picture. No personal experience, no sense of lived experience, just really painfully average web content. And then they were targeting newspapers with just really, really trashy wannabe viral stories like what happens when you die on a cruise ship. And what happens when you fart on a plane.

Obviously, you've uncovered it now, so people are much more aware of it. But it was obviously getting into, not these examples specifically, but a lot of these examples were getting into The Times, The Sun, The Mirror, the tabloids, the broadsheets. What is the process whereby this is not kind of thoroughly checked?

I mean, I think it's sadly a reflection of where journalism is in 2026. You know, a lot of newsrooms have been gutted. In a lot of places, there's a real culture of, you know, do 10 stories a day, or you're fired. And also young journalists who are actually putting up most of the stories, you know, they're treated more like data entry clerks than anything else. So it's like questioning a story probably isn't gonna win you anything in a newsroom bar being shouted at by somebody. But, you know, in a lot of these cases, the experts were just disgracefully bad. Like, you know, it was just so blatantly obvious that they are AI-generated. My best example is Fiona Jenkins, who worked for the website Price Your Job, and she's allegedly a gardener, but there is no picture of her anywhere in a garden or even wearing gardening gloves. And the only profile image is a just staggeringly bad AI-generated headshot with just a plain gray background. I mean, it's just, it's appalling. And she featured 170 times in the British press, far more than any real gardener. And she was just focusing on stories that people knew would go viral. So it's like search-based queries, like when should I stop mowing my lawn for winter, et cetera, et cetera, just to sort of capture those search terms. And the problem is that editorially, people are making the decision, this sort of content is performing well, let's just do more of it. And then it gets more and more links for this site. And you know, that gives them the incentive to do it. The guy behind this company was paying Cision, which is a press release service, to host. I mean, they normally did about 2 or 3 press releases a day, and in total, that company had around 700 features in the British press. So for a press, your job, they were also using journalist response services like ResponseSource. So they're paying out a decent amount of money, but the amount of value that you can get in terms of search engine optimization, it's hugely valuable. So it's worth it for these organizations to target sort of depleted newsroom I mean, I've sort of described it, it's almost like a cybercrime. They've found a vulnerability, which is that these sites want to run viral content and they are targeting that. I mean, Barbara Santini, for example, used to just answer frankly bizarre questions that had been put out by newsrooms and she would always jump on them. It's like, what are the psychological benefits of playing darts? Now, no real psychiatrist is actually going to answer that question because it's so enormously stupid. But Barbara Santini can get a link book by answering it in the persona of a sort of psychiatrist.

But presumably there's a lot of harm that could be done there, particularly if you're a psychologist and you are AI-generated. I mean, we know that these LLMs hallucinate, so you're literally hallucinating advice that's going straight into press media.

Absolutely. And I mean, and that's widespread. There was another case where it's a sort of sex advice person, and I believe that they're using an LLM to generate that, but they're giving medical advice. And, you know, when you give medical medical advice. I mean, particularly with Barbara Santini, she's giving advice in publications that are read by millions of people, and it's advice that is just made up. And, you know, when you do that, you can kill people. These guys are deliberately targeting stuff where it seems slightly low stakes. It's targeting kind of soft feature content in order to get those links that are so valuable to them. But, you know, the thing is that when it's medical advice, you know, it's incredibly dangerous to give that And a lot of the advice is just bad. You know, these guys are going for viral content at the expense of absolutely everything else. So for instance, there was a, a pest control expert who also worked for Price Your Job who'd been featured more than 50 times, and he was giving advice like the best way to keep rats away from your house is to have onions because rats can't stand the smell of onions. And I contacted the British Pest Control Association who said that his advice was just ridiculous old wives' tales and No pest control professional would actually say that. I mean, you know, I'm not sure whether people really were going out and sort of putting onions around the perimeter of their homes to keep rats away, but it is, it's just irresponsible, bad advice. And the problem is that the sheer volumes at which this stuff was going up, it was swamping out real advice from real people, making it harder when you go on the internet or when you, for instance, use a tool like ChatGPT, which reads available content, it means that the actual human advice is becoming more and more unavailable.

