Meet the AI Actress Built by a Performer ft. Eline Van Der Velden (Particle6)

Speaker A: Our this Week in XR podcast is sponsored by our friends at Zapper, the world's leading augmented reality platform and creative studio. With over 11 years of experience working with the world's biggest brands through Zapper Creative Studio. Zapper also has an award winning web AR platform, Zap Works, that lets you create your own mobile AR magic. Finally, check out their Zap box, the most affordable mixed reality headset on the planet. Start creating AR over at Zap Works or talk to them about your next ar project@zapper.com Good morning everybody. I'm Charlie Fink with Ted Chillowitz and Rony abovitz for the aixr podcast. It's Friday, June 26, 2026, show number 297. Great to see you guys as always

Speaker B: from, uh, an interesting week at awe. So, Charlie, I know you have a lot of reporting on that, which is exciting.

Speaker A: So we also have a fantastic guest today, Elaine Vander Velden, who is the founder of Particle 6, an AI company in the UK which trained an AI actress named Tilly Norwood.

Speaker B: Yeah.

Speaker A: And created quite a sensation. Yeah, Tilly was going to join us also, but I don't think it's quite ready. Oh, no one has really solved, no one's really solved the latency problem, I think. And, and they also promised, uh, an exclusive to a much bigger news outlet, so maybe she'll come back with Tilly. But meanwhile, um, Elaine has a lot to say, so let's get to the news real quickly though. AWE was an interesting year. Uh, even though this was the year I joined you two gentlemen in the AWE hall of Fame, I thought it was a little bit of a down year. What did you think, Ted?

Speaker C: Congrats. Congrats, Charlie.

Speaker A: Oh, thank you.

Speaker B: Yeah, congrats. Uh, I don't know if I would characterize it as a down year. Um, I think it was an appropriate year. I think people are sobering and realizing that this industry, while an important and valuable industry, is not like the computer industry. Right. Or not like an industry that's going to take over the world. I think that was sort of a, an impression 10 years ago, 15 years ago. Um, and now it's sort of taken its rightful and important place in the realms of where this form of computing and use cases are. And I think that's actually healthy. I think it's important to recognize where things fit in the world and where they don't fit in the world. And I think, uh, you know, our friends at Meta are realizing that. I think even our friends at Snap are realizing that when launching the spectacles. Um, because I think.

Speaker A: Hang on there, hang on there.

Speaker C: Okay.

Speaker A: X was the big news of the show.

Speaker B: Okay.

Speaker A: 20, 20, $200. Uh, they're standalone. They're a little bit bulky because they're standalone.

Speaker B: Yeah.

Speaker A: And uh, they have a 51 degree field of view, but it's vertical, not horizontal. Uh, which makes it feel much bigger. It was, that's a great innovation. Uh, it's, it's quite bright. They say you could watch a movie. Evan Gaba, uh, a terrific keynote in which he highlighted some features that don't exist on the developer version that we all have seen that they introduced last year. So we'll see. But they're available for pre order. They're on the pricey side. Snap has uh, you know, a billion monthly users, so that's quite a lot of people they can market to. So we'll, we'll see. But uh, I thought it was exciting. They're obviously the only see through glasses that are out there, um, or soon will be out there, uh, after the HoloLens and, and magic Leap left the field. So, uh, it'll be really interesting to see what they're like to use and uh, whether they get some pickup. I, I thought they were shooting for like a hundred thousand units, but other people have said they're actually thinking much like more like 15,000. I don't know.

Speaker B: I think it's that small. Really.

Speaker A: That's, that's what somebody told me. They actually, Dave Lorenzini told me that he got that from an insider.

Speaker B: Huh.

Speaker A: But before he said that I was thinking like 60, 90,000, something like that.

Speaker B: I mean, I guess both thesises are potentially correct, um, that at that price point it does fit into what I would kind of refer to as the gray zone. Right. Where it's got a lot of interest, uh, across uh, developer sectors. Potentially some enterprise for its use case. Although it doesn't really feel to be like targeted toward enterprise use cases, it feels more targeted towards entertainment and light sort of duty wear use cases. So it sort of doesn't kind of fit anywhere. Um, and that's, that's a little bit of a, of a problem. And so maybe that is kind of their target. I mean I can see it being used for LBE for theme park, you know, location based use cases and very bespoke industrialized use cases where someone has figured out that this is actually the best tool for the job. Not the Google Glass industrial version or the various industrialized versions of headsets.

Speaker C: So I'll give You my theory on this, I think Evan has the same bug I had which was the all day everyday wearable that's like consumer and media and fun and games and not like just text but like the full, you know, digital things living in your world and characters running around and world recognition and computer and it pretty much does. It's like a repeat of what we did at Magic Leap. It just doesn't have a chord almost, you know, 52, you know, degree diagonals like the Magic Leap One. And I could, I could sympathize with that because once you see the rest of it in your mind, you could put that on and you're filling in the gaps. The problem is if you're not that of a true believer, you want the end game and you're not willing to take the mid step. And I think, you know, my guess is Evan and the team at Snap are hoping that the true believers will do enough with this. Uh, like, like weirdly we did with laptops and phones, people were willing to like suffer through the bricks and all those other things along the way here not as many people are willing to suffer. They just want the end. Uh, it's kind of like this is really different from all other computer things where people were able to do the early and mid and then you know, kind of you have your MacBook Pro, it's like kind of perfected. But you know, people lugged around giant 40 pound computers for a while. Uh, here, I don't know, I mean there's a small cult that's willing to do it and it's like cult, cult, cult, cult, cult finally breaks open.

Speaker A: The question is if will, will people accept that amount of friction and uh, prosumers maybe, uh, developers maybe, uh, but consumers $2,200, I'm skeptical. Uh, but I try that. Uh, I have not tried them.

Speaker B: They didn't have any actually working models.

Speaker A: And the keynote to make it clear, we've all tried the developer versions. I have multiple times.

Speaker B: Yeah, and that's good.

Speaker A: And, and the keynote he talked about features that are not in the developer version. Uh, so you know, he's talking about being able to watch movies like sitting in your pool, which by the way would be epically huge if you could do that. Um, but I, you know, that's not in their current version.

