B (3:19)

Oh, we call them digital twins, actually. So, you know, they are digital replicas of the real world content creator. So, for example, for UK listeners, about three weeks ago, we relaunched Jordan with Katie Price. Now, those of you might know that Jordan used to be Katie Price's, I guess, early 2000s glamour model Persona. We effectively took training data or shoot imagery of her from back in the day, use that to train up a digital twin of herself and then put that live on our site so fans can come on and they can consume new content for the first time in 20 years of Jordan. That can be risque. It can be even a little bit more than that. They can engage in written conversations. We clone the voice so they can receive and send voice notes to the character. So I guess it's as close to an experience of engaging and interacting with the real content creator as interacting with them in real life and with them in person, basically.

A (4:10)

So Jordan is now alive on.

B (4:12)

Oh, yeah, she is, yeah.

A (4:14)

And you can come and talk to you. You can. You said sort of engage. Yeah, but it sounds like. But, you know, potentially more than that. I mean, is that. Is that something that she's happy about? I mean, that she's obviously signed this, Hugely excited about.

B (4:28)

I mean, she's huge as most of our creators.

A (4:31)

Jordan's pension, is it?

B (4:32)

Well, we hope so. I mean, she certainly thinks so. Yeah.

A (4:35)

How does that work?

B (4:36)

I mean, I think for all of our creators.

A (4:38)

Does she get it was she gets royalties or you get royalties?

B (4:41)

Exactly, yeah. So what I'll go through two parts, kind of why the steps of creating the twin and how that works and then why they love it and they.

A (4:47)

Think, all right, so Jordan's very happy with this. She's created her AI twin on your site. Talk me through it.

B (4:53)

So basically the first thing to say is all our digital twins, anyone that rep sort of, that is a replica of a real world person. It's all consensual. So we have licensing agreements between us as the platform and the individual represented in order to create a digital Twin of themselves. We need a few things. So we need imagery. So 50 images, ideally, at a bare minimum, high quality images of themselves. Full body, different outfits, different forms, full.

A (5:15)

Body, clothed or naked.

B (5:17)

Well, to be fair, that does depend on the level of the twin you want to create.

A (5:21)

Level.

B (5:21)

Yeah. So internally we have four different levels. Users won't come across this necessarily unless they go onto a character's profile. But level one, which would be sort of fully clothed, safe for work as we describe it. Level two, which would be partial nudity. Level three, which would be full nudity. And level four, which doesn't exist on the platform, but I think is something that we certainly have creator appetite to do, would be full, what I guess would describe as adult content. The creator chooses which level they want their twin to be represented at, and they provide content to us that allows us to train a replica of them at that same level. So they provide us static imagery, they provide us a clone of their voice. So we have a voice onboarding bot where we ask them questions, they reply with a voice note, and we use that data to clone their voice and then we actually use the content from those answers to fill out their Persona and their backstory. So that if. If I'm a fan of Katie or a fan of Jordan, and I know a bit about what she was like, and the content of her answers to us allow us to fill, sort of fill a profile that accurately represents and portrays her so that the illusion isn't broken too much that I'm sort of chatting with an AI, and then why.

A (6:28)

They people think they're talking to Katie Price?

B (6:30)

We don't say that. No, I mean, we very much call them digital twins. It's very much an AI site, but we certainly want the experience to feel like you are.

A (6:36)

Yeah.

B (6:37)

Why do they like it? Which I think is such a great question. So a lot of our creators are on OnlyFans. And one of the limiting factors of OnlyFans, ultimately, that AI solves, is scale. So you as a creator can only talk to so many people at one time simultaneously before you run out of bandwidth. And AI can massively solve that and make it more scalable. So, you know, our digital twins can be having 2,000 simultaneous conversations at any one time with thousands of people around the world without that creator having to engage or actually put any work into the interaction. Their work goes on at the start when we make the twin, but they're not involved day to day. So for them it's kind of a scalable asset where I can earn while You Sleep is obviously a piece of narrative we like to give out in the pitch to them, but for them it's an ultimately scalable asset that gives users the opportunity to create unique content with them whenever they want, wherever they want, within the boundaries of the creator reset. And so it's an exceptional experience for the user and it's a much more scalable way to monetize that unique IP that a creator has for them than a real world interaction.

A (7:39)

So on my way in here just now, I was recognized on the tube by someone who used to work for me and he noticed I dropped a card and pass it back and I said, hello, silly, but if I was on your platform, this would be happening sort of everywhere, potentially this one in. In a kind of weird fashion, wouldn't it? I mean, you just described the upside, but the downside could be quite sort of. Well, I think dystopian.