I guess my favourite one of all of them, and there might have been some since that I haven't picked up on, but the Royal Cleaner. I'm sorry, that is, I mean, that is sort of comedic and tragic. Do you want to explain what happened there?

Yeah, sure. I mean, I still don't know who's behind those ones, but I have a suspicion, and they came from a fake PR agency called Signal the News. So, If you go to Signal the News's website, it's kind of like half finished. It's like a template website. And people from Signal the News were sending out like dozens and dozens of fake people. They came to our attention because there was a story about lottery winners, all of whom had lost their lottery tickets in amusing and sort of quirky ways, like their child had thrown it in the bin, their dog ate it. And then lo and behold, there was some great advice from a casino website called Play Casino, and there was links to Play Casino in it. And And because of that, Signal to News were on our radar and I'd had response from quite a few journalists on tabloids and broadsheets saying that they were just getting submerged in emailed press releases from Signal to News. And so I was sort of chasing to do a follow-up when I noticed that one of the stories that had come from Signal to News was about a royal cleaner. And this one, I mean, it was, there was a photo of of a woman. It didn't actually come up as AI-generated, but to me it looks AI, so I suspect they must have masked it.

Is it the only photo available of her?

Yes.

Yeah, right.

There's only one photo. I mean, that's a sort of hallmark of all of these. No social media, no LinkedIn, no trace of the person anywhere else, and only one photo of them, which is used everywhere. And I mean, in this case, She was sort of offering viral cleaning tips and little bits of insider information about the royal family. And I mean, she got an interview in the paper edition of The Times, which, I mean, that is a hard piece of coverage to get. And, you know, they clearly just haven't done due diligence on checking it. I mean, it used to, you know, I'm pretty ancient and I remember the days when newspapers used to check every single fact, at least double source it at least twice. I remember checking pieces on the Mail, and we were told you had to double source everything, and we should count ourselves lucky because on the Telegraph you had to triple source everything. And I mean, clearly that's just not being done. And she, she went everywhere. There was— she had things about the, the cleaning products that Prince Charles bans in Buckingham Palace, Kate Middleton's morning routine. I mean, I was baffled by that one because, I mean, how a cleaner is supposed to actually be privy to Kate Middleton's morning routine? What, is she hiding in the room? When Kate comes in and does it. I mean, you know, it's like, I mean, it was just baffling. But I mean, she featured more than 20 times in newspapers dispensing sort of viral tips about Buckingham Palace and their cleaning routines, et cetera, et cetera. I mean, and I mean, this organization was putting out stuff from wellness experts, police officers, chefs, all of them didn't exist. And all of them were promoting clients like Eddic Suits, Plum World, Ski Vertigo, et cetera, et cetera. I mean, I know for a fact that Signal the News or other organizations related to it are still sending press releases now because somebody sent me them last week. I mean, that's a part of the problem here is there's no punishment.

Well, that was gonna be my question actually. Who regulates and then who enforces it? Is this the domain of IPSO or someone else?

Ipsos have commented on it, and obviously there's now a campaign from the PRCA and the MCIPR, which are the two sort of main trade bodies for public relations in the UK. There's a campaign against this sort of deceptive practice led by them. But the problem is that it's very hard to regulate against this stuff. A lot of the stuff is coming from outside Britain. For instance, the guy who actually ran Barbara Santini is in Kenya. And the people who own that company are in Lithuania. And I was speaking to one of the journalist response services and they were saying they dealt with between 70 and 100 fake people per day. And the two main countries from which these things are coming are Pakistan and Bielorussia. And so these guys are just sort of aggressive link building SEO guys who are using you know, this vulnerability to build sites sometimes with a view to selling them on, or they build up sites to drive affiliate revenue. And so they have spotted a weakness, and it's very, very hard to regulate against people who are, you know, in countries that, you know, would— don't extradite. I mean, in the case of Belarusia. And so it, it's almost like cybercrime. It's extremely hard for these guys to be punished. I mean, the one thing that I hope our campaign has done is it's punished some of the brands that have used this.