Speaker C: I mean I'm having so many echoes of like walking into the magically parking lot in like 2017, 2018 wearing like Magic Leap One and then magically do prototypes and having like a drive in movie theater screen Floating in the parking lot of like, wow, this is, this is a 300 foot screen. And doesn't everyone. And like actually watching movies like that and sending out a launch. I was doing stuff like that. I'm like, everyone will want this, but I think everyone wants it when it's perfect, you know?

Speaker A: Right, exactly. And they want it. But not. But not at the price of $2,200 and lots of other functional friction. Now, there was one other release at the show. I don't know if you had a chance to put on Project Aura from X Real Ted.

Speaker B: Yeah.

Speaker A: Uh, but I finally got to use it and I have to eat my. Eat my hat.

Speaker C: What do you say?

Speaker A: It's good, Ronnie. It's really good.

Speaker C: Is it the same bird bath or is it something else?

Speaker A: No, no, is. And somehow it isn't because they seem to have gotten hand tracking and computer vision. Right. So.

Speaker C: But is it 70°? They claim 70°?

Speaker A: Yes. It's big. It's really big. It's a really big field of view.

Speaker B: It's still bird bathroom. It's still. But, uh, let's, let's go back to this, to the snap thing for a minute, because actually that's more important, I think, than bird bath optics at this point to sort of wrap this, this discussion point. Rony makes really, really good points. And this is kind of where I refer to. He's honing in on why I call it the gray zone, is that until you can actually make them look and feel and weigh like glasses, like the glasses that Rony is wearing now, they fit into the gray zone. They're not in the green zone. People, even the kind of cool, funky, stylish. And they did like a fashion campaign with a fashion photographer and famous people wearing these glasses. They're just too big and too clunky and too heavy to be a daily wear scenario. Right. And the Magic Leap was the same and the Hololens was the same. But like, I'm a perfect example of the person that was willing to put up with that friction because there were so many interesting things about the Hololens and the Magic Leap that made them valuable. The Hololens. There was a piece of entertainment called Fragments. I spent like 10 hours, you know, in this, like, uh, the, um, investigative crime scene thing. And we've talked before on Magic Leap. The best piece of content, mixed reality, I think, ever created is the Dr. G experience. Um, but the people that had a chance to do those, and those are very bespoke use cases. Right? Use them for a specific thing. You're not going to wear it all day long and it's not replacing your smartphone until we really figure that out, which is a huge gap still that we have to figure that out. These things don't fit into a marketplace. Um, and that's sort of the TED

Speaker C: let's, let's go there for our listeners because a lot of people don't have this experience all three of us had. But when we were building it, there was one theory of don't put anything out in the world until we hit the final form factor and just like be uh, a X lab type thing for like 10, 15 years. The other one was put things out, get feedback, keep iterating, keep iterating with the idea that people have accepted alpha beta, uh, and iterative products along the way in cars, in phones, in so many things. And I guess it's true here, but in this much different curve than people appreciated, I think the tolerance of like what you call the gray zone, I think a lot of people are waiting for the green. I think the company that gets the green is going to have billions of users. But the gray zone is like tens of thousands, maybe hundreds of thousands, uh, hundreds of billions. To walk through the gray zone valley of death, it's not costing billions. I think everyone did not realize. I mean here's the thing I tell folks when they call me, oh, should I go build an XR company? I'm like, you're going to walk through the gray zone, valley of death. You're going to spend more money than it took to build pseudo AGI and go to space. It costs more money to do this than to do SpaceX. That's mind boggling to me, right? That uh, in the same period of time, the same brilliant kinds of people, one is like landing on Mars and one's doing electric cars and all kinds of stuff. Why is this harder? Like that is so mind numbing. And I think Tim Cook and Mark and all these people and Evan throwing billions and tens of billions and hundreds of billions are like, why is this harder than space? It is, it is like no one saw it as the thing. Harder than space. Harder than like making opening eye work. It's, it's really unexpectedly difficult. Um, and that's just like the chosen field I guess we like.

Speaker B: And it's just because the form factor expectations are so high. When you put something on your face, it's got to be super light and super fashionable.

Speaker C: Like 40 grams, 30 grams.

Speaker B: Exactly. If you look at the current ad campaign that Meta did with uh, one of the, uh, Kardashians I think it is or something. And they have these like stylish new meta, you know, just the audio glasses and they actually almost look like glasses now. Um, but of course there's no visuals yet. So they're, they're driving the audio component because they know that the visuals still don't get you there. Right. In terms of like, okay, but they're getting closer. Like they're getting closer, but they're making other trade offs to get closer.

Speaker A: So we have Elaine coming in in a few minutes. I want to pivot to a couple of other stories. Um, Cosm. You guys know Cosm, right? Of course. The Semi Dome Entertainment new uh, movie theater of the future, I guess I would say, uh, they just got $100 million from Sony.

Speaker B: Fascinating to me. I would, I would love to have been a fly on the wall from the aspirational side of what Sony thinks is a good idea. You know, they bought Alamo Drafthouse, so they're, they're looking at other sort of out of the home forms of entertainment as a, as a business vertical. But the true, like scalability and viability and the business model of Cosm, I'm, I'm a little kind of like, huh, huh. I wonder what Sony sees that I don't see about the viability of that business at that scale. I think it's a cool little exotic thing I think they can have.

Speaker A: They're gonna put one in every, they want to put one in every city because they think that they've got content that will be of recurring interest to local people. It's going to be a mix of sports and I guess films that are, I mean this is your uh, this is right in your strike zone, Ted. I mean, I think bespoke made for Cosm, right?

Speaker B: Yeah. I think the sports side is interesting. You know, the SP experiences in there are really interesting and valuable. I just question the amount of scale that they can, they can drive. These things are expensive to build and, and exotic to run. But maybe like literally, I would love to know, maybe they, maybe they see a business that I just don't see.

Speaker A: I, it's, it's super duper premium. Right. You go there to watch a football game, you're going to spend $200.

Speaker B: Yeah, it's like, it's like as much as actually going to the game.

Speaker A: Exactly. Um, so, you know, that's the thing that I, I wonder about. But you know, they're, they're, they've got scarcity on their side. Right. It's a very unique, you know, viewing experience. So, you know, maybe there's a, an audience for that.

