Virtually Anything Goes - a Made To See Podcast
This podcast focuses on interesting conversations. Each series has a different theme. Episodes cover a wide range of topics including marketing, AI, communications, healthcare, addiction, sleep & insomnia, sales, livestreaming and many more.
Get to know our guest, their story, and their expertise. Learn something new, be entertained, and discover something new.
It's in th ename: Virtually Anything Goes.
Virtually Anything Goes is a MadeToSee.com podcast
Virtually Anything Goes - a Made To See Podcast
CEO Benjamin Field on Virtually Parkinson, AI, Ethics, Deepfakes, and Leadership
Step into the future of media with the creator of Virtually Parkinson, the world-first podcast hosted by an AI recreation of legendary broadcaster Sir Michael Parkinson. In this episode of Virtually Anything Goes, Deep Fusion Films CEO and Co-founder Ben Field reveals how he and his team brought the iconic interviewer back to life through cutting-edge AI and groundbreaking production workflows.
AI That Thinks, Listens, and Interviews
Ben takes us behind the scenes of building “AI Parky,” trained on more than 100 hours of classic interviews to hold entirely unscripted conversations with celebrity guests. He shares how the custom-built “Squawk” software allows the virtual host to react in real time and creates surprisingly personal, even therapeutic, discussions that feel authentically Parkinson.
Setting the Standard for Ethical AI
Beyond the show’s wow factor, Ben is shaping the global conversation on responsible AI. From writing ethical frameworks for the BBC’s Gerry Anderson: A Life Uncharted to advising broadcasters and policymakers, he’s helping define how AI can enhance creativity while protecting intellectual property and artist rights.
The Journey Behind the Innovation
Ben also reflects on the career twists, from acting and BBC comedy writing to award-winning directing, that prepared him to lead at the frontier of technology and storytelling. His message: bold experimentation, trust in your team, and clear ethical guardrails can turn big ideas into reality.
If you’re curious about where AI and entertainment collide, or how policy and creativity can coexist, this conversation delivers inspiration and insight in equal measure.
This episode is part of our Leadership Stories Series, where we speak to leaders from a variety of different backgrounds, including AI, Strategy, Marketing, Executive Coaching, Healthcare and others! Subscribe and check out our other episodes on Youtube at @madetoseemedia
Connect with Benjamin Field on Linkedin or find out more at https://www.deepfusionfilms.com/
Connect with Lev Cribb on Linkedin
For more information, content, and podcast episodes go to our Youtube Channel or https://www.madetosee.com
Monday afternoon I rang Jamie and said hey, jamie, how do you fancy resurrecting your dead dad?
Speaker 2:Hello audio listener. This is your host, lev Kribb. Thank you for choosing this episode featuring our guest, ben Field. If you prefer video, you can also find all of our podcast episodes on YouTube or on our website at made2ccom. But now I'll get out of your way and hand you over to well me. Hello and welcome to the Virtually Anything Goes podcast. This episode is part of our leadership story series, where we speak to leaders from a variety of different backgrounds, including AI, healthcare, software, strategy, executive coaching and others, and if you like what you hear in this episode, be sure to subscribe so you don't miss any upcoming episodes and, of course, check out our previous episodes too.
Speaker 2:Today I'm talking to Ben Field about his story into leadership, what got him to where he is today and which obstacles he encountered along the way. But we'll also talk about some of the projects he and his team at Deep Fusion Films are working on, including the unique Virtually Parkinson AI podcast. Ben is the co-founder and CEO of Deep Fusion Films, creators of Virtually Parkinson, a podcast hosted by the AI version of the renowned late talk show host, michael Parkinson, and we'll definitely be talking about that. It's absolutely fascinating. I think you'll see that when we get to perhaps a clip as well, but first I want to tell you a bit more about the company behind it.
Speaker 2:Deep Fusion Films is a production company which blends technology and creativity to generate high quality storytelling with a focus on the ethical use of AI. It produces a diverse array of content and that captivates audiences across a variety of platforms. And on top of that, ben and his team are market leaders in the ethical use of AI, as I mentioned, within the creative process. Ben is also a broadcast award-winning producer and when he has time, also turns his hand to acting and singing and, in his own words, occasionally in tune. Ben Field, very warm welcome to you. It's a pleasure to have you on the show.
Speaker 1:Thanks, david, it's very lovely to be here. Thank you for that very lovely introduction.
Speaker 2:Yes, You're very welcome to be here.
Speaker 1:Thank you for that very lovely introduction. Yes, occasionally singing in tune is, I think, the nicest way of describing me at any point.
Speaker 2:Those were your words, I'm just quoting you. They were yes.
Speaker 1:You know I'd forgotten. I'd said them and it's very lovely to be reminded.
Speaker 2:Excellent. Well, if this is your first time watching the Virtually Anything Goes podcast, or indeed listening to it, stick around until the very end, where we'll have the Virtually Anything Goes question, where I turn control over to Ben. He can ask me anything at all, any question whatsoever, and I will have to answer it. There is nothing prepared about this, and when I have answered my question, given my answer, then Ben also has to answer the same question. That's my safety net. Good, so, ben, you started out as a voiceover artist and actor and now you find yourself as a broadcast award-winning producer at the forefront of ethical use of AI in the creative production space. Producer at the forefront of ethical use of AI in the creative production space. Did young Benfield imagine this would turn out this way? What were you like as a young boy or young man? How did it all start?
Speaker 1:Yeah, no, I never saw it being this way. I mean, I think, you know, when I was sort of 13, 14, uh, when I, when I first encountered sort of gcse drama, I guess you know, uh, or uh, acting as as sort of it would have been put to me at the time, I guess, um, you know, amateur dramatics and all the rest of it, my, I think my heart was set on becoming the next James Bond. Um, obviously, uh, that that didn't, that didn't go to plan, that, um, but you know, honestly, for you know, probably the first 10 or so years from that point, you know, definitely, acting was my first love and and, yeah, you know, uh, I, I did get some roles and and I, I played fagin on tour and that was very lovely and you know so it was, um, it was, it was a key driver, um, but, and my, my mum and dad at the time, who are, you know, definitely realists, I think it's fair to say uh, would always try and stress that I needed a backup, I needed something else to do in case acting didn't work, and I was, you know absolutely well. Of course it's going to work. I mean, have you met me? And it turns out they were right, which is more annoying than perhaps anything else.
Speaker 1:And so what happened was that I, you know, I just had to diversify after I worked out that this wasn't going to pay the bills. I wasn't going to be able to have the life that I wanted. I couldn't really even move out of home without additional support. So I had to get a very regular job while I kind of sorted myself out. And that was, I actually went into banking, um, and not, not in not London style hedge fund banking I, I mean, I was a cashier for Barclays, um, and, and that gave me some time to sort of readjust.
Speaker 1:And what I realized was that, um, I wasn't going to give up on the dream, uh of of performing and acting um, and that actually it was everybody else's fault for not casting me um, because the the problems weren't right. It's obviously everybody else's fault, um. So I said, right, well, in that case I will, I will just write my own shows then, and I will, I will, you know, I'll direct those and I'll I'll make them, um. And so I did that, and in fact I worked with a comedy writing partner who I worked with probably two or three years, named Jason Smith, really really nice guy, incredibly funny, far funnier than I was and we took a show up to the Edinburgh Finch Festival and it won a Critics' Choice Award. But you know, but obviously, being the arrogant watsit that I was, I took the lead role in that, in that play, as well as directing it and co-writing it with Jason. And so, yeah, you know, we ended up in this scenario. I was like, yeah, see, I can do this, I absolutely can.