Like who?

Well, if you search for MyJobQuote, one of the first things that comes up is our article explaining that they put 700 fake experts into the British media. So it's going to make you quite a lot less likely to book services through MyJobQuote because they're seen as a disreputable company. But I feel like there needs to be more kind of regulation and enforcement on this. Because like I say, Signal the News, which carried out all the most outrageous ones, the fake lottery winners and the royal cleaner, Signal the News is still sending press releases to— well, one of the editors on The Sun was sending me her inbox just a couple of weeks ago, and it's still absolutely brimful of trashy press releases from Signal the News.

You see, this is what I think is one of the perversions, or paradoxes if you want to be a bit kinder, about the Online Harms Bill and this sort of government policy and regulation is that we as real people are being asked to identify or authenticate ourselves, even just to use, you know, a social media network in future, right now, and more and more and more in the future. Meanwhile, there's this entire sort of parallel universe of complete AI identity slop, and the two are sort of coexisting. And by us being asked to identify ourselves isn't really helping eradicate any of the other stuff. That they're like two completely different worlds coexisting.

Totally. And what I feel that when it comes to, you know, for instance, PR organizations who want to get their clients into publications, being able to prove that people is real is now a sort of added value situation. You know, I know at The Times they took the royal cleaner extremely seriously. Several of the organizations I work for have instituted new policies which that you have to authenticate every single expert that you interview before you do it. I mean, you know, hopefully that will reverse some of this culture around, oh, just get the story up because that's been the problem. I mean, you know, there's plenty of them where they were just, the expert was just anonymous. There's like stories which ran in newspapers, which went, which quoted a pilot and you're like, no, no, no, which pilot? I mean, it's just, you can't just say, our pilot. And, uh, it, uh, you know, I mean, that really shocked me because each of these stories, it'll have been spotted by a copytaster, it'll have been handed to somebody on the back bench on the publication, and then it'll have gone to a journalist. And all of those three people thought it was okay to publish a story with an anonymous expert. And it's just like, well, you know, what's the point of a story where the whole story is the expertise but you don't know where it comes Am I right in thinking that that pilot was giving advice on the best times to travel?

Yes. Or something. I mean, that wouldn't happen anyway, would it? If one has the time to stop and think, your point about the newsroom being, you know, working at such speed and under such pressure, actually sometimes if you sit back and actually think about it, you would say your own brain, your own cognition would say, well, that doesn't really scan. That doesn't really make sense. Why is a pilot, you know, commenting on that. They wouldn't be asked to comment on that and they wouldn't have a view on that necessarily.

Totally. I mean, yeah, I mean, pilots probably book flights less than anybody else because, you know, they have to do it for their job. It's not like they're sort of going to be sitting there obsessing over when's the best time to book. I mean, it's just baffling. I mean, the ones that really, really made me laugh were the ones where, so they're just desperate to get the link here and they will just just marry up an expert with something that they have nothing to do with. So one of the ones I spotted in the last lot was it was a supposedly Royal Horticultural Society trained gardener who was working for a personalized number plate website. And you're like, what? And, but that was going into newspapers and getting links for these guys.

I know the world's post-normal, but that really is stretching it, isn't it?

It's just extraordinary.

But it does sort of, one can imagine a world in which these companies don't really exist, these ones that sit behind other companies that sit behind other companies, they don't really exist. They're pumping out content created by AI-generated people, quotes that are going into media that's probably AI-generated itself contextually, and then maybe even being written up by an AI journalist. So I'm like, this is literally a fantastical non-reality that sort of exists.