Speaker B: I mean, the only thing I think is, is maybe people like me were like, huh, topgolf interesting. Like, can you really make that a big business? Like, it's like a fancy driving rage. Like, why is that a big business? And you know what, it kind of worked out like they made it a profitable business. It's owned by Callaway now. It like a, it's a thing. Right. It's like another version of going bowling. Um, so maybe KM has a business model that Sony business people sort of, you know, reverse engineers and looked at it and said, oh, we could maybe 10x scale this and we can see a return of a couple hundred million on a, uh, on a $100 million investment. So maybe they, maybe they're right. I just, I don't see it, but maybe they're right. Like, I, I want to self question my own belief structure that it doesn't seem like a great business to me, but maybe it is.

Speaker C: Let me, let me throw this out to you guys. What if you have this recurrence that you have at sports bars or bowling alleys and there's food and beer and then you have that Sony branded content which is what you get at a Disney, but it's in every cool city. Like no one's fully cracked that code. It's like, it's not Chuck E. Cheese. That's like low end. But, um, you know, if Sony can correct that code, they have a counterpoint to Disney's giant theme parks. Right. Well, this is much more local and you know.

Speaker A: Yeah, maybe Wolf is doing something similar in that.

Speaker B: Yeah, yeah, yeah.

Speaker A: So really interesting. We'll see if, uh, if sort of a new kind of habit out of the.

Speaker B: I'm super excited about it. I'm motivated that they're doing this. I'm involved in all kinds of things like this. I think it's great. I just hope it doesn't turn out that it didn't make enough money and then they bail and then the whole thing kind of pulls.

Speaker A: What's, what's the premise here? It's something comes after the movie theater.

Speaker B: Right.

Speaker A: That's what people are thinking.

Speaker B: I, I truthfully think it's more for sports, live events and concerts. I think the movie side of it is an interesting piece of the puzzle. Maybe they've seen the success of wizard of Oz as a, you know, as a, as a. Brought it back, um, and turned it into A viable like, you know, wizard of Oz is making a lot of money, right? Because the cost of making it, it's not like a live concert. It doesn't cost a lot of money to put it on, um, once you built it and you can print cash. So maybe someone at Sony studied the wizard of Oz model and said, look, we've got a lot of really good movies in our arsenal and so do our other, uh, studio partners and we can build cool.

Speaker D: I don't know.

Speaker A: You're going to get the same amount of people in to see the man who Would Be King as came in.

Speaker B: Well, obviously it's not that, but you know, maybe it's Jumanji and maybe it's like they're like, you know, maybe it's going to see Raiders of the Lost Ark and ET and certain things that are like all the different studios.

Speaker A: Isn't the dome in Las Vegas just a, you know, financial failure based what they had to spend to build it.

Speaker B: We should bring in an economic analyst to actually answer that question because I think there's both sides of the equation. So you know, the Jimmy Dolan had to spin out the company from MSG because it was sucking all the cash. So he had to make a separate company called Sphere Entertainment or whatever it is. Um, but from a standpoint of the actual, like does it fill up the venue? Do they have a successful concert? Um, strategy and this movie strategy with wizard of Oz of which they'll do others, that has been working. Right. And it fills audiences and I think it's one of the most successful concert venues in the world now, albeit the cost of running it and the cost of building it. It would be interesting to bring in someone that studies the economics of this and ask. I don't, I don't really have an answer to the question other than I think there are two sides to the answer.

Speaker A: All right, so, uh, I see Elite is in the green room. We'll get to her in two seconds. I just want to also point out two other stories this week that we're not going to discuss. Google invests in a 24, the film production company. Uh, supposedly their engineers are going to be uh, collaborating with them to make tools.

Speaker B: We, we do have to discuss that for a minute because it's a 75 million dollar investment to make tools.

Speaker A: I don't know, I don't know what that means.

Speaker C: By the way. That's like Ted, that's like you buying chewing gum at 7:11. I mean that's a good for Google. It is like it's like money perspires off their forehead every second. ET is actually, it is, it is not even worth the money to invest the money because they make so much money kind of money. It's, it's, it's insane because they're like a multi trillion dollar company. But 75 million is a nothing burger. And it's really sad to say that out loud, but for them it's nothing.

Speaker B: Well, potentially the thesis that makes that make sense is the two big hit movies of the summer are these tiny little indie movies. Obsession and Backrooms. And uh, potentially Maybe both are A24, but certainly one of them is an A24. I think it's Obsession. Um, and that thesis of building content for almost nothing that resonates with youth culture. Using the technology of the day. Right. Using AI as the, the right thing, not the wrong thing. Like the way to. In, in integrate it effectively and organically with filmmakers as opposed to sort of taking over filmmakers. That's potentially an A24. Tries to figure out how to use.

Speaker C: And then Google buys them and you make YouTube cool because it's a 24. YouTubers. Yeah, I mean they need cultural cool and a 24 is cool and.

Speaker B: Right.

Speaker C: I wouldn't be surprised if at some point this is a test to see if Google should buy them or not.

Speaker A: Um, interesting. Well, it'll be interesting to see the direction YouTube goes as it becomes more and more of the dominant network, if you will. Um, one last thing. I'm going to skip over this one and not discuss it, but I thought it was a good one for you. Roni Midjourney has launched a medical division, uh, and it's going to use its AI technology to uh, I guess create kind of a casual, less invasive MRI that can be used more sort of for monitoring than for serious clinical application.

Speaker C: Let me just say this MRI light. David Oltz is a, uh, friend and a founder. I've known him for 12 years maybe.

Speaker A: Yeah.

Speaker C: So he, he's a brilliant, genius guy. And I just want to give full kudos to someone that is, um, didn't take venture money, uh, is profitable with a small team, and is actually using the money he's made from Mid Journey to try to do something good for the world.

Speaker B: Oh, interesting.

Speaker C: You know what I mean? Like this idea that he's, I mean he, he's, he's making so much cash on Mid Journey, but he's pouring it into trying to actually help people. Not talk about it, not let's pretend the cure can, but really trying to build something innovative without being stuck in the bureaucracy of a big medtech company. So if, if you know anything about medical imaging, which I know some. It's fairly brilliant design they've got. And he obviously has to build this company go the FDA clearance. But they showed something that's. It's like 30 seconds, you immerse. It's a full 360 ultrasonic view of the company, rebuilds it instead of Mr. And CAT scan. This could be a game changer for so many people around the world.

Speaker B: This is a real thing, Haroni.

Speaker C: This is real. This is like. We haven't had something like this in imaging since, like, the invention of cascan.