Speaker 1:And what resulted from that was the fact that jason and I then went and worked. So it was noticed by the bbc, and so the bbc said you know, do you want to come in and do some writing with us? Uh, so I ended up, you know, sort of being a comedy writer with the bbc for for a short while, um, and and because writing wasn't really my key skill or core skill and Jason wanted a real job, we weren't able to carry that off, but I'd then been introduced to a new skill set, if you like, of directing and I'd taken directing modules at university and I'd taken directing modules at college, but, again, not something I would have said was was a key skill, um, but it was whilst I was writing comedy, um, with the bbc put out a call for is there a, is there a type of sketch show that blends live animation, live animation with live action, and can it be done in front of a live audience? So nothing, really, a tiny, insignificant step in the means of content production. Everything needs to be live, everything needs to be in front of a studio audience.
Speaker 1:And so I stumbled across a company called Musion who made holograms, live holograms. So Musion, for those that don't remember or didn't know, in the first place, they were behind the Gorillaz live show, if you remember. So Gorillaz was obviously a band that was, uh, animated. There were cartoon characters and they performed live with madonna at the mtv music awards, um, and that was kind of their big, that was their big moment. And I, I stumbled across them and went hey, actually I wonder if there is a, you know can, can there be a comedy act between cartoons and people? Is this the answer? So unbelievably expensive to do, but really cool tech. It sort of died a death, but I then contacted Musion after that point and went hey, look, I'd really like to make content for you. I think this is a really, really interesting phase. And so for three or four years years.
Speaker 1:I was then writing and producing content for uh musion shows, uh, which culminated in me doing what was described at the time as the digital resurrection of um morcombe and wise for children in need in 2012. Uh, so we recreated more wise dancing with uh chris miles, uh on on children in need in 2012, um which, looking back now, um, you know it's almost predestination, isn't it? The the you know what what I'm then known for many years later is, is yet another. You know what has been described by people who are less kind as digital meat puppetry, um, and, and being a digital grave robber. So, uh, yeah, it's sort of you know it. It all feels part and parcel of the same journey, but it stemmed from me wanting to be a performer and then finding that I needed to diversify in order to actually make any money right, yeah, it's fascinating and we'll get to obviously the, uh, the, the, the virtual parkinson side, um, and the digital meat bubble.
Speaker 2:Well, it's not mine or somebody else's words, probably, but uh, yes, we'll come to that because it is. It is absolutely fascinating. But, um, I just want to, I suppose, dive a bit deeper into you. Look at your profile and, um, it is is quite obvious also from just what you said. You know that you have a passion of for storytelling and you know that that clearly comes through in in the work you and your team do.
Speaker 2:You know the way you just spoke about? You know the the work you have been doing, even when you were earlier in your career. Why is that? Why do you feel so passionate about storytelling and where did that come from?
Speaker 1:I mean storytelling, I suppose, do you know what? So I think Lisa Nandy actually mentioned it just in the last couple of weeks which is actually that, um, you know, amateur dramatics and community theater, uh, which you know the league of gentlemen gave an absolutely awful kind of rendition of. I don't know if anybody remembers the league of gentlemen doing, doing community theater based sketches, um, but community theater is, is essentially what, um, the majority of all entertainment is based on. You know, if you, if you go far back in time, what it is is it's about bringing people together and sharing stories, um, and and bringing people closer together, um, and sharing experiences, and that that is, I think, one of the things that's never left me, because I've I've always been, I've always felt the most included when I'm surrounded by people who are enjoying themselves, um, you know, and I I think that's probably the same to everybody you know, we all have friendship groups, because when we're together we feel safe and we feel, you know, we and, and the extension of that um is that I've always enjoyed um trying to make people laugh. I enjoy actually making people laugh more than I enjoy the trying to make people laugh, um, but but again, it gives you that, that sense of oh look, everyone's enjoying themselves. I'm, I'm having, I'm a nervous wreck, but everybody's enjoying themselves and that's and that's really useful, um, and and I think there's, there's something.
Speaker 1:I was a very, very socially awkward teenager, I don't you know. I wouldn't say that I was confident in myself and you know, probably pretty rubbish at making friends, to be honest. Pretty rubbish at making friends, to be honest, um, and and so it was a coping technique, for for an awful lot of that time was was actually about, you know, right, if I'm making people laugh on on stage, well, first of all, they're at arm's length. So you know, they, they, absolutely you know they can't attack me for this. And if I was making them laugh, then that that probably meant that I was, I was doing a good thing and that, you know, nobody, nobody beats up the kid that makes them laugh, you know. So that's a, that's a really handy thing to to learn, um, doesn't, doesn't hurt with attracting women either, um, so that's always handy, you know. It kind of gets you through that first period and I suppose I've always had that need to be liked, I suppose, by being able to entertain, and I was able to turn.
Speaker 1:That, you know, quite, I suppose, quite toxic trait. In many ways it's sort of quite as a level of self abuse. I think to that in needing to entertain others all the time, you know, I don't think that's a particularly great trait for anybody to have. I think that you know that could potentially be quite a dark place to end up. But I was fortunate enough that I was able to turn that skill set into something that was a lot more healthy and recognize my own failures, but was then able to turn that skill set into something that was able to make me money and that took a long time and go through many, many different iterations of what I actually did for a living.
Speaker 1:So I taught myself pretty much every element of either theatrical production or TV production, largely because I was rubbish. You know I didn't, I didn't go. I didn't do media training per se. You know I didn't, i't go, I didn't do media training per se. Uh, you know I didn't. I've never been to the met film school, I never did the nfts or anything else um. So I had to, you know, be self-taught um, and so nobody would employ me. So I went around doing all the jobs, um, and managed to get employed on, you know, on other people's projects, doing individual elements of that, but I was never really trusted to be a producer or a director on anything that wasn't my own production.
Speaker 1:Uh, probably until you know, 2019, I guess you know I mean, I started going into. No, that's not necessarily true. I did, uh, 2016, um, and I'd done a couple of couple of projects prior to that, but, you know, the, the big shift was around 2016, 2019, um, you know, and those that that journey has just been one of self-discovery, really, uh, but the, the driving need to always understand everything, um, and be able to do it, you know, has sort of led us to this point. But one of the one of the really interesting ones, um, that I, I, I did was um, to give you a, you know, an absolute case, case in point. So I, I used to be, um, I told a, a um, I used to be, I told a.
Speaker 1:I was invited in to do a job off the back of the resurrection of Morkman Wise. You know, that's essentially an animation job. That was to a large degree, and so the guy I was working with on that said, oh well, actually a series now for for cbbc. It's an animation series, um, can you take a couple of? They're five minute shorts, uh, can you take a couple of episodes for me and can you animate them? Um, I had never worked in after effects before um, and and he just went. Well, here's a copy of After Effects. You can have this and you can work from that. Can you give me the episode back? I need it in 10 days. So I had to teach myself After Effects and deliver an episode of After Effects in 10 days, an episode of CBBC content in 10 days. And that I mean, my God, it must have been awful. What I delivered must have been awful. What I delivered must have been awful. Um, but I did two episodes, uh, and in fact I think I went on to do three or four um, but that gives you an idea of the type of person that I am. You know, I rather, rather than openly admit no, I can't do that I will always say yeah, okay, fine, I'll work it out and invariably, to you know, one degree or another um, I did.
