hello, i'm katie razzall. and i'm ros atkins. this week on the media show, we're looking at the disruption caused by ai to the advertising industry. meta has announced new tools that will allow people to make their own ads without needing an advertising agency. and we'll be looking at how china may see an opportunity in budget cuts to new services funded by the us. it's all coming up on the media show. but first, dan snow is the tv historian turned media mogul whose network history hit celebrates its 10th anniversary this month. it offers over 1,000 history documentaries as a streaming service. it has over a million subscribers on youtube and a very successful podcast. dan snow has been telling us

all about it and explaining why it's boom time for history content. first of all, i was aware that there were interlocking revolutions going on in the media. one was around reaching audiences. you know, muppets like me could just flick on the phone and reach people all over the world now. we had a great advantage of speaking english. we had this huge advantage of kind of english, british, uk soft power, cultural power. people are interested in british content. so, that was going on. it became easy, but these platforms existed. you could get people to upload credit cards into platforms, have a little netflix of your own, like we do at history hit. you could also miniaturise production. so, my first project, the bbc, 2003, went to egypt. excess baggage, £16,000 on excess baggage. we took cameras and cranes and sound kit. now we can travel on lime bikes with backpacks - and other bike companies are available - and we can also... and the quality of content you're achieving is high. because you're not taking any camera equipment. we're taking tiny cameras. you're taking little

small things like gopros. you're taking things that can shoot in higher resolution than those ones we took to egypt. you can take drones that pack down. so, you know, a helicopter in your backpack - it's extraordinary. so, these various revolutions were going on. and at the same time, i also thought, "this is madness. "my whole career and life depends on "some commissioner at the bbc liking the cut of my jib." and after 15 years making lots and lots of documentaries, i guess i sensed that i needed to jump. i needed to take this risk, this gamble, jump into this new world, because, a, i was still sort of young enough, i thought, and sort of resilient enough to do it. and, b, i just felt i'm probably not going to get asked to make this many history documentaries for the next 60 years, unless i'm david attenborough, which i'm obviously not. so i thought... did you get the sense that they had stopped quite liking the cut of your jib quite as much as they had done? i think so. why? we think you're great. that's very nice of you. the world is full of tv presenters furious at commissioners for sort of not seeing their greatness. and i think, with me, i'd never had a smash hit, you know, i'd never delivered a sort of planets like brian cox. so i think i was sort of knocking out documentaries that would do fine. and i think they thought, "well, let's see

"if someone else can..." and my view is, "fine." i mean, we're all... we all get hit by lightning. we all win the lottery to become tv presenters. i'm incredibly lucky. my view is live by the sword, die by the sword. so i took what... the extraordinary benefit i got from working all those years and tried to parlay it into this digital empire that i could sort of control and have some ownership of. so you did that, and it appeared to work. when did you decide to go beyond podcasts? well, yeah, it might appear to work because i was very unremittingly enthusiastic about it. but deep down, behind the curtain, it was terrifying. was it? yeah. when you fire up a youtube channel with zero views, you know, you're used to being on the bbc and discovery and american tv, and then here you are sort of talking to your mobile phone from a kitchen, and three people are watching. one of them's your mum. you know, it's... but i bet she liked it. well, she was unremittingly... yeah. she was very positive about it. i didn't really trust her opinion after a while. so where did you find your audience, then? how did you build it? well, the podcast was the first thing, because the joy of the podcast... all these other things, it's nice to go viral on these social media platforms. they don't pay you any money. and the great thing about the podcast is you can

monetise it and advertise it, because people tend to listen from start to finish. people tend to develop quite strong relationships, like your audience have great relationships with you guys. or something about audio - they take you with them on runs and walks and in domestic quiet spaces. listen to you when they go to bed. so, for some reason, advertisers, thank goodness, think... i don't know why i'm telling you this when i'm sitting next to one of the greatest advertisers in the world. but anyway, um, we can hear his opinion in a second. not advertiser. no, no, but a man, you know... wish i was an advertiser. that would make it easy. that's sir martin sorrell. but... so, podcasts, quite early on started to attract quite a lot of ad money. and so i would read out all these ads and so suddenly you were getting paid for things you were doing on the internet, which after years of putting pictures on and trying to go viral and messing about, it was very exciting, that actual link between virality, between numbers and listens and eyeballs and money was very exciting. so, the minute you establish that, you can grow and you can start to employ people, employ people cleverer than you are. and then, you know, i partnered up with a guy called justin gayner, for example, tom clifford, and they helped me build this