Speaker A: It's, it's a little bit controversial in medical circles, so that makes it kind of dramatic.

Speaker C: You know what? When you see the flak, you're near the target.

Speaker B: I'm sure the original MRIs back in the day were also controversial. Right.

Speaker A: Let's, let's, let's bring in our guest, Aileen. Um, Hi, guys. Welcome. How are you?

Speaker D: Good. How are you?

Speaker A: Good. Pleasure to introduce you to my podcast co hosts, Ted and Roni. Um, we were all very excited to have you, uh, on the show. I'm sorry, Tilly's not going to be joining us today, but next time.

Speaker D: Yeah, she's. I mean, she's nearly there, though. She's nearly there.

Speaker A: Well, I said that you. You're an actress and a performer, so you could play both roles.

Speaker D: Can't play Tilly, that's for sure.

Speaker B: So do we pronounce it Eline? Is that how we pronounce your name?

Speaker D: Yes.

Speaker B: Where. Where are you calling in from?

Speaker D: I'm in London at the m. Moment.

Speaker B: London.

Speaker A: Got it.

Speaker B: Okay, great.

Speaker D: Yeah.

Speaker A: Elaine's the CEO of an AI company called Particle 6. So she segued from being a performer and a producer to being a Ph.D. in physics and starting an A.I. uh, company. Did I get that progression right, Elaine?

Speaker D: Yep. Maybe not the right order, but. But sort of the right gist. Yes.

Speaker A: Well, maybe you can give us the right order. And what led you to starting Particle six and then subsequently led you to making a synthetic actress?

Speaker D: Sure. Now. Or did you have another question, Ted?

Speaker B: Well, no, do that, and then I'm going to ask a question. Go ahead.

Speaker D: Okay. Um, yeah, so first I did musical theater, then I did physics. I was going into nuclear fusion. I wanted to do, like, save the world stuff.

Speaker B: Uh,

Speaker C: are you sure you're not a Marvel character? Because that's a backstory of a Marvel superhero musical theater. I did Any nuclear physics. And I became a fighter pilot and I did AI Anyway, keep playing sucker interrupting. But that is a Marvel backstory. But it keeps saying,

Speaker D: um, yeah, that's how I became a Hollywood villain. No, um, I then, because, you know,

Speaker A: the overlap between the performing arts and nuclear physics is so broad, it's just a no brainer.

Speaker D: That's what you go into. Honestly, I was sat in those labs and I remember just thinking I might. I don't think I can do this for more than the year that I'm meant to be here. It was like a year of research and it was so boring I couldn't do it. Um, but anyway, so then I went back to acting after finishing my masters and I think I did quite well in my acting career. I really enjoyed it. I was working. I had. Eventually I became a YouTuber and I had a show called Miss Holland and then it got picked up by the BBC and won a load of awards for best online comedy. And then after having my own show, I realized that actually I could make more money if I wasn't in front of camera. Because you could be in five shows, you could be producing five shows at the same time, but you can be in five shows at the same time. So I started a production company. And then four years ago when the AI stuff came out, I just got really excited and I think the physicist inside me was like, I. I think I can do this. And so I converted it to being an AI production company. And that's, uh, where we're at.

Speaker B: Interesting.

Speaker D: And then Tilly was just a normal progression of AI content where I thought, well, you'll have characters and some of them will be the AI A Listers.

Speaker C: So I want to question for you how there's been, um, some decades of people trying to build digital humans, uh, epic with unreal Lucasfilm, there, uh, was the Tupac. We built a digital person Micah at Magic Leap for a long time. But when you dropped Tilly, Hollywood seemed to have a, uh, crazy, uh, reaction that they didn't have over other similar things over the last 20 years. What, what caused it? What do you think?

Speaker B: You know, because I can give you my opinion. I'd give you my opinion. Coming. Yeah.

Speaker C: What? Because I was like, oh, very.

Speaker A: Why did nobody care about little Michaela? But suddenly Tilly Norwood is like a nuclear bomb and has to be or

Speaker C: two pack or all these or animated characters that have been around since the 90s.

Speaker B: By the way, before Eileen answers, everybody did care about all these things. And I will give you that perspective. But go ahead, I want to hear your perspective.

Speaker D: Um, I think Tilly was the first one that looked really real. And so it was the first time. But also there was, there was a lot of bad things that have happened in Hollywood. And so emotions were running high and she was like the lightning rod. And as an, uh, like a creative artist, which is what I, um, am. I was representing that fear in one person. Right. Like this thing can really replace us, which is what everyone was thinking about. AI. So, you know, that's what we do with art. We represent the, the zeitgeist back at us. Um, so I was, I mean, I've always made stuff that made people sit up. I think people sat up a little bit more for this one.

Speaker B: I think that's, that's the right perspective. So I have a whole litany of questions as sort of the, the pretend Hollywood insider in our group here. Um, insider. You've been hanging on sort of. I, I'm sort of not, well, disruptive Insider. Yeah. Um, the first thing is maybe the most critical. You actually glossed over a little piece of your little language, uh, there. When you said, you just, in your, in your talking, you said, and then we become the Hollywood villain. Um, and you just kind of glossed over that. That's actually very Hollywood villain, you said.

Speaker D: Oh, yeah, I said that. No, because, I mean, people think I'm the Hollywood villain, which I'm not.