B (8:03)

Well, I think it could be quite a statement. I guess that's baked into the decision making process of the creator.

A (8:07)

Yeah, that's their shout.

B (8:09)

That's their shout, yeah.

A (8:10)

So you have these different levels. How many AI twins are sort of alive and well on your site at the moment?

B (8:17)

So on the platform we have about 30, 35 or so live. We have nearly 100 signed and many.

A (8:23)

More people who are coming through.

B (8:25)

Correct. And actually what we're about to launch, which is really exciting, is at the moment the process of outreaching and creating a twin is manual in the sense that the AI does the training, but there are developers in our engineering team that are making that happen. But soon we'll be automating the full process. So you as an existing creator can come on, you verify your identity, you verify your age, you verify that you have an existing profile on another, more established content site so that you are an existing creator. And from there you can upload your images, you can record your voice, and our software will train your digital twin automatically and let you know when it's ready. And you can go live on the platform with no human intervention at all. So for us, that's a way for us to scale from 100 to 100,000, you know, ideally.

A (9:06)

So how soon is that going to be possible?

B (9:08)

Oh, within, within four or five weeks.

A (9:10)

Oh, right.

B (9:10)

So this is huge.

A (9:11)

So, so, so you have to be on another site already to get set up.

B (9:15)

Not long term necessarily. I think this, I think this goes.

A (9:18)

Into what you're talking about. If you're on Only Fans, you can come on here.

B (9:20)

Great example.

A (9:21)

Is that what you're saying? Great example, if you're on LinkedIn, does that still?

B (9:24)

I don't think we've extended it to LinkedIn yet. Well, we might, we might.

A (9:28)

Okay.

B (9:29)

I think not to kind of make it.

A (9:30)

Well, you can go all four levels.

B (9:32)

Yeah. Not to sort of bring it down to another point, but I think. I think the protections around this are quite important to us.

A (9:38)

Yes.

B (9:39)

At a certain point, we would of course, like to find a safe way that if anybody wanted to create a digital twin of themselves, they can do. But right now I think it's about baby steps. So verifying the age, verifying that that creator has already engaged in creating that type of content on already established platform. I think it just is the next sensible step for us to ensure that we are creating digital twins with people who understand the nature of the business they're getting into and the nature of the interaction.

A (10:04)

What would be an established platform? And I mean, give me a couple.

B (10:07)

Only fans would be a good one. I mean, fans Lee Fan View, who would be kind of OnlyFans type competitors would be good, good examples. I mean, when level four content comes out, to be honest, established tube sites like pornhub and others would also be.

A (10:19)

So this is really adult content that they're producing.

B (10:22)

Not all of them, but like, that is certainly part of the offering. Yeah. And to do so in. I know we can get onto moderation and protections and stuff, which is an important point, but to also do so in a highly moderated guardrail environment, if.

A (10:36)

Like me, you run a business or manage a team, you'll know how important it is to hire great people. Over the last 65 years, Reid has worked with thousands of businesses around the world, helping them to recruit and to grow. This is what inspired us to create Reed AI, your new AI hiring agent. It's here to save you money, to save you time, and to help you get the right match. Whether you're hiring your first team member or your 50th, read AI is here to make it easier. Try it. Free for 14 days, exclusively available during our beta period. Go to Read AI and get started. Read AI your higher power. So. So you're going to be using a lot of computing power on you to train AI to recreate all these twins.

B (11:29)

Yeah, a huge amount.

A (11:31)

A huge amount.

B (11:32)

A huge amount for us, actually, that's. That's the biggest.

A (11:34)

That's a huge cost.

B (11:36)

It's a huge cost, is a huge challenge. I mean, the funny thing is, at the moment, you know, so we have about 250,000 users, nearly 300,000 registered users of the site since we launched about nine months ago. But given that, you know, we're still not fully optimizing with that number of users, the extent of our GPU stack. So the GPUs effectively power the generation of the AI models and the experiences. So up until a certain point, the better we utilize those and the more effectively we do, and we do that through growth, the lower the individual cost per interaction will become. So AI used at a large scale on a per interaction basis can be cheaper in terms of a text message sent, an image generated, but yeah, at scale, it's going to be hugely consuming. And that's just on the consumer side. Don't forget we have, you know, API clients who are much more established businesses than ours, who are looking to power AI experiences through their sites, and they're already doing multiple millions of users a month in terms of throughput. So the infrastructure for us technically is probably the biggest challenge that we have to face.

A (12:38)

So you started this business in the uk?

B (12:40)

Yeah, Right, correct.

A (12:41)

And how big a team have you now assembled?