Yeah, no, absolutely. It is just baffling. And I really think that the serious part of this is that people need to put the brakes on and ensure there is human content. Otherwise, I mean, I think we are all seeing it in the internet around us. The sort of human touch is disappearing from it and AI is becoming the sort of new normal. And I think that people really, really, journalists, PR people, anybody in the information ecosystem needs to fight really hard to keep humans in it and to keep human expertise at the forefront.

But you recently did a story, I think, about Brian and Callum, who were AI-created authors or journalists. Is that correct?

Yes. Now this one, I mean, this, because, uh, the organization involved is extremely litigious. I'll, uh, not say their name, but the organization behind that, they've made a habit of buying websites and then they shift to the editorial focus of the website and turn it towards casino content, reviewing casinos. And then they put links for casinos into it, which earns them money from affiliates. Get revenue and boost the search engine rankings, et cetera, et cetera, for these websites. And in the past, they bought up lots of gambling websites, they've bought lots of sports websites, and they're now buying video game websites. I mean, what is truly depressing about it is a lot of these were viable businesses, but they're bought out and sacrificed just to get search engine boosts. And the amount of money in it can be calculated by the— they bought a very famous gambling website And apparently they paid £12 million for it. So, you know, they're paying market rates or over to buy up publications. Then they replace all the journalists with AI and shift it to casino content in order to drive people to non-GamStop casinos, which is, GamStop is a voluntary thing to stop problem gamblers visiting your website. But there are non-GamStop casinos which are basically designed to target problem gamblers. And these are, that's what these guys are advertising.

So, I mean, we talked about the purposes of this and getting the backlinks and rising yourself up the rankings. Is Google doing enough? I mean, you do report on and obviously explain really well the way in which Google has its experience, expertise, authority, and trust sort of vectors. Is that being modified or evolved? How does Google take account of this? And are they taking enough accountability or responsibility for it in your view?

Yes, I don't think they're doing enough at all. But, you know, the day when Google actually gives a damn about the publishing industry is the day when hell freezes over. They have never cared about publishers and I don't think they're going to start. I don't think they're doing enough. I think that they are the people who could really change this. And, you know, they single-handedly could get rid of this. If they sort of really clamped down on this practice and clamped down on the domains using it, not the people running it, but the people who are actually paying for it. So if, you know, your plumbing website pays for a dodgy SEO that uses this tactic, then they should be downranked. And I do think there are some manual penalties going out. So for instance, the gambling organization that we were just talking about, a lot of their websites have been manually penalized by Google. So there is penalties out there, but I just don't think they're being used enough.

Yeah. I guess it's all about incentives, isn't it? If your job depends on not doing enough or not understanding how to fix the problem, then you'll never fix the problem. But you do talk about and work with, I think, some, would you say they're forensic, digital forensics companies or sort of companies that are studios that are looking at trying to find deceptive PR, maybe these synthetic identities. How evolved is that? Is it that helpful? Is it that sophisticated? And where do you think that might go in the future?

So yeah, I mean, I work with people who are just using digital tools to track down this kind. I'm working with another set of people who are helping me sort of hopefully zoom in on more of these, because I mean, I, like I say, I think we've only seen the tip of the iceberg. I think there's quite a few more of these out there. So I'm looking for more of these. I mean, I think that the guys who I worked with, they're a sort of a NeoMam Studios. They just had a sort of bunch of data guys who were happy to work on it. And they're in the digital PR space. And I think within that industry, I mean, I've had a good response from a lot of people within that industry. There's a lot of appetite to see this stuff stamped out because they correctly, I think, see it as quite existential for them. If people are so annoyed with just this avalanche of just fake rubbishy content, then they're not going to deal with any PR agencies regardless of what they're selling. So, you know, there's a lot of appetite within that industry to push these bad actors out.