Speaker B: And that's really, really interesting and critical to sort of everything that you're doing and others are doing, uh, in this realm. Because, you know, we're all storytellers that work in this business. Right. So when we can latch onto a story and a narrative and create you as the bad guy, that's a powerful narrative. And when we can create a synthetic being like Tilly Norwood as the bad guy, the bad girl, that's a narrative. Right. That people can latch onto. So let's kind of drill into that a little bit because it's fascinating to study it and to see if we can elevate past that. Right. Um, the first thing I'm curious about is something that is akin to what you're doing that has become a company called Metaphysic, which started out as the. I'm the deep fake Tom Cruise, and the overt story was Hollywood villain. The real deep story is actually much more fascinating where the kid that played the actor that turned into the deepfake Tom Cruise, was a relatively good looking kid that isn't quite as good looking as Tom Cruise. So he could never get parts because they were like, you kind of look like Tom Cruise and we don't want to hire you. So there's like, there's a really interesting story. We don't need to go into the whole thing now, but you can read up on it online. Just like how I became the deep fake Tom Cruise is a fascinating journey either into madness or success. Right? And this metaphysic technology, it became a company, what this, he as an actor and his developer in somewhere in Eastern Europe created, became this company that became extraordinarily interesting and valuable to Hollywood. Right? So started out as villain and then they start to study it and say, maybe this is a tool. And it became, uh, most famously used in a movie. Unfortunately not a great movie, but used by a really great director, a guy named Robert Zemeckis to de age and re age, uh, Tom Hanks and Robin Wright in a very, very unique high tech way. Very different than a traditional VFX pipeline. Right. So that to me fits into your narrative because it's a little bit akin to what you're doing, but you're taking it even one step further. Um, I think what concerns Hollywood so much is what you just touched on is that the perception that these synthetic beings can actually replace actors, can actually replace humans and what humans do, and you as a former actor, maybe still an actor or actress must, um, have some perspective on that. Right? So I know that's a lot to throw at you, but I think that's, that's getting to the crux of why Hollywood is both concerned and fascinated with what you're doing.

Speaker D: Yeah.

Speaker B: Well, do you have anything to, to add to that? Like any perspective for us to, to latch onto? Because just, just being like, okay, we're all concerned, it probably isn't the end of the story. Right?

Speaker D: Look, uh, no, I think you have to also, I think what people didn't realize was I was showing Tilly, you know, every, every week to probably 40 different people. Right. I was presenting all of last year because I felt passionate about educating the creative industries about what's coming and that this tech is here. And I created Tilly as like a showcase. Like, look, this is what we can do. Um, and when obviously the first time. So in the uk, when we released her, it wasn't like a big deal because everyone knew who we are. Because I've been in industry for a long time, they knew I was a creative and they were like, oh, this is really cool. We're using AI as well. Like, thanks. Because I feel like now's the Time to educate people. Because now there's a transition period where we can retool and reskill and prepare ourselves for the transition that's coming. If we ignore the tech, if we put our heads in the sand and just pretend it's not here and oppose it, and we will be out of a job, right? I will be out of job. My business won't exist. There'll be nothing for me to do. And so that's why I was like, okay, this is what I can do. And my contribution to mankind is to be able to educate people and show people that this is coming and get them to retool and reskill and prepare themselves.

Speaker B: Here's the Hollywood sort of perspective is that we're all interesting. And, uh, I'm just sort of using me as the conglomerate of all people in this industry that isn't really even an industry anymore. Hollywood is just a figment of our, like, previous imagination. It's a story that doesn't really even exist in the way it used to exist. So that's even a fiction in and of itself. Right? But the, the. The way that people feel about this, and this is, by the way, always a sliding scale, is that there's a bridge of what technology can do and as it relates to humanity. And then there's a bridge too far. And by the way, that bridge too far is a sliding scale. It always changes, right? It's like what we perceive to be a bridge too far. 50 years ago, 75 years ago, 100 years ago, 10 years ago, 5 years ago, 5 years from now is a different bridge. This bridge, the idea of a fully synthetic character that not only can just say the lines you tell it to say, like an actor, but can actually be a human replacement, something that can go on talk shows and talk itself. Like, it doesn't have to have a puppeteer behind the scenes. Right? That has a weird connotation and a feeling today. And I'm not making a judgment call on this. I'm just giving sort of a sense of it, the feeling today of a bridge too far. It is different than animation, where there are talented people behind the scenes bringing these characters that are synthetic obviously to life. Right? There's a human.

Speaker D: But why are there no talented people behind this character?

Speaker B: No, I think there are, and that's why I said there's no value judgment here. But it becomes the perception of a bridge too far once you let it become fully autonomous, once it is created and then unleashed to the world. But it is not being puppeted anymore. Somehow Psychologically, we as humans get concerned about that. Um, and that's this. But to the point of, like, the Tom Cruise deep fake thing, it was a bridge too far. And then a huge Hollywood director, Robert Zemeckis, said, this is not a bridge too far. This is just a tool that I can use to make my art in the way that I want. But there's a nuance around Tilly, unless I'm getting it wrong about Tilly, is that eventually what you want Tilly to do is run without humans behind the scenes, like, humans created her and are part of her. But then it can just do. She can be who she is without any human intervention. Am I incorrect about that or correct about that?

Speaker D: This is so, again, like, I didn't create this stuff, right? This is already happening. I am merely just showing the creative industry. I'm like a. Like a communicator from the tech industry to the creative industry. That's what I feel like. This is already happening. I'm not the first to create an AI actor. I am not the first to be making her autonomous. This is all already happening. I am merely communicating this is happening. How are we going to do it? We have an ethical framework in place as to how we think it can be done ethically and how our industry should adopt it. But, you know, I'm just showing it and I'm where it's at.

Speaker C: Let me ask you a question, totally other side of ted, just to see, uh, and love to hear your response on this. What is wrong at all that you, as a producer, build your own stable of amazing actors and actresses and your own IP and your own movies, and if it costs you a thousandth of what it costs, you know, Warner Brothers. But you become very successful because you've outsmarted their production process and you've broken their stranglehold on movies and you become successful. Why is that bad? Like, you simply, uh, are like the automobile beating the bicycle, right? There's one mafia, like, hold over entertainment, and you break it. Like, if you become a Warner Brothers studio and you have, like, your own CAA of actors and actresses and you make great films with them and you be. And you could direct things that you may not get the chance because you need a $200 million budget and this actress won't work in your movie, you make your own stable about, like, what's wrong with that? What's like, I personally go, if you're building your own United Artists and you can outsmart other people, like, go in, like, what. What is the Issue other than people losing to someone, winning. Like, it is a free market. Innovation keeps going. Uh, I feel like these are the people with, like, horse and buggies yelling at the car people. They're going to lose. There's no chance they win.

Speaker D: It's an industrial revolution, and that's exactly what's happening. And you can either decide to dig your head in the sand. Fine by me. I personally have not done that because I enjoy my work and I have actually turned to these tools and enjoyed creating with them. Um, I totally agree with you, Ronnie. Like, there is a way in which storytellers can put their films out there that they were never able to do before. And that's wonderful. There was a guy who just had his film, AI film at Tribeca. These things are happening. There were AI actors in that film. This is happening all over the place. There are AI influences all over your social media, and there are AI characters where you build an autonomous brain and they can communicate. That still has all the fundamentals of human creativity behind it, because humans have to create this. The machine has no impetus or desire to do anything by itself. Tilly would not exist if it wasn't for me wanting to create her today.