B (12:44)

So we're about 25.

A (12:46)

Right.

B (12:46)

And that's all remote. So we have an office, which I basically am the only person that works out of most of the time. My business partner is based in Lisbon for the majority of the time and everyone else is remote. That's probably two thirds on the engineering technical side, a third on the marketing side. I think the vision for us actually is that we should be able to scale quite a bit further than where we are now without having to increase the size of that team too much. And there are two ways we do that. Talent density. So finding the best people and the right people, so that we can maximize the value of any given position within the business. And for us, a bit, I think automation has to play a role in it, particularly on the development side. You know this feature I mentioned before, where we'll allow creators to come and create their own digital twin completely autonomously of human interaction that's powered through a piece of software which we built the front end of in three hours for an AI coding tool, and that would previously be three to four weeks of active development time for a team of two to three developers. So I think for us we see massive scalability being opened up and cost savings happening by pairing top tier excellent talent with automation and AI.

A (13:52)

Now, Felix, when I went to look for O, I didn't find it that easy to find, and I think we ended up finding it by finding you on LinkedIn and clicking through. Clicking through to it clicking through to it on your bio. But I think The URL is ochat.com ochat.com yeah. So it's this discoverability point and a few. Is it, is it platforms suppressing this.

B (14:17)

Content or discoverability is an issue. I mean search engines like Google don't suppress our content in the search engine format, but it does from an ads format. So for example, we can't serve Google Ads like traditional B2C sites can. We can't serve meta ads in quite the same way that other sites can. So we have to be a little bit smarter about the way in which we grow a consumer brand compared to how you might have done historically. We're not prevented as a, as a business and, and we do share, you know, you have to sit within certain content guidelines which we do of course. But generally the moderation teams of these businesses take a bit of a dim view of any company that, that operates in the kind of not safer workspace. So we, we have different methods of, of growing our business but the creators is, is one of the number one ways of doing that. So I mean OnlyFans is explosion was driven obviously through the power of the creators in their network. That is they had millions of followers and, and they were promoting their, their OnlyFans accounts and other accounts on a daily basis to millions of people. And it's the same sort of network effect that we're looking for through our website. If you look at a graph of social media activity overlaid on top of user spikes, you can make exact correlations to the moment that some of our creators post content on their social media sites and usage spiking on the platform. So accessing their network is a massive asset to us. And then there are other ways we have to grow our platform, but it's.

A (15:38)

So you're relying in a sense on them promoting their digital Twitter.

B (15:41)

100%. 100%. I mean the quid Jordan will be.

A (15:44)

Telling people on her Instagram.

B (15:46)

Yeah, absolutely.

A (15:46)

Out there. Exactly.

B (15:48)

Doing interviews. I mean actually we launched Jordan. It was by far our biggest campaign to date. The sun put her back on page three or her AI digital twin back on page three, which is where she made her name. CNN did a piece. So I think the appetite for this kind of conversation is huge and we just encourage creators to capitalize on that and build their own brands in the process.

A (16:04)

So, so researching this, I saw that you had Carmen Electra.

B (16:08)

Yes.

A (16:09)

As one of your AI twins. And Carmen Electra was obviously a Baywatch star, well known actor and, and media personality. And then I was thinking about Pamela Anderson.

B (16:22)

Yeah.

A (16:23)

Who. Who's just had huge success with her film Showgirl. And they seem to be going in sort of diametrically different directions. You have Pamela Anderson speaking against pornography and talking about the need for sort of people to connect and more old fashioned ways, I suppose. And Carmen Electra doubling down on. On this. Do you see this causing sort of social and mental health problems? I don't.

B (16:53)

I don't think so. And one of the reasons that I. So I think this space or anything in the not safer workspace, generally people take a dim view of it. And I believe that the primary reason for that is lack of content moderation.

A (17:03)

You call it not safe for work.

B (17:05)

Yeah. Nsfw. Not safe for work.

A (17:07)

What's that mean?

B (17:08)

Something you wouldn't get out at work.

A (17:10)

Okay. You wouldn't show it around in the office.

B (17:12)

No. You might not want something. Your manager looking over your shoulder.

A (17:16)

We block all this stuff on our system.

B (17:18)

Not everyone does. Yeah, yeah.

A (17:21)

So you can't look at it at work typically.

B (17:22)

Yeah, exactly.

A (17:23)

All right.

B (17:25)

So I believe one of the primary taboos or reason that I think payment processes most in traditional business infrastructure take a dim view of adult businesses is because of poor content moderation.

A (17:36)

Right.