Speaker 1:There was a. You know, there was a another example where, um, this was after that uh, we were making um green screen sets for a comedy series. Um, that was, uh, it was called, that's english and it went out in spain and we were just making the green screen sets for them and then the exec producer on that series said, oh, you know, do you know anything about comedy? I was like, yeah, you know actually. Yeah, quite, I mean, dear God, it's about the only thing I do know properly. And he said, oh, okay, well, this is a comedy series. And I said, well, have you got a director for it? And well, this is a comedy series. And I said, well, have you got a director for it? And he said, oh, have you ever shot multicam studio stuff before? Then, from a gallery, I was like, yeah, yeah, unfortunately, you know, he'd seen BBC on my CV, so he sort of made the middle bit up. The trouble is, I needed to make the middle bit up as well because we were shooting in like three weeks time. And he said, oh, okay, well, in that case, if you're making the sets, you know fine, yeah, you, you shoot the, you shoot the series, um, or you shoot the sketch series.
Speaker 1:And I rang my mate, nick Wood, and Nick Wood, uh, was a BAFTA winning director who was working on Two Pints of Lager and a Packet of Crisps at the time. Um, and I said, nick, nick, I've, I've, I've done something bad here. I've committed to doing something that I have no idea how, how you do, what you do I don't know any of the processes um, and he was very generous, um, and he, he said immediately well, actually I'm still shooting two pints at the moment and I forget which series it was. Um, he said why don't you come in and shadow me, uh, on the production? So I went and shadowed, uh, nick, on two pints of lager.
Speaker 1:Um, for you know, I think I was any. I think I could only do one episode before I went into my own production, but yeah, so I had to. He sent me home with the camera notes and the camera diagrams of Two Pints of Lager that we were shooting on. He said just read that, learn that and transfer it into your own production. You'll be fine. And that is what I did.
Speaker 2:And I worked. There are two episodes one of Two Pints of Lager and a Packet of Crisps and another programme that look identical in terms of the camera settings.
Speaker 1:Yeah, they're remarkably similar.
Speaker 1:It was different content. Yeah, slightly different content. It was one of those, really, but it was a really interesting process. So I worked on that show for five years, um, and so that was.
Speaker 1:You know, that was one of those godsends where you go yeah, okay, fine, I said yes to something and and rather than blag it, you know I will always take something on say yes, we'll do it, and then I'll work out how to do it. Um, and that has I don't think that's ever properly caught me out. Um, it's been close a few times and I'm and I remember once, uh, once I uh I had to tell somebody that I didn't know what I was doing, and they were then very supportive, um, because you know we just got on, but that you know that that is the closest, but those are the. You know, that's the route where, which I've, I've always taken, which is kind of right, let's, do you want to do that? Yes, you do, right, you have to learn it at some point.
Speaker 1:These opportunities are not going to come around again. So we'll just say yes, um, and that that is again exactly the same as how deep fusion films came about in. In truth, yeah, um, that that approach to working is exactly how deep fusion happened and again, the very, very accidental way that I became the, you know, the, the trusted face of ai within media. Uh, in 22 23, which was off the back of uh, jerry anderson, a life uncharted, where we suddenly decided we would put a deep fake in. In 22, 23, which was off the back of Jerry Anderson, life Uncharted, where we suddenly decided we would put a deepfake in Jerry Anderson Life Uncharted. We were contractually obliged to do it after we suggested we could, but in truth not one of us had a clue how to make a deepfake.
Speaker 2:Yeah, let's get into that, because obviously we're talking pretty much present day make a deep fake, uh. So we yeah, let's get, let's get into that, because that's obviously, you know, we're talking pretty much present day. Um, you know, you are the, the ceo co-founder of deep fusion films. Uh, as I mentioned at the beginning, there is a a very strong focus on the ethical use of ai and and you know you've developed policies around that and publish them. And just how do I mean? You hinted at slightly how it started, but how do the two go together, from a sort of creative film production company and then and then ai and and being sort of almost an expert well, you probably are an expert in ai policy they don't seem naturally good to go together. How did that come about?
Speaker 1:uh, no, you're right, they. They don't necessarily seem natural bedfellows, do they? Um, especially given how much sort of furor there is around ai and media generally, and certainly ai and creativity and the difficult relationship between them. Um, okay, so, so what happened is when, uh, so I was working at a company called the format factory, um, who were uh I worked for for a number of years, um, and one of the productions we we were uh pitching was one called jerry anderson, the life uncharted, and this was about the thunderbirds pioneer, uh, jerry anderson and jamie he's uh, one of his uh sons, um said that there was a, you know, an untapped audio archive of something like 36 hours of unheard audio, um of jerry talking to his biographers, and it was deeply personal and it had never been explored before, um, and so we said, yeah, it seems like the absolute perfect basis for a documentary. It's wonderful. And then we said, okay, so what archive are we going to cover this with? And jb said, but I don't really have any um, or certainly not enough that was going to be able to cover the dock. And so we said, oh, that's that's. You know, that's going to be a challenge. And so let's imagine that we have this conversation on the Friday.
Speaker 1:On the Monday morning the deep fake Tom Cruise thing came out on TikTok, and Monday afternoon I rang Jamie and said hey, hey, jamie, how do you fancy resurrecting your dead dad? Um, and that you know that that was pretty much verbatim, because you know these conversations are going to be awkward, so why not break them up with some low-level humor? Um, and he, jamie, responded very positively to it. He said well, do you do you know how to do it? I said no, but you know how hard can it be? Um, it turns out very hard. Uh, you know, we, we actually weren't so, we, so we went into pre-production in april, we ran some tests in april and we worked.
Speaker 1:It wasn't until the end of october before we had a working deep fake that actually was able to, to actually fool jerry's daughter, uh, linda, uh, to think that it was footage, that that actually existed. Um, that has never seen before. Um, and we, we didn't intentionally fool her, just to say that, you know, I mean, I did say before presenting her with a one minute clip of jerry talking jerry, in inverted commas. Um, I said look, you know, this is not your dad, this is not an interview he ever gave on camera. It is his voice, but it's not the visuals, and she watched it and you know there was a pause and she went. So when did dad film that? So you know, even though she was told it was not, she couldn't believe it. It wasn't um. So that that was the time that we kind of went, okay, we've got this right. And then then, of course, we have to deal with the presentation.
Speaker 1:So jerry um, jerry was was a genius in many, many ways, um, but one of the things that he was probably best at, uh, was falling out with people. He was incredibly good at falling out with people, uh, be that a personal relationship or a business relationship. He would find a way to ruin it, um, and I'm being perhaps a little bit glib, but but it is. You know, it is broadly true, um. And so what we realized was the fact that the documentary was placing jerry in in quite a positive light, um, and that a lot of people would not necessarily um sort of align with with the way that he was being presented in in some ways, even though they were all his own words, um, and so what we needed to make sure was that um was the people who were watching. You know, once they knew that there was ai involved, there would be the opportunity for people to say, ah, jamie's making a documentary about his dad. He comes out really well. It's all with ai. You know, it's all a load of rubbish. They've just twisted things to make make him be positive. It can be seen in a positive light, um.