Speaker B: Well, you're touching on the nuance of why people feel concerned, not necessarily about you specifically until, but why people concern about this perspective. And this kind of gets into the larger discussion of LLMs being this giant, monstrous thing that even their creators can't exactly tell you how it actually does what it does. They can give you the broad strokes and they can tell you it functions this way, but we don't exactly know why it comes up with the outcome that it comes out with. Right. That's the concern. That's the perspective of the Bridge. Too Far is. And I keep using this term of, uh, puppetry. Right up until now, we as a culture have used front end creative, like visuals with a back end, whether it's synthetic or not, a back end of someone, then in the human perspective, still kind of knowing what's going on and being in control. It's a bit like the concerns around fully autonomous driving. Right. Where it's all good until the car runs into a house.

Speaker D: We don't know how your brain works and we let you do stuff. Right.

Speaker B: That's an excellent point. So that's a really excellent point.

Speaker D: That's interesting here. It's the new electricity.

Speaker B: Yeah.

Speaker D: And we have.

Speaker A: Well, that is true. The future is not optional. Uh, let me just jump in really quickly. I have, like, five things that I want to say, because everything here has been so provocative. First of all, with regard to the effectiveness of the press release, um, there are two things going on. One is, yes, timing.

Speaker B: Right.

Speaker A: You could have done that press release last year and nothing would have happened. Right. So it was a, a moment of heightened, uh, anxiety, which is persisting, um, about not only Hollywood and AI, but AI generally. Um, so. And I do want to talk to you about what it's like to be a lightning rod for that sentiment. Right. Because much of it misunderstands. And this was the other point I wanted to make as the only person here, I, uh, guess, aside from Aileen, who actually directs AI actors all day long. And let me tell you, the performances that are put together by Kling and Vo are, uh, surprisingly good.

Speaker B: Mhm.

Speaker A: Um, and I don't. I would like to credit myself as a director, but the truth is I'll put in a prompt and it'll come back with something and I'll go, whoa, that's like real acting now. It isn't coherent. Right. It's not, there's not a logic that goes from shot to shot. But yeah, it fakes acting. Fakes acting really, really well. Really well.

Speaker C: But, but all of this is.

Speaker A: Yeah, and so what I am. Mhm.

Speaker C: True. But all these things you're talking about, problems, they're going to get solved.

Speaker A: Yes, I do think they are going to get solved. And I think one of the ways they can get solved is by giving the act the AI that you, you know, taking the AI and separating it out into agents. So you have your cameraman agent, then you have your actor agent, and you're interacting with all of those agents much as you would interact with individuals. Um, so, you know, I think that way you can get, you know, real directors who have the amount of control. Because when you talk about, I'm not a real director of actors, I'm a guy who's good at prompting, um, and you know, has, has some experience putting shots together. Um, but I think, you know, the other thing is that AI actors belong in AI movies with other AI actors. I don't, you know, necessarily see a world where Tilly Norwood is placing Robin right on a movie.

Speaker D: No, that was never the point.

Speaker A: Yeah, that was never the plan. But again, that was immediately a lot of the hysteria in Hollywood, some of it driven by the Actors Guild itself was almost an idea. Yeah, exactly.

Speaker D: Misunderstanding of AI filmmaking. Right. You wouldn't have a real film and then replace one AI actor in the, In a normal film. That would never happen.

Speaker A: Yeah, you would do the opposite, right? You would have your AI actor actually dubbed and performed by a real actor.

Speaker C: Well, let me ask you, this is the tougher question. You will over time create dozens, hundreds of amazing actors. And some of them may get to be so good. They are Academy Award level. They're method actors. They emote, they make you feel things. And then at some point they might displace an Academy Award winning actor from a role because yours are not aging. They could be any age, they could have any color. They can do what you want. They'll endlessly do takes. If you're a director, like you're a producer director mode, they're probably going to be easier to manage than an actor that's like, you know, getting into a car crash or doing drugs or not showing up on time over time. Are not human actors somewhat vulnerable to this? And I just mean that as a very like, not an objective statement. Like, aren't your actors going to be easier to direct eventually? They might even be better. Like it sounds like heresy. But you must believe that in some level if you're doing this, it's, the

Speaker D: transition will be even bigger than that. Way, way bigger. Uh, because it won't be like displacing. So you'll have, it'll be just a new form. Right? So you'll have, you have an. I see it more as animation than I do live action. It's basically we're making animation, but it's hyper realistic animation. That's how I see it. And people will still want to go and see real live action with real actors. Fine. But eventually, yes, the acting that an AI actor can do. And there's no, there's not one AI actor is going to be a better actor than another because they'll all have the same exact training. Right? They are all trained on the exact same, um, information that's out there on the Internet. But they're trained on more emotions, all emotions that humans have ever emoted. So they can't be compared to real actors. Does that make sense?

Speaker B: But is that what it is today?

Speaker C: But is that going to be true three or four years from now where you can have an actor trained in this method and another actor trained in a different method? And I think what's true now is things are changing so fast. I just don't think what's happening today is going to happen three or four years and it's just going to get so much better, so much more quick, quickly that you're going to have exquisite control as a Director, producer, over everything your digital actors do.

Speaker D: Yeah, you're completely right. You either have control or you. Or it will just be naturally better. I think we'll. Have you guys tried Suno?

Speaker A: Sure. Yeah.

Speaker D: You will get to the same process with feature films as you can with sooner in creating a song. So you will be able to have a one prompt and a feature film. That will happen eventually.

Speaker C: What do you tell actors, Eileen, when they come to you? You are an actor, so when actors come to you, are you giving them guidance how to shift their career? Like instead of being the thing guided by a director, are you telling them to become writers and directors and producers? Like how are you guiding folks? I'm wondering if they're coming to you behind the scenes, yelling in public. But like Elaine, tell me what to do. Are you getting some of that?