You see my, well, there's lots of concerns and worries about this, but one of them is that it reaches up, you know, away from sort of mainstream news, which is bad enough, but into academic publishing that these sort of fake experts really do enter at the top of academia. Somehow, especially when people are much more connected globally and there's many more experts, you know, over in different countries across the world. You're never necessarily going to meet them face to face, but they might have a profile. They might have a profile on a, like a research site and they might build up those contextual links to make it look like they are a real person, a real academic, and then pump out, you know, lots of academic knowledge, put in scare quotes. And the more and more of that that exists, and I think we saw a lot of it during the pandemic, you know, people coming out with expert opinions and then being sort of grabbed into the media, but they were hosted because they were academics, because they were seen to be much more trustworthy than perhaps anybody you would see as a sort of more mainstream commentator. Now, have you noticed any infiltration into those sorts of layers of expertise?

I think that there's always been an element of that because there's that temptation to become a media personality. And, you know, in every case, I always just think, okay, What is this person getting out of it? Like, you know, so when you, it's an academic and they suddenly become very, very prolific in the media space, they are looking to earn money in a way that is not academia. So, you know, suddenly a guy who's a very serious gut health researcher is like producing supplements or whatever. And there's always been that sort of permeability. I mean, I think one of the things that we're going to see next time around is that we're going to see fake experts that are just a lot better. Because quite a lot of the first round, they really didn't make that much effort with the fake expert. And so I think what you will see going forward is they'll make the LinkedIn page, they'll put them on the company website, they'll create fake social media for them, and it just becomes a lot less easy to see that this person isn't real.

I totally agree with that. I think if you look at what ElevenLabs are doing, and not to cast any aspersions on ElevenLabs, it's just the technology that people are using to adapt to to, you know, whatever uses they have in mind. But it's not just now sort of synthetic voices, it's the whole synthetic world that goes around that. So they're going to create the tools for people to world-build. It'll be like being in Roblox, I guess, you know, that you will have, you know, your friends around you, you will have a situational context that has been created by AI, generated synthetically. And it's going to be really, really difficult not just to compare and contrast and authenticate people and a very specific person against an AI version or a bot with the whole of the world that they've created. So it's going to get really, really difficult.

Yes, I mean, 100%, this is something that's going to get a lot worse. I was interviewing somebody who's a very, who's a big expert at deepfakes, and I was like, well, how good is deepfake technology? And she goes, well, it's this good. She pressed the button and turned into a completely different person with a different voice. And I was like, oh my God, It literally, had I not been known that the second person wasn't real, I would've thought they were 100% real. And that technology's not exotic, it's not expensive, it's not limited, it's there. And, you know, people can use it and they are using it. I think we're gonna see a world in which proving that you are real is gonna become a big asset for people.

Yeah, totally. Yeah, it's been coming a long time now. One of the other consequences of this is In this world of machine readability, where obviously this is a whole industry, a new industry growing up to make itself machine readable and therefore visible in the online environment, mainly through SEO now, but potentially in the future through LLMs and agentic AI, real experts are getting crowded out. Do you have any points of view? I mean, you're an expert, you have worked In all the—

Wouldn't go that far.

Well, you've worked across all the main media really, and in science and technology a lot, obviously. How do you, in the future, sort of compete with a lot of these synthetic AI experts, some of which will be found out, but others which they won't be and they will pass? And actually it's becoming very, very difficult for a human to make themselves machine-readable, to show up in SEO, or actually in the future to show up in these LLMs as an expert in a particular area. You can do all the code injections you like, but actually it's really competitive.

I mean, I think the way that humans can, human experts can stand out is simply to do things that machines can't. At the minute, I'm doing a story which I can't talk much about, but it involves a number of very prolific AI journalists who, they're fake people. And basically to catch them, I'm going to get a friend on a publication to commission in one of them and then ask him to go to a flesh and blood meeting.

Exactly. That's it.