Speaker 1:And so we said, you know, this is really ropey. We need to absolutely make sure that we write a framework that allows for um, for this um sort of for this documentary to exist and for people to not be able to criticize the production techniques. So we wrote an ethical framework for things we would and wouldn't do, and that involved so the, the audio quality, uh, of the tapes was really ropey. And we got in touch with a company called respeacher. Respeacher have now won plenty of emmys, but at the time and I think they, they, I think they've even won an oscar now, uh, but at the time nobody had really heard of respeacher.
Speaker 1:They were just the only company we could find who were capable of recreating the voice of somebody, um, and making it, you know, um, authentic or as authentic as possible. We thought, right, so we could recreate the voice and so, therefore, clean the tapes up that way. This is before adobe podcast was a thing, and adobe podcast would have been able to clean these tapes up, no end, um. But we didn't have that option, so we we would have only been able to um, you know, recreate them. And we said no, you know what. You know the pauses, they're all really important, they're integral. You know, there's as much said in a pause as there is in a line. So if we, if we get the intonation wrong or something, it changes the meaning. No, we need to be transparent.
Speaker 1:So, rather than change the audio, we created the visuals. So all of the segments with Jerry talking are all placed in TVs so that we contextualize the crap audio rather than recreate or try and clean up the audio. And we ran processes like that in order to make sure that the audience could trust what they were seeing or hearing. And we ran title cards at the top so everybody could understand that some of the visuals were AI generated, but all of the audio was authentic. So we placed these sort of frameworks in and when the film came out, it was universally accepted that it was a really great use of AI Russell T Davies dropped me a message on on Instagram.
Speaker 1:I didn't know Russell at the time, still can't claim to know very well now, but you know I know him better than not at all. So you know the use of deep fake was inspired it's. You know it absolutely changes what documentary can do and and so on. And pretty quickly after that I became the face of AI for BBC. So every time there was an AI story, so Deep Fake Neighbor Wars came out on ITVX so BBC had me on all the time talking about the ethics of AI. The SAG-AFTRA strikes.
Speaker 1:Yeah they rolled me out again. I was on BBC radio, I was all over the place, and then that is when, and because I was the only known face at that point, or one of the only known faces, that meant that all the policy writers sort of said well, actually you seem to know what you're talking about, can you come in and help write our policy? And you know policy then turned into okay, can you advise on some legislation, can you advise on copyright? Can you advise on this? And so over the last five years, very accidentally and very incrementally, um, my knowledge base within this area has has escalated. We set up deep fusion to take advantage of the fact that ai was coming and that if we had written the policies and were largely trusted as being the ethical faces of AI, then when the production capabilities came around, we would be in the perfect place to be at the forefront of that. And so we set up Deep Fusion in 2023, which was probably 12 months early, but it did give us time to set all the foundations in place, get all the relationships going. You know, really leaning into ai generated or ai enhanced productions, or even where, you know, the machinations of a show or a format can be underpinned by ai. Uh, looking at you know how data can be translated into visuals or or what have you, how it can impact the way that that we live um and and therefore represent those changes on screen in some way.
Speaker 1:But really, the big tipping point was virtually Parkinson, which was a massive leap of faith for all involved, but that was as far as I can see.
Speaker 1:There weren't any other commercially successful AI productions, certainly not in the UK and I would go so far as to say not really globally.
Speaker 1:And so, after Virtually Parkinson went out, we were able to prove that this was a success, and a success in many, many different, by many different metrics. It was a success in many, many different, you know, by many different metrics. It was a success and that changed the perception of how AI can work within production and indeed that we were the trusted face to do it. Because, again, you know, virtually Parkinson, despite the fact that it was noisy and a lot of people had very, you know, differing opinions on whether or not it should even be allowed to be a thing or whether it was worrying or concerning um, actually a lot of you know it, it did a lot of good um, because we were able to answer all those guardrail questions yeah, I mean, I want, I do want to get into that um, both in terms of how you created it and we'll touch on what you called squawk um and then the the kind of the policy framework for workflows that you came up with as well afterwards.
Speaker 2:But when I first met you and I first saw you, this was at the podcast show in London and you demoed and you spoke about Virtually Parkinson, together with Mike Parkinson, the son of Michael Parkinson, your business partners obviously as well, and I mean we still talk about it here in the office today. It was that demo was just gobsmacking, it was super interesting, but it was just unbelievable and because you'd explained what it was and how it worked and that it wasn't scripted and so on, we just couldn't believe it and it still sticks in our memory today and it's obviously why I reached out to you to come and talk to us here today as well. But just, I don't want to show a clip, but just before we do, perhaps can you just explain for those who don't know what Virtually Parkinson is, what is it, how can people picture it, it, and then we'll show a clip as well.
Speaker 1:So Virtually Parkinson is at its essence is a chat show that is hosted by an AI representation of Sir Michael Parkinson, so it uses Sir Michael Parkinson's voice, um, and it is uh, all of the unscripted conversation um is driven, uh on the ai side, by an ai brain, uh, that is trained on sir michael parkinson's interviewing techniques, uh, so, and that is from the itv tapes, or what we call the itv tapes uh, which is the stuff that parkinson productions owns um and ran on on itv uh.
Speaker 1:And so what we had was we had a live celebrity guest uh in the studio and we had, we built a piece of software which, as you mentioned, was called squawk, which allowed for a guest to have a conversation and be interviewed by AI Parkey, and that was, in essence, what it was. And then, either side of that conversation, there was a wraparound element which was a discussion about the leaps in advancements we'd made in the model, in the AI Parkey model, and the improvements we'd made to try and make AI Parkey do a better job of interviewing each week, and what that did. And that sort of wraparound element was done between Mike, jamie and myself, and every week we would end it by saying, right, okay, how do we improve? And it was also exploring the celebrity guests' relationship with AI. You know on either side, so we would always do a debrief with them that said, okay, so how did you find that? What was the most interesting? What was the bit you didn't like? And it was a fascinating journey into the exploration of our relationship with AI what we're comfortable with, what we think, the potential upsides and downsides are A new podcast featuring an AI recreation
Speaker 3:of the late Sir Michael.
Speaker 4:Parkinson's voice. Good morning to listeners of the Today programme. This is Michael.
Speaker 3:Parkinson or at least a version of me in this rather curious AI form.
Speaker 1:If you didn't know that that was AI, you would never have thought it wasn't him. He's going to be interviewing real guests.
Speaker 2:I received this following message it would be an absolute pleasure to have you as a guest on my new podcast called.
Speaker 1:Virtually Parkinson. On behalf of humanity, I'm going to turn this one down.
Speaker 4:Ladies and gentlemen, please welcome my first guest on Virtually Parkinson, Jason Derulo.
Speaker 1:Good morning to listeners of the Today program.
Speaker 2:How did your family respond to your dreams? Did they nurture this rather curious AI form, you know?
Speaker 4:without those same sacrifices that he made. You know I wouldn't be here today.
Speaker 3:Wow, I received this following message it would be an absolute pleasure to have you as a guest on my new podcast called Virtually Parkinson by part of humanity. I'm going to turn this one down.
Speaker 2:I'm curious about how you found this interview.
Speaker 3:The way that you were able to ask the questions felt like you knew me.
Speaker 1:How did your?
Speaker 4:family respond to your dreams. Did they nurture that creative spirit in you? Great?
Speaker 3:question. You know, without those same sacrifices that he made, know, I wouldn't be here today. Wow, uh, that took a left turn. I fractured my c2 vertebrae. I had to make a way out. There's also a part of me that doesn't want my son to feel the way that I felt yeah, I mean I love this video.