really from the podcast, a netflix for history. so, a subscription channel where you can make tv shows that people around the world pay for. and that means i've achieved the sort of nirvana of creative people, which is we now...we sit down with our team and we decide what we want to do. we say, "have we done anything on magna carta? "what have we done on the inca recently? "let's go out to peru and have a look at the inca." and we can do that because we've got this... ..pretty predictable, you know, subscription model and cash flow that means you can fly out to peru. you can make podcasts out there. you can do instagrams, tiktoks, youtubes out there, the youtubes monetising beautifully. and then... and it all... so all this one piece of content, keep subscribing, and the joy of history, folks, unlike your politics and sport, is that history is evergreen. and do you, dan, ever see yourself as... are you seeking to share an opinion? some of your podcasts, you look at current affairs, obviously through the lens of history, whether that's, you know, recent ones, why isn't canada the 51st state? or who owns greenland? we know these are political

zeitgeist-y things right now. are you expressing a political opinion? the zeitgeist-y stuff works really well. there's big... audiences like it. audiences are... they really want to know the context of the things that are going on in the world around them. as the world gets, frankly, to many people, more disturbing or crazy, they want to go, "hang on a sec. well, actually, what is the status of taiwan? "what is going on in israel-palestine? "what is going on in greenland? panama?" trump is good for business, but... and of course you want to express opinions. i mean, i would say that every time i've done that, it just enrages portions of the audience. and what's interesting, i have to say, to be honest, is, as it's got bigger and more commercial, the pressure has been on not to express opinions because you alienate chunks of the audience. so, therefore, we've gone from this kind of fun insurgency where i slagged off brexit all the time to actually thinking now, "oh, well, actually we have huge numbers of listeners "in america and lots of them probably "lean slightly, slightly trumpy, slightly right." so what are the things you're avoiding? what are the most challenging ones that people don't like? well, at the moment, you... well, i think the key thing is to get the tone right.

like you guys have to work on every day here in this building. it's actually very interesting to do, "why isn't canada the 51st state?" but without just egregious trump-bashing. you can kind of let the history and the context do that for you, i think. but do you feel like you've lost income? i was watching a very high-profile youtuber the other day talking about how he felt he had lost income into his business because of opinions he had expressed. do you feel that there have been times when it's cost you? yes, yes, i have. because... so, interestingly, the way that findability on spotify and the other big platforms works is if they see a weird spike in negative reviews, that will inevitably impact... the computer will go, "actually, that seems a bit weird. "let's not push that to as many people as we should," which is kind of understandable. therefore, very small groups of well-motivated antagonists... i probably shouldn't tell people this. they can affect the way your podcast is seen and heard and found. ai is upending every industry.

and this month, an announcement by mark zuckerberg raised big questions for the advertising world. meta is promising a range of tools that will allow anyone to create their own ads without a traditional ad agency being involved. does this pose an existential threat to the big players in the ad industry? a question we asked sir martin sorrell, famous in the industry for founding wpp and s4capital. if i think about it historically, the first golden age was around globalisation. we can say that's finished. dan may agree or disagree. you can say it's checked, or stuttering. more fragmentation coming. the other big force has been technology, and we've had the internet revolution in the '90s, we've had the smartphone revolution in this millennium. and now we're going through an ai revolution. it's not just ai, by the way. it's around quantum computing and it's around blockchain.

i don't get crypto, but that crypto is getting increasing relevance, for good or bad. warren buffett would say bad and charlie munger would say the same if he was still alive, but... so, i think it heralds a different era. and i think it's a new golden era. s4, or its brand name, its branding, operating brand monks, really is focused totally on the digital age. i mean, just to put it in perspective, our industry is a $1 trillion industry globally. about 700 billion of that trillion is in digital and 300 billion is in traditional - free-to-air tv, etc. newspapers in the old form. the traditional industry is going backwards by about 0-15%, depending on whether you have live sports or not. the digital piece is growing by 10-15%, dominated by google, which out of that trillion is 250 billion of ad revenues. meta's 150 billion, amazon's 60 and tiktok outside china