Because that is what a machine can't do. They can't go somewhere and have a coffee with someone because, you know, you're gonna notice if a massive robot appears or whatever. You know, they can't do that stuff. They can't get on the phone. I mean, in this case, I think that these fake journalists, there's the people who are operating them actually impersonate the fakes who, even if I did a phone interview you, they could still get away with it. But it's the actually going somewhere, appearing, and trying to be the person that you say you are. That is what machines still can't do.

Which I guess is fine if you're in the same country, but it's not so fine if you're, you know, halfway around the world.

In Belarus, yeah.

Yeah, well, exactly, which Ray would be. So, I mean, I just think it's fascinating. This is surely gonna be just become a bigger and bigger and bigger story. Two questions for you, just before we finish, Rob, what can people do now if they want some tips or tricks or heuristics really for recognizing if they might be dealing with an AI-generated identity?

I mean, I think that it's just, we wrote in our articles about what to look for on Press Gazette, but I mean, most of the time it's just common sense. I mean, a lot of these ones that are out there at the minute, You know, it's just, they've made absolutely zero effort. I mean, I would say the hallmark that always marks these guys out is that there's no way to book their services. So Fiona Jenkins, the gardener, you can't book her as a gardener. Like there's no way to contact her to do that. Whereas, you know, I'm a freelance journalist. It's very easy to find me and book me. But with these guys, there's no way to do that. And it's like, it makes no sense because people have to make a moving. And if you are just a talking head that all they do is comment, there is no way to actually book your services in what you're supposedly an expert in.

That's an excellent tell, isn't it? And so one of the things I'm working on is as we become more and more machineable, if you like, and everything that you have just been discussing is about this world of machineability. And actually, if you can't turn humans into machines, you just invent human-like machines. That converse and communicate. We need to be even more unmachineable in very important aspects of our lives. And you've talked about some, you know, with judgment, with adaptability, lots of different areas. Where do you think we will net out 10 years from now? Will everything be much more machineable and we won't really be able to recognize reality, but we're kind of in it anyway, so we have to go with the flow? Or will there have been a backlash and a kind of reversal, if you like, of unmachineable traits?

I think both things would probably happen. There'll be a move towards a sort of very sort of technology-centered, AI-driven world, but at the same time, there'll be a backlash. And I think that things that have been out of fashion, like actually picking up the phone and talking to people on the phone, will come back into fashion because people will want to authenticate. They want to approach things in a sort of more human human, less screen-mediated way. And I think that we will see a significant kind of wider cultural backlash against AI everywhere and towards a sort of more human, more flesh-and-blood approach.

Do you think you'll still be writing about this in 10 years then, Rob?

Well, I mean, I have to say it surprised me that, it surprised me I'm still writing about it a year later and it doesn't seem to be going away. So who knows if I'm still around in 10 years, then yeah, probably. Brilliant.

It's brilliant to hear the background it. Everybody's got to follow your stories because they're absolutely fascinating. And I think it's fair to say that we're populated by probably more virtual AI personas than real people. People were worried about population explosion in the '60s and '70s, but they didn't realize there was going to be a virtual one.

Indeed.

Thanks, Rob. Brilliant to speak to you. Thanks for your time.

Excellent. Thanks, Tracey Follows.

My thanks to Rob Waugh for joining me. What Rob's investigation shows is that the problem isn't misinformation, it's something far deeper. Some people call it synthetic experts, I prefer to call it identity slop, because this is the moment when personal identity becomes just another form of highly commoditized and commercialized scalable content, produced really only because because the system needs it. And that raises larger issues for all of us living in AI-mediated information systems. Machine Dialogues is a special series from The Future of You. Find us on Spotify, Apple, and YouTube, or at mechine.com. That's M-E-H-Y-P-H-I-N-E dot com, or at traceyfollows.com and search for podcasts. We'd love for you to leave a rating, if you like the show, and do subscribe so you never miss transmission.