Speaker 2:Obviously it's edited together to kind of show various aspects and promote the episode, but I think the first thing that struck me was just the actual, also self-awareness of virtual parkinson to know that he's in this curious format, as he puts it, and and I remember you saying at the time when, when I first saw you on at the podcast show uh, this is not scripted, you're not feeding into the voice to tell it what to say. It generates that through through its ai right.
Speaker 1:Yes, yeah, so it listens to what you say, um, it analyzes what you say and it comes up the question based on its you know, based on its learning of, of you know, previous uh interviews that michael gave.
Speaker 2:Sir michael gave yeah and yeah and um, obviously, that that process can go in, therefore, any direction, because you're not telling the ai what to ask and what, what to say, so it can go in certain directions. Um, I mean, for anybody who hasn't watched an episode before, they are, uh, obviously available on on youtube as well and take a look there again. As I say, absolutely fascinating, but just to get an impression of not just the, the type of questions being asked, it's the camera and the tone and the voice, it's the, the short breaks in between where he sort of you know that style where he slightly readjusts himself and repositions the question a little bit, and it is just so accurate. So how did that develop over time? And obviously, you mentioned I mentioned Squawk as well.
Speaker 1:Tell us a bit more about how that actually was made possible.
Speaker 1:Yeah, I mean that was essentially it. We took 100 hours of ITV interviews to work out the interview style that we fed into the model. The voice was trained via 11 labs and that was essentially. I mean, the training of the voice was actually the easiest bit. I mean I really expected that to be a very, very difficult process, but it you know actually, that that voice that we have there is the first version of that that I made. You know, as soon as I made it, it worked. We had a horrible moment when Eleven Labs updated its internal systems and it didn't work for a week. That was terrifying because we couldn't remake it. So to some degree it's a bit of a one-off, so we are very fortunate in that regard. So we are very fortunate in that regard. I mean, I've not tried to remake it in the last eight months, so it may well be that I could remake it now, but at the time I couldn't, and yeah. So basically, the way that Squawk works is that you take in a voice, it's fed into the model.
Speaker 1:Ai Parkey will analyze what's been said. Go back to the research notes that we provide on the guest. Correlate what's been said, with the research notes and generate a question and come back with it. So there is, as you will have seen at the podcast show, there is in real life life. There is a delay of around about two to three seconds between you know, uh answer and next question, uh, and. But those were the only edits we made, uh, in the in the podcast. We never took anything out, um, because we one of the one of the scariest and yet one of the loveliest things about that show was that I would press go on the software and I would have no idea what conversation was going to come next, and neither did the guests. They knew that we had input on the research nodes, but could have gone anywhere. Um.
Speaker 1:What generally happened, um, was that it turned into a bit of a therapy session, because normally when you, you know, when you're being interviewed by you know, by another human, you as a, as an interviewee, respond to micro expressions, uh from the interviewer and you're aware that there is an agenda to some degree. So, as an interviewee, you're always trying to please the interviewer and give them the information that they want. So there's, you know, there's almost a requirement for that, for that to take place. Um, you know, for instance, you, you are nodding at me now in in that micro expression that suggests yeah, no, I understand, keep going, I want you to keep talking about this. And similarly, I know that you know, if you were to trying to interject, I'd try and hurry up what I'm saying so I could finish. But what happens, uh, with um ai parky is that there are no micro expressions because he's not in the room. Um, he's literally just listening to the words. And so what happens is you, you get like a stream of consciousness, and that stream of consciousness is not something that is hindered, then, by an agenda. So, uh, ai parky is not going to go. Oh god, get on with it. You know, fine, I don't want to talk about this element of your life. I want to talk about the things that you know that I want to talk about. I'm going to get you back on track so that doesn't happen.
Speaker 1:But what does happen is that, you know, let's use jill scott as an example um, you know, jill scott uh started talking about the fact that she, uh, she writes poetry, um, and so ai parky uh said oh, can I hear some poetry? You know, I mean, you know, my ai parky was slightly more verbose than that, as he was in all the answers to all all these questions were all always slightly longer than they would ever need to be. But, you know, he, he said you know, can you read the poetry? And so she did, um, and she has never read her poetry to anybody ever. Because she feels judged by you know, or the potential for somebody to judge her for them. Um, but there was no judgment in the room because there was no body in the room. So she read her poetry uh, stephen moffat, um, dr who writer.
Speaker 1:He started talking about childhood bullying. Uh, you know, and how that changed, who he was growing up and why he became a writer. He doesn't talk about that, you know. No, no human is ever interested in that. So he says, um, because they always wanted to move it back to Doctor who or whatever. Ai Parkey wanted to understand the human context of being bullied, you know.
Speaker 1:So you end up in this very, very, very interesting, you know, deep conversation that we did sort of expect because when we did our early tests, we actually did it with Nick Briggs, who's a friend of the show, so to speak. Nick Briggs is the voice of the Daleks, among many other things, and in fact, actually the podcast show Lev. It was Nick Briggs, who we brought on stage with us then as well Again, useful guy to have, so, yeah, so Nick was the first to do it and we fed in, you know, a couple of pages of research notes and what actually happened was that nick started to reveal very personal information that he completely didn't expect to, because there's almost a requirement to just keep talking. You know the question arrives and you don't really have time to deal with. Well, what question is this? You just then go on to the next one, and so it became almost the most.
Speaker 1:Who was it? Chris McCausland, when we interviewed him, said it was the most personal podcast or personal interview he'd ever given. Joe Dracini said it was the only ever podcast he would ever listen back to. Uh, because, you know, because he was just so intrigued by it all and what he talked about. Uh, he ended up talking about pantomime and the relationship and and a and a meeting he had with an old woman at stage door that really affected him or really touched him not not affected, but let's say um, and he said I'd never expected to be talking about that, I expected us to be talking about this, but it was because you know what you give into it is what you get out of it.
Speaker 2:Yeah, but isn't. Isn't that fascinating, though, because that's perhaps from different angles, but that is. You know what sir michael's style was like as well. He was able to bring out conversations and aspects and things that nobody else could or that you wouldn't expect in in that particular format.
Speaker 1:But it's well and and and this is exactly what we said to mike in the early stages um was that, you know, once we'd worked out that there was something in this format, what I said to mike was you know, because it was Mike that approached us to do a podcast, not the other way around, and I rejected the first idea that he had, and we sort of wrote it off and shook hands and said well, you know, if anything comes up, we'll come back to you. And so, when something did come up, I said look, actually, mike, your dad revolutionized the way that the celebrity interviews were held. You know, it became much more relaxed. It wasn't over a table or anything else, it was a really relaxed environment and the style of questioning was very, very different. It kind of makes sense, therefore, that once again, his voice and his interviewing style is again presented in this way.
Speaker 2:Yeah, it's fascinating and obviously AI is a is a very top, very topical topic. It's very popular at the moment. It will, yeah, it is becoming part of of everyday life, but it's also relatively new, I guess, in the grand scheme of things. So we mentioned you're you're very active in in in the area of ethical AI. How do you, seeing as you're working with it, seeing as you've worked a lot with it and advised people on it, what excites you most about AI and what makes you most concerned about AI?