is 40 billion. so, those four platforms are 500 billion - half of the market and 70% of digital. that's the big shift that's taken place. so that shift could lead to what you say may be a golden era, another golden era for advertising. but are you concerned that there's products being proposed by meta and others that would allow people to generate adverts much more easily than they could have done before, that it will simply remove the business model of huge sections of your industry? no, we have consistently said since we started s4 six years ago - and mediamonks and mightyhive, which are the two core companies inside s4, date back further - we've consistently said since inception that we believe the market was moving to a platform-dominated market. the four companies that i mentioned, plus alibaba and tencent, so you have three in the west and three in the east, and they would dominate the industry and the investment needed for ai, which is about half a trillion dollars a year that the big...the mag 7, or the maga 7

as it's now called, are investing in ai capacity. we consistently said those platforms will continue to dominate, and the role of the agency is as validator for the algorithm. i mean, traditionally, you wouldn't go to rupert murdoch and say, "here's my media budget - invest it." you're not going to go to mark zuckerberg and say the same. as i said, meta is 15% of the industry. what you will do is you will look at the algorithm. you will look at what the algorithm spits out in terms of solution, and then you will ask an agency or a third-party independent, in inverted commas, to validate that. but there'll be people listening to us thinking, "ok, well, we're looking at this "from the industry's point of view, "but what will this mean for "the type of adverts that come my way? "are they going to be personalised? "what form will they take? where will i find them?" give me a second just to run into the sort of five things that we see going on. the first is around visualisation and copywriting. so, a 30-second ad used to take... it's very similar to what dan said about programming

in the old days. used to take two months, used to cost millions of dollars. we produced ads for puma just recently, and we're working on several other clients at the moment which are being launched where we produce the same amount of material, the same material, using ai, literally in days for hundreds of thousands of dollars. so, visualisation incorporating, time compressed, time to market. second is personalisation at scale. you mentioned netflix. what i call the netflix model on steroids, using first-party data, that is consented data that consumers have given our clients, plus the signals from the platforms, which says what we're doing on media, to create highly personalised ads at scale. for netflix, for example, who we work for, or amazon prime, or disney+, we'll produce 1 million, 1.5 million assets for a campaign. we can do multiples of that. third area is media planning and buying. planning that trillion using algorithms, just like they do in the equity industry, in the investment industry.

so, using computing power to improve how that's done. exactly. to improve the investment of it, general efficiency. and lastly, most importantly, i think democratisation of knowledge. the bbc is a very siloed organisation. if you can use ai to inform everybody in those bbc silos, if i can put it that way, about what's going on as a whole, that democratises the company, flattens the organisation, makes it much more efficient. well, thanks, sir martin. we're also joined here by alex dalman, who heads the advertising association's ai taskforce. the advertising association is the trade body for the industry and you, alex, are also managing partner of faith, vccp's ai creative industry. yes. welcome. thank you. it's great to be here. great to have you. just first, your take on what sir martin's been saying, but also more broadly about what mark zuckerberg has been saying about the future of the industry. yes, we all saw that article. i think it's quite interesting. these tech bros do go out there

and like to cause a bit of a ruckus in the industry, but when we think about it... i mean, it's a great idea for small businesses, you know, being able to create your adverts and, actually, get out there a lot quicker and grow your business. so, it's kind of democratising it for the small businesses. but the big businesses that we work with all day, every day, i just think mark zuckerberg is basically putting more kind of ai slop into the world and more average advertising, when actually the kind of beauty of advertising is, you know, you're going to raise that bar of creativity. and that's the reason why we started faith in the first place. we wanted to actually see how ai could be an accelerator of creativity and imagination, and not really just make it an average of everything. and do you think it can be? because i remember in la a few years ago being in an advertising event about ai with a lot of creatives in the room saying, "you're going to... this is going to take our jobs. "we're worried." and all the people on the panel saying, "no, no, we still need creatives." absolutely need creatives. i mean, these are just tools, at the end of the day. they are, you know, very,

very powerful tools and they're, you know, amazing and they're going to change a lot of things. they will change our industry. but, you know, you've got to put something into the tool to get something good out and that still needs to be led by a creative eye and a creative vision. and why should consumers care about what we've been talking about? i think there just needs to be a bit of a realisation that, you know, you're on social media quite a lot. you are going to consume ads whether you like it or not and actually you're probably going to scroll past most of them. i think the average at the moment is you have about 2 to 2.5 seconds to grab someone's attention, to get them to buy your product, or know your behaviour change campaign, or whatever you're trying to do. and i think, yeah, consumers, you know, the bar for an amazing visual now, an amazing, you know, video content that you haven't seen before on social media is so high. i scroll past so much stuff. you know, you're competing with so many different things. you're competing with your friends' content, you know, the publishers' content, the bbc's content, everyone. so i think the bar is so high.