Speaker D: I wrote a whole essay in Variety about this, speaking to actors as to, you know, what's coming because I want them to be prepared. So the best thing you can do is obviously if you're a famous actor or if you have a great look, is trademark your look and your name so you can license out a digital twin of yourself into these films. Because we will there you're already seeing licensing of characters happening into Roblox and etc so that owning your image will be really important and then, you know, live action will still carry on. But I think AI will be a huge, huge component of the entertainment industry and it will go beyond what we know it to be. Now look, in the uk, acting was always a, uh, you know, not a real job for many. For 99% of the actors, right, there were only a very few people that actually could make a full living out of it and we all have a second job next to it. That was our normal way of living as an actor here. That might not be normal in LA yet, uh, but it might become so like I don't know what the future holds, but I do.

Speaker B: It is hyper normal. It's the same exact scenario, you know, where we talk about brass ring phenomenon, right? It's like the amount of people that can make a legitimate living just being an actor, just being a performer is a very finite group, very, very small. That's a worldwide phenomenon. That's not a specific regional phenomenon. Um, that's just a knowledge base of there's only so much content that gets created even in a world of quote unquote, infinite content. Now with YouTube and everything else that's out there, most people that make content do not make a living making content. Right? Um, they, they are reaching for something. They're aspirational reaching for something. I don't think anybody would argue that's a bad thing. Um, but most people have to have second jobs, have to have another sort, uh, of scenario. I think there's a precursor to this that's already happened. It'd be interesting to sort of get responsiveness.

Speaker D: So hold on, just to get back to that. So it's only for a very, very small amount of people that this will actually change anything from the status quo.

Speaker B: Yeah, this may actually bring more opportunities to more people spread it potentially.

Speaker D: Like, you know, we might be getting in house actors. Right. Like people, uh, thought engineers would be out of a job because, you know, with, with Claude code you can code anything you want. Actually what happens is most companies now want an in house engineer to be working with Claude code. So I think you might start seeing in house actors in some cases as, uh.

Speaker C: Well, let me ask you a question. No one seems to yell at authors for sitting down alone writing a book filled with all these characters they made up. And that is, uh, an individual activity. But if you want to do the same thing with a movie, you have to sell your soul to a big studio. You're going to make a tiny fraction, they're going to edit your book, they're going to wreck everything and cast who they want. Unless you're auto writer, director, producer, billionaire, you get to make your movie. Isn't what you're really doing saying you can now write your book, movie anything you want with no interference? It seems that you are unlocking something that the Hollywood system has blocked creators from doing. Like no one stops them from writing a script, but the system stops you from directing your script and having enough money and casting who you want. It just is semi impossible. And even if you're a Spielberg, it takes you a decade to get something done. And then executives, you know, hear what you're doing. I'd love to hear what your thoughts. It seems that your idea, you machinate a uh, movie, a book, and you can manifest as a film with whomever you want to cast and it's what you want. Like an auteur. Is that, was that it a reason you did this or was like other, other reasons?

Speaker D: I mean, I didn't create the tech again. Right. Like you saw, you saw that. You saw, I saw that, I saw the potential, I saw the upside. It democratizes creativity. Exactly. As you say. Like it means that people and people are doing it. People are creating the films they want to be making. Great, great for them. Um, back in the day, we didn't all used to be able to write and read, right? We didn't all have a scroll and a thing. Now we all have a pen or a laptop and we can all write our own scripts that didn't, you know, that's a privilege. And so in 10 years time we'll, and we think it's normal we can all write a book now if we wanted to. In 10 years time we'll all think it's normal we can make our own film if we wanted to. I think it's just a new expression of oneself in a way that's not been possible before. And, uh, that's a wonderful thing. And what I do feel is we have to remember, and this is why ethically it's such a tricky area, is it's trained on all of our stuff, right? And that's the tricky bit. It's trained on everyone's work that's been out on the Internet, including my stuff, including your stuff, including all of us. The only way I have peace with that is because it democratizes creativity. And we all have access to those models. If we no longer had access to those models, which in some cases is the case, like Ben Affleck's company was sold to Netflix. I don't know what data they have trained on, or other things. Like we don't have full access to Anthropic anymore because the US has siphoned that off for them. Like that's when it becomes more problematic because it's trained on everyone, but not everyone has access. So that's a completely different scenario.

Speaker B: Ronnie, there's an interesting nuance that should be discussed. So you talk about, here's the author and they're writing and bringing their characters into the world. That's a form of puppeting, right? So they're writing their ideas, they're creating their characters, they're writing the dialogue, and then those characters come to life. And sometimes those characters come to life visually in a movie, right? They're still being puppeted by actors and creators and CG artists and all these different things. The question is, and this is where people get that ethically uncomfortable moment, is you're a best selling author or a new author that's never had anything published, and you wrote and puppeted this amazing world and created it, and then you plug it into a technology base and now the SQL is writing itself. It is no longer being controlled by you or written or influenced by you, other than the influence of the canon of work that You've already created and it is now doing it autonomously. It is doing it on its own. Right. And that is the concern dynamic is when you start to bridge that gap, when you start to go over that hump, um, where you no longer can control it and you are no longer the puppeteer. The.

Speaker D: We're already doing that. Ah, we're already doing that. So anytime there's a famous writer and they don't have enough time to write everything that they're meant to write, what do they do? They hire an assistant.

Speaker B: Yeah, they have ghostwriters. They do other stuff.

Speaker D: Or a ghost writer. It's basically the same thing. Right. So what you can do is you can train a brain on your writing method and then you can let it write sequels or books, which is exactly the same thing as training a human on your writing style and making it do ghost writing for you. So. And then you can check it, you can change it, you can give notes. It's exactly what we're already doing just with humans.

Speaker C: Lean, let me ask you the. You asked a different question about like the AGI systems we all have access to. Big question for you. I'm kind of a purist that I generally agree with the idea of what you're doing, but I don't think it should be using systems that have unlicensed ip. And I'm wondering if you're trying to move into a world where you can build your actors and your things where you know what's in it and it defeats the argument of like, all of our work is there. Are you thinking about how to get to a point where other people's work is not there? Um, and if you don't have rights to it. Because I think what, what the big AI companies have done is they have just claimed rights to people's ip, and I think that is the legitimate pushback. I actually think you can do what you're doing without having to do that. Do you think there's a switchover phase where you have like, you know, true licensed use cases and technologies that go away from that? I think we're in the Napster moment right now. But what's, what's your take on that?