Speaker 1:Things. Yeah, I mean, I think that those, those two questions, if you like, are a little bit like a ball of wool in my head. Um, because they are so intertwined, um, that it becomes a little bit messy. Uh, because all of the answers of as to what excite me also have areas of concern, and so there isn't necessarily a straightforward answer. So I'm capable of giving you an answer, but I'm very aware that, as I say one thing, somebody will have quite a visceral response or a visceral reaction to what I say, and I hope that I can then catch that visceral response and provide an answer to it. So, bear in mind I'm going to talk about this through the lens of media production, if you like. I think if you start looking at AI within healthcare, for instance, I think there are potentially a huge amount more upsides than there are downsides Within AI in the creative industries it all becomes a little bit messier.
Speaker 1:Uh, because generative ai was, it is obviously created uh, using what is referred to as data sets and data sets as uh, somebody once shouted at me on instagram instagram of all places. Instagram's a nice place, it's not twitter or x. Uh, it got shout stuff on instagram. I was very surprised, um, anyway, uh, there's a. It's not data sets, it's creative work, it's somebody's creative work. I was like, yeah, okay, I won't refer to it as a data set per se, but as far as ai training is concerned, it is a data set. But yes, you're quite right, it is somebody's life's work.
Speaker 1:Um, should, should AI data sets be, you know, not scraped and indeed be paid for? Um, and I think the straightforward answer to that is yes, I would absolutely advocate for that. You know, every single day, every hour, every minute of every day. In truth, that is unlikely to happen, but, yes, I would always advocate for that, because, at the moment, what we have is we have AI data companies, ai software companies scraping materials in order to create efficiencies within the creative process, and so they are, they are selling, they are, they're scraping data and then selling it back to the, to the creatives who they've just nicked the data from in the first place. So it's kind of like being resold your own creativity, and that is very messy, um, and it's. You know, it is not comfortable and I'm I'm working a lot with the um, with the costar national lab, um, and, and so therefore, feeding into uh dcms and and dc around, how we can make a sustainable and responsible model for ai software developers, um, and how we can get remuneration for creatives and creative work, um, and how we can respect copyright at the same time.
Speaker 1:So we acknowledge that there are massive challenges over there, but what is frequently not as well acknowledged in the same breath is that the commercial creative sector, by which I'm not necessarily referring to digital production, so TikTok, socials, all the rest of it, um, youtube kind of his heart has got a foot in the door there, but through a different, different kind of way. So when I talk about commercial creative opportunities, I am talking about traditional tv commissioning or street or or anything else, um, or branded content, because the TV and film sector has almost created a bit of a citadel for itself over the last 40, 50 years, which is that frequently you will hear a producer say I can't possibly make that show for anything less than £150,000. And that show being an hour long and will be something that goes out on Channel 4. And you go how have we got to a point where an hour's worth of content can't be made for anything less than £150,000? You know, you could potentially buy three houses in Barrow and Finesse for the price of making one TV program for one episode, and that feels completely disproportionate. So we are, you know, and at the same time, you then have digital creators who are quite capable of making programs, you know, even without using AI, they are quite capable of making, you know, programs that are watched by younger generations on social media platforms that last about the same amount of time, and they make it for the same price.
Speaker 1:As you know, tuba smarties, um, and you can, I love smarties, I would buy quite a lot of those, um, anyway, um, so you know, there is this disconnect between what the traditional industry is doing and, of course, then you have the problem that for the last 10 years, the traditional, say the traditional industry has been going oh, we need to attract younger viewers, so, yeah, okay, but younger viewers are all going online, you know. And and then we have to ask why are they going online? And the reason is generally because there is, you know, it is much cheaper to make content for online, and if it is cheaper, you can take more risks. So you've got, you know, one end, you've got 150,000 pounds an hour, which has to be risk averse because it's so blimmin expensive. So you can't take a risk on it. And then on socials, where you go, well, it's all riskier content. So why wouldn't you? Why wouldn't you? You know, as a you know, as it's always been, the nature of things is, is that actually, boundaries have been pushed by younger generations, which means that that's the kind of content they want to see. So we have this imbalance now, where AI is a massive efficiency generator. So, whether that's creating content, as in B roll, or getting you know, new voiceovers done, or even if it's just editorial or research, you can.
Speaker 1:I mean, I did. I did a short documentary for a red carpet premiere at Leicester Square in October 2023. And I did the entire pre production using using AI. Didn't do the actual production, but I did the entire pre-production using AI in order to do, because it was a two-week turnaround and the budget was very low, and so I made it in AI and then I remade it in real life so that I could get the commissioner to sign it off before I went into production, so that I shot what I needed in order to get the you know, and it varied very slightly, but ultimately it was all there.
Speaker 1:Um, and that wouldn't you know, without ai, I would probably have had a lesser quality output and it would have taken many more man hours, so the chances are that actually the production would never have taken place, so we'd have lost out the work. And that's the balance is that we now have to balance the idea of whether or not actually is it better to be using AI and putting more people into work on more productions than it is to not use ai and not make as many productions, because we we have a real problem that you know through and I don't know how away you are of this, you know, but through 2023 and 2024, there was basically a commissioning clamp down for channel four and channel five. You know, and and and also you know, to a lesser degree, you know, bbc and itv, um, and you ended up in this scenario where nothing was being made in the unscripted space, so you had this mass exodus of some of the the. You know this is happening in hollywood as well. I got an uber in la? Uh last year, um, and the uber driver was a hollywood producer who'd had a successful run of hollywood films out um. You know, and and so you're looking at this massive shutdown because there's nothing happening, because ad spend is down, uh, so they can't afford to make anything. So, but that is, they can't afford to make anything at what could be known as citadel levels. You know, no, I can't make it. You know I'm going to make one series of something. You know, scripted content now is often two or three million pounds an hour, you know, and they go on channels that nobody's ever heard of, and that's how expensive it is, and that's unrealistic.
Speaker 1:We now have to be looking at new ways of doing things. That gets people back into work. But the question is, are we going to get people back into work and pay them the you know, the fees that they were able to request or command five years ago? Well, realistically, no, we're not, because, you know, even if we talk about the efficiencies, we go, well, actually it's a different. It's a different time. We're in a very different time now, uh, and so we're going to have to look at what the next model of of commissioning and content creation is, and that is deeply uncomfortable for an awful lot of people, because it means that some of their jobs are going to be reduced to a point that you know.
Speaker 1:You now need to be doing an awful lot more of it to make the same kind of money, either because of the speed it takes to shoot that stuff in the first place, or because the work is just not there. Or you need to accept a lower rate because ad spend is down, because advertisers if you're talking digital content advertisers now recognize that it doesn't actually have to cost them the same amount as it cost them five years ago. So they go right. Well, the budgets are smaller. How you fill that gap is then up to you. You. So they go right. Well, the budgets are smaller. How you fill that gap is then up to you. You pitch for the work. Well, the industry has been in a downturn for two years. Everybody is desperate for work. So you sort of look at right, well, ethics. If you are desperate and you need to feed your family, ethics can do one, you know, because everybody fights themselves.
Speaker 1:What I'm trying to advocate for is a middle ground where we recognize the efficiencies that AI generates, but we also put safeguards in place to make sure that creatives' work is respected and that the ongoing sector is sustainable and responsible. So we're kind of looking at anything that's happened prior to today. Well, that's kind of gone. You know, we can't change what open ai did with sora. We can't change mid journey. Uh, let the let the courts deal with that. You know, for anybody forward to take mid journey to court. Or you know, um stability ai was taken to court by getty, you know, let let the corporates deal with that and let let the legal systems play out in the. In the interim, every individual creator needs to be aware of what, the, what terms and conditions they're signing up to when they start using ai software. Uh, because most of the most of the ai generators so.