and what about when we think about advertising or when i think about advertising in my role as culture editor here at the bbc, i think about those incredible directors who started in advertising, and then it was this pipeline, this crucible? frustrated film director. exactly. well, they then became film directors, right? with sir ridley scott or, you know, david fincher, david puttnam, a lot of them started in advertising. do you think ai will actually close that pipeline, that we're not going to get our film directors? absolutely not. i think it's going to make probably more people who couldn't get into the industry before able to get in. it's going to lower that bar. you know, i don't know how all these film directors started out, but there is this thing in, you know, the industry and the film-making industry that you have to know people to get far. now, anyone who's got access to the tools, you know, everyone's got a smartphone, pretty much. everyone can be a tiktok creator. everyone can... you know, this exists. what doesn't exist is that kind of high-level creativity, and you need to push that, and they'll be the people... just to underline, if you're personalising ads at scale,

and a lot of people are doing it and people are being bombarded with ads, the creative differentiation will become more important, not less important. right. and when we talk about personalised ads, alex, this idea about british ads as a sort of cultural force, you know, up to the '80s and even '90s, everyone was watching the same tv, the same adverts. people will remember, if they're older, you know, whether that's the hovis ad or the milky bar kid or the tango man, milk tray man, all of that. this will all obviously be lost on younger listeners, but do we think we just won't get that universal recognition of the ad in the way that we used to? do we not already? have we already lost it? i think we have slightly lost it. but what that does mean for advertising agencies is you have to work a lot harder to get people's attention. and in the spaces that they're in. we talk about a lot about big ideas for small screens, because it's no longer just, you know, ok to create kind of average ads. you know, you want the things that cut through, that get into talkability, that people actually want

to share. so, i think, yeah... but what is... ..what is the problem in using technology more effectively? i mean, what the agencies have to do is two things. we have the same argument about data. data destroys creativity and intuition. the luddites had the same argument a long time ago. exactly. and so it creates better insight, data, through data. technology can produce better work. it's not a question about displacing people. maybe we will lose employment... but there is a counterargument, isn't there, which is that data will give you an analysis based on everything we've always done, we've done already, what audiences and consumers of advertising have done already? it won't necessarily be able to predict the thing that you didn't know. well, that's a charge against llms because it's based on all the knowledge that we knew already... it's a charge against data analysis as well, isn't it? not that data analysis shouldn't be used, but... i disagree. by examining the data in detail, you might uncover an insight that gives you, you know, an idea that is mould-breaking. dan. sorry, sorry.

i just wanted to bring dan snow in on this as well. i mentioned the luddites, but can you draw any comparisons between the technological changes we're living through now and what's happened in history? i'm not accusing you of being a luddite. my head just exploded. i can do those things. but i also... interesting with data, if we went with the data, we'd make... i'm giving away a secret. we'd make every single piece of content we have about anne boleyn. because that's all... i'd listen. well, there you go. but so we... for some reason, the world is so hot for anne boleyn. how many different pieces of anne boleyn content have you made? did she really have six fingers? we have to... no! we have to... we have to take her out of our top list otherwise it skews all our... but that's the point. so, there's a place for simply, "we need some more anne boleyn, lads." and also then just throwing in a few hail marys and discover what the next anne boleyn's going to be. people don't know what they don't know and don't want, so we have to have that creativity. what's the number two on the list, then? i can't...i can't tell you that. she's so far ahead of everything else. but in terms of... obviously, yes, i mean, the story of our species is one of extraordinary technology.