Speaker D: Um, I think they've tried that. Um, I believe Lionsgate tried it with their limited data set and unfortunately the results are very bad. So that's why that's never going to happen. Um, they've used a fair dealing approach to train on all these different types of data, whether you. I mean, I don't agree with it, but that's the only way to achieve a certain level of perfection that's required. So, so the thing is just like a human brain, right, because we do exactly the same thing. We go to university to learn everything that's come before us and to become experts at a certain thing. And we read lots of books to become a good writer. We work for years and years and years to train our brains to do all this. And we don't reference everything that we've been trained on, right? So it's very much the same thing. And the thing is we can't go into these black boxes and see what they've been trained on. Just like with humans, you can't track everything they've been trained on and everything they've absorbed and redigested and put out there. So unfortunately I've decided that that's a lost cause, but this will be a

Speaker C: fight for some time.

Speaker B: You touched on that loss cause dynamic. Uh, and a lot of people in our industry are discovering that very, um, robustly they have an aspiration that they want to use small bespoke models that are only trained on license paid for, you know, providence level, like tracked IP pieces. And what happens is they try and use that material and compare it to a sea dance or these giant models that are doing it. And unfortunately what happens is the giant models that are trained on everything, that just scrape up everything, just do a much, much, much, much better job uh, at delivering on that promise of creating a realistic actor, a realistic scenario. So maybe that can be solved over time. But right now the, the, the sort of like scrape the world model is working because it works better. And that's both, uh, exciting thing and a problem at the same time, right? So we, we don't quite know how to grapple with when the thing that we don't necessarily agree with, which is take everything, don't worry about all the IP and who owns what, it'll all just get shaken out. And then there's a like psychological sort of.

Speaker D: So we have a really clear AI policy in place, right? Because just like with a human brain, if I say to you, can we have a Mickey Mouse like character? And then it walks into like you know, a film like the Lion King, right? I'm referencing two pieces of, of IP there. If you call those bits of IP in the prompt, it's going to recreate something quite similar. And we do this all the time in film and tv. We, we reference stuff, right, in order to communicate what our ideas to someone else. Now in our prompts, when we use it. We do not pull on those specific pieces of copyright when we are creating our work. That's an AI policy we've put in place. And we've got very clear, like, ethical guidelines that we've decided for ourselves. And that's why Tilly wasn't a replica of another actor. The whole point was to. She was not.

Speaker B: She's a whole new entity. Yeah, yeah. And, and Google, by the way, tries to do this too. Right. So if you create stuff and prompt into things and say, I want it to like, pretend like it's this actor and like, it'll, it'll sometimes get over a threshold and say, you have hopped over our privacy policy, our restrictive policy. We can't create exactly that. We can create something like that. We can make something that looks sort of like Tom Cruise with an action hero sort of feel from Mission Impossible. But, uh, we can't do sort of what you're asking, which is give me a Tom Cruise replacement from Mission Impossible and put it in my movie. So Google is trying. I don't know if they're always succeeding, but they, like, you will, you will cross that moment where they say, no, no, we can't do that for you. Um, so there's people in the background working on that, trying to figure out where that line is, which is interesting. The other piece I was going to see, if you wanted to comment on, which is interesting, is there's a version of this that's happening already now at a commercial scale, which is just on the voice acting side of things, which is when we do, uh, dubs for, um, uh, other countries. Right. And so movies that are typically made in America, they want to be translated to 200 different languages. There used to be and still are to a certain extent, but this is now starting to fade away into an industry in a category that is no longer a valid job anymore, is there are jobs, actor jobs, to dub it in Italy and Germany and Russia and all the different places. Right. Um, and typically, you know, the lips don't match and it's not a great experience, but the actors in those regions are known like, you know, the Tom Cruise in Italy is a guy that's had his job for X amount of years being the Tom Cruise in Italy. Right. That is now being replaced with a technological advantage of using Tom Cruise's actual voice and synthesizing that in any language and doing lip synthesis so that when he's speaking in Italian, the lips match perfectly. So now it's not a bad dub job. So there's A pro and a con to that. Right. We. We've replaced a human workforce or starting to and probably wholly will replace that human workforce. The actors themselves get more work. The ones that are actually the working actors, if they're smart enough, they actually get paid for every language they do. Um, but that's going to their pockets, not the dubbing actors pockets. Right. So everything's changing.

Speaker D: Um, yeah, there's a huge transition coming. Huge. It's. It's so big, we really have to prepare for it, and we're not ready at all.

Speaker B: Yeah, yeah.

Speaker D: The, uh, biggest thing, what we see is we get these Hollywood or traditional film TV people come in here. I mean, we're all filmmakers, but there, you know, other filmmakers that haven't touched AI and they come and sit with one of our AI editors for half a day and they leave and they go, this is completely not what I thought it was. And that's the big disconnect is I really recommend that, you know, if you're an AI skeptic, you go and understand it before you have a devil is in the details.

Speaker B: Right. It's like understanding what's really going on. And, like, I think I've used the word nuance in this podcast more than any other of our, uh, almost 300 podcasts I've ever used, because a lot of these are nuanced discussions that do not have binary answers. Right. There's a sliding scale of where we cross that line.

Speaker C: So I would almost say stop using the word AI and just call it, like, you know, animation. Advanced animation.

Speaker D: Yeah.

Speaker C: I mean, Pixar does advance. So does Lucasfilm. If you were just advanced animation, no one would ever. It's what you're doing. You're using tools to do better animation.

Speaker D: Absolutely. I couldn't agree more. The funny thing is, though, uh, as much as, you know, the backlash has happened. Loads of amazing people have reached out to us as a result. Right. So it's sort of this, like, secret interest, um, as well.

Speaker B: Look, AI works in your favor, right? I mean, you're able to make an impression. People. If you didn't talk about AI, no one would know about you. You're just another person doing really interesting, great stuff with characters. So you play that to your advantage. That's well understood. Right.

Speaker D: You see a whole load of new filmmakers coming out who are experts at these tools. I'm really excited about all these new filmmakers that are rising to the top.

Speaker B: So this is obviously quite a fascinating discussion. We lost Charlie's audio. He is trying to figure out a way to talk to us, but he'll, uh, see you next week on the AIXR podcast. Thanks to Aline calling in from London. It's a fascinating story. The controversy both is interesting and challenging, and that's great. And, uh, we'll see you all next week for next week's episode of the show. So for listening.

Speaker C: Thank you, Elaine.

Speaker D: Thanks, guys.