Speaker 1:So let's use Midjourney as an example. If you upload an image to Midjourney to restyle it or to adapt it in some way, or just use it, especially now that they're doing generative video as well, by agreeing to the terms and conditions, you are saying that any material you upload, you own the copyright or the necessary rights to upload it that allow for the service to train their models on it. And bearing in mind that mid journey is currently being sued by disney because mid journey's original training data was supposedly trained on Disney material, well, if Disney are suing Midjourney, it's only a step away that Disney will then ask for the details of anybody that's uploaded content to Midjourney In order to work out? Actually, has anybody else been uploading our content, claiming that they have the rights to do so? So you are putting a level of personal risk on that because it goes down to the individual, doesn't? It's not corporate risk, it's it's individual risk, so we need to be careful about that now. In truth, the chances of that ever actually happening, the chances of Disney going after one individual, is just not worth anybody's legal time, but the risk is there, so we need to be careful about that.
Speaker 1:Those are the things that we're looking at at the moment and going right. There needs to be new, safe commercial models around AI, training, data and greater levels of transparency around copyright, but it's a very, very difficult process. But in terms of the very quick answer is there are huge opportunities to create a new media creation model, but the corporate support or the, the infrastructure and, um, the copyright situations need to be addressed in order to give uh, creatives the security they need in order to actually use them, and in the middle of that will be everybody just giving it a go and hoping they get away with it. Um, we, we aren't doing that.
Speaker 1:We have spent, so we haven't lent into any generative content that isn't licensed. Uh, virtually, parkinson is all licensed and all paid for, um. But so we've spent 12 months, uh, looking at the various generative models and working out a pipeline that allows for us to retain all the copyrights that we need to we're not giving away licenses where where we don't own them, and we're able to make generative content, um, but that has taken 12 months. But we have now finally done that in a way that has been signed off by the lawyers and we've got insurance against it. So we're now, you know, we are now legally entitled to make generative content using our particular workflow. Um, but not everybody has one of those. They just go for. Well, what's the easiest, what's the quickest?
Speaker 2:thing I can do, yeah, and, and I mean that, that's, that's also to your knowledge, a world first right first, first workflow to be, in short, to my knowledge, yeah yeah, you, yeah, I mean it's very difficult to difficult to actually sort of, you know, nail that to a wall.
Speaker 1:But yes, to my knowledge, that is the first.
Speaker 2:You're not the millionth to do it. You know you're at the forefront of that and clearly we are yeah.
Speaker 2:And so I suppose, on various levels, with that and other things, you are in a leadership position and I'm just curious, kind of to expand this a bit further out again. What you know this is obviously the leadership series of our podcast what have you learned over? You know all the things you've done, all the projects you've worked on over your career. What have you learned about leadership and what have you learned about being a leader along the way? What can you share with us on that? Yeah.
Speaker 1:I think it's very, very important to surround yourself by people that you trust, and trust can obviously be a very broad term. The people that I will trust with my finances are very different people than I would trust with you know their creative opinions or their business opinions. But you know, I think you have to, I think you have to be very selective over who those people are that you trust, because there are lots of people who take advantage, but one of the and this, this takes a very long time but one of the most important people that you need to be able to trust is yourself, is that you need to trust your own judgment, um, and in you know god, I mean even even over the last you know, 12, 18 months, two years, three years, um, I've had to. I I've sort of had to tell myself off for trusting people that I really shouldn't have done, um, and actually my gut instinct would have served me better. But you take risks. I think we all have to take risks. You have to give people the benefit of the doubt in order to earn that trust in the first place. But certainly trust in one's gut, important, um. But also don't don't ignore culture, um, because a lot of.
Speaker 1:There are some brilliant people out there, but what you need to build is a is a team that can talk to each other in a respectful and in a powerful way, and sometimes that means that you don't always hire the person who looks best on paper. You have to hire the person who you think will fit into your team, and deciding who is going to fit into your team is something that you know is is paramount, and you have to build them around the, the type of you know the type of entity you want to build. I I want deep fusion and and jamie uh, who was my co-founder, is my co-founder, sorry and um and andrew uh. We wanted to build a team that allowed for honesty and fluidity of thought, so that meant that they were everybody is able to challenge and everybody is is allowed to express an opinion. Um, and as long as those opinions are done respectfully and as long as you know we we laugh a huge amount of the time at deep fusion because we just enjoy the.
Speaker 1:You know the people we're surrounded by, and I think that you know if you then end up hiring somebody who doesn't necessarily have the same kind of outlook that you do, um, that's going to change the dynamics. Um, so you, you need to be responsive and alive to different personality types and make sure that actually, the people you're bringing in, uh, are all going to work together cohesively, and the problem is the fact that the larger the company becomes, the the more difficult it is that you're going to work together cohesively, and the problem is the fact that the larger the company becomes, the more difficult it is that you're going to have those interpersonal skills that are going to directly align with everybody. But that is certainly something to be mindful of. But, yeah, those are my key areas.
Speaker 2:We started off with a question about storytelling and your passion about it. I want to come back to that, perhaps for this question. If, in 20 years from now, somebody were to tell the Ben Fields story, what would you hope it would be?
Speaker 1:I would like somebody to recognize that I may have made many, many wrong choices, but I made them all with the best of intentions and I'm very cognisant of my faults. But I would always try to make sure that everybody around me was supported in making their own decisions, even if I was driving something where I made a misstep. I think broadly, I'm always trying to do something for the betterment not just of the company, but right now I'm doing an awful lot that I feel is for the betterment of just of the company, but right now I'm doing an awful lot that I feel is for the betterment of the sector and the industry. Um, and others may disagree and others, may you know, may see me working in a very different way, um, but I truthfully, I think I'm I'm genuinely trying to steer a path that is very, very challenging because of the amount of new information and the amount of disruption that it's causing.
Speaker 1:You know, there will be a wake of destruction around AI for a great number of people and it will take a lot to reset that balance, and it will take a number of years probably to get everything back into a status quo, which, which means there's going to be a lot of upset people, um, but you kind of need to look at what the end goal is, which is that people are still employed and that there is still an industry for people to be employed in, and if that means that some people are left alone way, and those are the people who don't wish to change or don't wish to see the world in a different way, well, the world is going to be a different way, but right now we have to shape which way that is, and I'm quite capable of getting that wrong, but my intentions are honest and that's how I would like to be perceived, I think.
Speaker 2:Thank you for sharing that, and we're not quite done yet, but we still have the virtually anything goes question. For anybody new to this podcast, it's where I turn control over to Ben. He can ask me any question he likes. I don't know what it is, and it doesn't even have to be related to what we spoke about today. I do have to answer it, though, even though I don't know what it is, and it doesn't even have to be related to what we spoke about today. Uh, I do have to answer it though, uh, even though I don't know what it is. Uh, the only caveat is that and that's my safety net once I've given my answer, then ben has to answer the same question, answer his own question as well. Uh, so, ben, this is, um, this is where I turn it over to you, and, um, let me know what the virtually anything goes question is the virtually anything question for you, lev, is what's the belief you hold about the world that you're most afraid people would misunderstand?
Speaker 2:wow, you put some thought into this, didn't you? I did yeah yeah.
Speaker 1:If not, what's your favorite color? That would be a rubbish.