what we're going through at the moment is obviously radical, but it's not unprecedented. for 200,000 years, homo sapiens are knocking around. then about 5,000 years ago, we mix tin and copper at 1,000 degrees centigrade and discover bronze. 5,000 years after that, we're flying a drone on mars, you know. so, actually, we're living... this is just the latest spasm of an extraordinary technological upheaval that has seen, i'm afraid to say, populations wiped out, religious transformations. it's seen north and south america settled. 90% of north and south american indigenous people dying within 100 years because of this technology that was introduced. so, i mean, you know, the one that people like is the printing press, of course. you see protestantism sweep across europe within a generation. you see a vast spike in women being executed for witchcraft as a result, so unintended... so unintended consequences. we have seen again and again and again, nobody understands the nature of the technology we've released. now, in march, we discussed cuts announced by the trump administration to

the us agency for global media. this oversees and funds a range of news organisations whose content is primarily targeted at countries where media freedom is restricted. the cuts have resulted in a reduction in services, and the washington post has now reported that china has filled the gap by adding new services and jamming the frequencies previously used by one of them - radio free asia. to understand the implications of this, we've been speaking to michael bristow from the bbc world service. if you look at the chinese state media, they're reporting after president trump's decision to withdraw funds for radio free asia, chinese media were cock-a-hoop about what was going on, really saying, "this is the end of fake news in china." and so it wouldn't surprise me at all if china, the authorities there, hadn't spied an opportunity, moved in to take those frequencies, those short-wave radio frequencies that had been used

by us radio free asia previously. they've done it before. a few years ago, the australian broadcaster abc, they withdrew some of their broadcasts and chinese state media used those frequencies. so, they have done that before. and how easy is it to find out if they've done it this time? i don't think it's very easy. the washington post article, if you read that, it was using information it had collected anonymously, it had been given that information, but it sounds certainly like something that the chinese authorities would do. you're certainly not going to get the chinese authorities telling you about that because that would indicate to the rest of the world that they're interested in essentially stopping the western world reporting information about china to chinese people. and that's certainly not an idea... certainly not a narrative which they want to promote to their own people. right. and i think we do know that radio free asia is no longer broadcasting to many of its listeners, as i was saying, including to uyghur and tibetan listeners.

can you offer any examples of how chinese state media already broadcast to or indeed about those audiences? well, ever since the chinese communist party took over china 70-odd years ago, they've been engaged in a massive amount of propaganda and controlling information that its people get, essentially trying to wall off any bad information. that's only become more severe since xi jinping took over in power in china a decade or so ago. they've got the great firewall, so they've walled off the internet to their own people, so people just can't access... you can't log on, for example, say, to the bbc chinese service in china and look at what's happening there. you just don't have access to this information. so, this is part of a long-running decades-long battle to really limit information to their own people. at the same time, interestingly, china is doing another thing. it's trying to get

its message out about what's going on in china to the rest of the world. it wants... it believes that media organisations such as ours, the bbc, essentially changes the message of china and represents china in a really negative light. it wants to change that. so it's spent a lot of money over recent years funding english, french, spanish, russian, all kinds of language services to get the chinese message out and bypassing traditional western media organisations. so the message goes two ways. and just finally, michael, is it too simplistic then, based on what you're saying, to say that if western-funded media operations like radio free asia retreat, not just within china, but more broadly in the region, that leaves opportunities for china to explore those ambitions to exert influence via media outside of china? oh, exactly. definitely. they'll be absolutely thrilled that president trump has taken this action

because china... particularly in east asia, the countries around east asia, they seek to present themselves as an alternative to the americans. "we're your neighbour. we will help you. "don't look to the americans." the americans withdrawing gives them an opportunity to kind of fund media operations and just get their message out far more easily without any contradiction from alternative voices. so they'll be absolutely thrilled at what's happening in america. that's michael bristow from the bbc world service, the last guest on this week's edition of the media show. yep, that is it for this week. thank you so much for your company. goodbye. bye. and if you'd like to hear a longer version of today's show, search bbc the media show wherever you get your bbc podcasts.

live from singapore and this is bbc news. us president donald trump lie just to make a decision on us involvement in the iran is real conflict within the next two weeks. iranian and western leaders will meet for the first time since the conflict began at a key summit in geneva. russian president putin offers to broker a deal between iran and israel saying a solution can be found. in other news thailand's coalition government is on the brink of collapse after a lengthy phone call between the prime minister and a foreign cambodian leader. in the business of a protest