Speaker 2:That would be rubbish. It's blue, so yeah, I think I mean I have a faith, I have a Christian faith, and it's for me, it's very strong, it makes perfect sense. But I'm also very aware that not everybody shares that opinion and I think in a you-do-you world, I think everybody chooses their own path and that's their right to do so. I think with that there is probably, I think, a worry from my side that perhaps people don't share my belief. I'm not forcing that on anybody, but I do see the positive it does in my life so, and many other people's lives as well. So I think from that perspective, that's probably the one I'd have to say, that's the strongest I feel about that kind of yeah. If it's a worry or concern, but for me it's the most obvious one, I would say yeah.
Speaker 1:Do you think that's because it's about judgment? Do you think other people judge you for what? The stereotype of what a Christian belief might be?
Speaker 2:you for for what? The stereotype of of what a christian belief might be. No, I, when I, well, so I've, um, I haven't always become. I became a christian when I was, um, when was it? 2007, I think. So, um, I didn't grow up a christian, so I made that choice as an adult.
Speaker 2:And but even back then, I think, when I was younger, yes, probably I would have been more concerned about that. Now I've, I've just turned 50 and you know I've, I can, I can wear, you know, red trousers and red socks and not be concerned. You get to that age where you don't really care what people think. So you know it's, it's. So, in that sense, probably not so much anymore, but but there is an element of that, I guess. Yeah, um, but also, I think, I think there is an element of not not wanting to um force.
Speaker 2:You can't force beliefs on people. People have to explore that for themselves. Um, and but then, at the same time, when, same time when you're passionate about something that doesn't have to be the Christian faith it could be anything. You would talk about it, wouldn't you? You would tell people you go to a party or whatever. You tell people about it and you're passionate about it and people think, oh, that's amazing, that sounds awesome, but there isn't that same freedom really for any belief. Really, you know it doesn't have to be christianity, could be anything. Um, from a religious perspective, that that is, I guess, in the way you know, not done so. I think in that sense it's probably not the way of being judged.
Speaker 1:I guess it's just the the way it's handled in society in general so there is always an opinion about it, and those opinions are generally always quite strong or visceral anyway yeah, and it's the same with politics, and it's you know.
Speaker 2:you go to party, you know if you tell them you're a labor supporter or reform supporter, or you know a democrat or republican, wherever you live, there will be a, and it's it's more polarized these days, isn't it? Um, and and I think there's stronger opinions, stronger reactions, but that doesn't mean it's any less important to believe in follow your belief, than it has been Very interesting.
Speaker 1:I could have many theological debates. I really enjoy that as a topic of discussion.
Speaker 2:But now is not the time for that.
Speaker 1:We'll do another episode we'll do another episode, just just theology.
Speaker 2:Yes, please, that'd be excellent yeah, I'll turn it back to you, um just okay just just uh, read the question out again and then um so the question was.
Speaker 1:The question was what's the belief you hold about the world that you're most afraid people would misunderstand?
Speaker 1:Um, and my, my.
Speaker 1:Obviously I've had time to think about the answer as well, so it's a bit of a cheat really, um, but but the belief that that I have is is already one that we've discussed pretty much at great length, which is the idea that, whilst there are many, many conversations around AI use in the world, I think that AI is going to be a huge force for good in the coming years and the coming decades. I think there's a there are lots of difficult questions and difficult conversations to be had, um, and I think that there are always going to be good actors and bad actors, uh, with you know and I'm not referring to my acting career here, um, but you know there were um, you, but there are always going to be people who take everything one way or the other. But overall, artificial intelligence is going to take a while to play out and for the ecosystem to sort of settle down, but at the end of it, it is going to largely be a force for good. The only side where I think that there is a real risk is within military use. I do have a genuine concern around that. But for the majority of normal humanity I think that there will be a lot of resetting done, but I think the world will be a very efficient place afterwards.
Speaker 2:And do you think generally that you know the concerns around job replacement? You see, you know five jobs, the only five jobs that will remain in the world after AI takes hold. You know those kind of sensationalist I mean, are there sensationalist claims? Do you see people's concerns that they might be warranted?
Speaker 1:No, I know, you know, I understand them. Ai is not the first time that a technological advancement has created a shift in the workforce. Um, you know, let's use the media industry as an example and I'm only using this as an example because I can't use other sectors, because I'm an idiot um, but the statistic is something like 40 years ago, 60% of the jobs that are done in the media industry now were not jobs. So we've lost 60% of the jobs that were jobs 40 years ago, and in another 40 years the same will happen again. And in another 40 years the same will happen again. Um, but yet it is still the strongest sector that the uk economy has. You know, with with whatever the numbers were that were banded around in in february that I can no longer remember, but you know as by uh, economic value, the media sector is the second largest sector in the uk.
Speaker 1:Um, and it has undergone many, many technological changes. Yes, this is another big one, but actually, for humanity to roll over and for financial poverty to set in at a widespread level would require a level of human passivity that has never been seen, would require a level of human passivity that has never been seen. So, yes, there will be an upheaval and there'll be a shift, but humans are very, very resilient and very, very good at finding new ways of creating value and I don't mean value in an economic sense, I mean value on a personal sense. They will provide a way everybody will find a way of doing something new that has value for somebody else, because the AI will be a new tool that they can use in that process. So, no, I don't think it will be the great reset that everybody thinks it's going to be, but it is going to be. It is a new tool to use.
Speaker 1:The worry would be is the fact that actually, it's going to take some time for the majority of people to work out how to, how to bed in with that. So you know, there will be a short-term contraction of the labor market, I suspect, but then, when people find out new ways of working with it, you will hire new people. You know we we are as as you. You know, as we've discussed, there's an awful lot of stuff within Deep Fusion that looks like it's AI driven, but yet we are hiring people and growing faster than any other production company that I know at the moment, certainly in the unscripted space, and that is because we need people to do the work that we're doing.
Speaker 1:And, yeah, okay, will that change over time? Yeah, and our needs will change over time. Do the work that we're doing? And yeah, okay, will that change over time? Yeah, and our needs will change over time, but the jobs that we'll need doing are still going to end up being jobs. It's just what tools are they going to use to be able to do it? That will change.
Speaker 2:That's the only difference, one tool in the list of engine and internet and other things when we just adjust as humans and carry on and roll with it. Absolutely well, ben, it's it's been. It's been such a fascinating conversation and, um, I appreciate you, uh, coming on to the show, sharing your thoughts, sharing your story into into leadership, telling us more about how you work, what you do, um, obviously you know seeing that that clip of virtually Parkinson um is as fascinating as when I first saw it demoed to me as well at the podcast show. So I really appreciate you being there, um and and and sharing that with us. So, um, thank you and, of course, thank you to our audience as well for listening.
Speaker 2:If you enjoyed this episode, there are plenty more in this series and previous series and, in future, in upcoming series as well. So, if you liked it, share it with somebody that you think would would like it too. It helps us reach more people and grow more um, but, of course, subscribe so you don't miss anyone, uh, any of the episodes going forward as well. So, ben again, thank you very much and we'll speak to you soon and I'm sure we'll see all our listeners as well, very soon. So, thank you very much and take care thank you for joining us on this podcast.
Speaker 4:We hope you enjoyed it as much as we did for other interesting topics. Go to your favorite podcast platform or watch the video versions on youtube. Just search for the virtually anything goes podcast. See you next time.