B (17:37)

And one of the reasons that content moderation is something that businesses struggle with is two. Two reasons in my opinion. Number one, I'm not sure that many of them are that keen to do it because I'm sure their businesses grow as a result of it. But number two, it's quite hard. I mean, our CMO actually was previously the global head of marketing at OnlyFans for about 2, 3 years over their big growth period. So we have a little bit of insight onto what went on during the growth periods there. And they had hundreds and hundreds of full time content moderators on the clock at any one time, 24, 7, going through what was being published on the platform and taking stuff that went against their guidelines. And I think it's often, if you talk about porn, it's the exposure to the extreme stuff, the stuff that shouldn't be so easily accessible to underage people or anyone, quite frankly, online that breaks moral and legal boundaries. That I think is the number one crux of why people find this challenging. One of the reasons we went to AI is this ability to moderate at source. So all of the models are our own fine tunes on top of open source foundational models. It's a technical term, but what it basically means is that somebody else built the foundational text model or Image model, they posted the code and architecture of that on an open source in an open source format so that other people could use that. And we effectively build on top of that foundation to create our own bespoke AI model. And the reason that's important is that all of the content that's generated on the site is generated from our models, which are built from our training. So through the training process, through the guardrails that we put in, we can moderate content at source. So you can't create underage content, you can't create violent content, you can't create content, which goes against our policies purely because of the way in which we've trained the AI. So I think when it comes to moderation morality and how we protect users from harmful experiences, we can't do everything. But I think we can certainly ensure that we are doing everything we can to make sure what is consumed is safe, that users and creators are age verified, that there's consent in place on the creator side, there's clear ownership and contracts in place for use of their image and generating that content, and make sure that at least in our environment, it can be as enjoyable as possible. It's also being as safe as possible.

A (19:41)

There's a lot there. I mean, I'm thinking about, you know, you've only been running nine months and you've got hundreds of thousands of people engaging. What sort of, what's the data telling you? I mean, do people get addicted to this sort of on it all the time, or do they come back regularly? Or men use it more than women or in what way do they use it?

B (20:00)

So men are a large chunk of our user base, which perhaps unsurprising.

A (20:04)

Large chunk, what, 90%?

B (20:05)

About 80%? Yeah. Perhaps unsurprising to a lot of your listeners. Usage patterns are quite curious. So there's a general consideration that men are more visually led when it comes to consuming this type of content. Women might be more narrative led, and we do see that in the data. So let's say we break down the interaction formats, text, voice and image down. Generally, our female characters, the ones being used by our male users, are always making up the top 30, 40, 50 most popular characters in the text and in the visual category. But when it comes to audio, it completely flips on its head. And actually we see some of our male characters generally dominating the top five charts in terms of number of audio messages sent and exchanged with users. So that tells us that that kind of hypothesis around how people engage is being carried out.

A (20:52)

So you have charts, so you can see who. Which of your twins are performing?

B (20:56)

Yeah, absolutely.

A (20:58)

Who are Top of the Pops at the moment.

B (20:59)

So Carmen Electra has been top all the way up until Jordan launched about five, six months ago. Both have always sat within the top five. We have a couple of quite prominent, but probably lesser known from a public perspective, but maybe big on TikTok and Instagram creators like Isabuscemi who have over 5 million followers on Instagram. Rio Sage. We have Kylie Thorne, who's a sister of Bella Thorne, who I think in her OnlyFans days was the top earning Onlyfans or something. Earn something like $3 million in her first 24 hours on OnlyFans. So quite prominent in that sort of younger demographic community. But generally at the moment, so far, those that can marry a social media presence with a genuine media Persona or personality. Katie Common.

A (21:43)

But the men, the names of the men, you know, I was looking at them as Robin Hood and Sinbad the Sailor and the, the. I mean, they real. I mean these are Personas the way you've talked about Carmen Elector and these are real women.

B (21:55)

Yeah.

A (21:56)

Are there any men who would stand up and say, here I am, this is me?

B (21:59)

Yeah, we do have some men on the digital twin side. So it's actually an important distinction. We have digital twins and then we populate the rest of the platform with what we call original characters. So they're fictional, like you said. You mentioned some Sinbad the Sailor, Robin Hood, for example. And that's for. For two reasons. One, there is appetite to have kind of completely fictional interaction.

A (22:16)

So people want to have a relationship with Robin Hood of some.

B (22:18)

Or interaction. Yeah, absolutely. And actually this is where AI becomes quite cool. Right. It becomes more than just replicating Robin Hood when it's a bit of a. What can you do with Robin Hood extrapolated out? You know, imagine we could create a AI digital twin of Kevin Costner when he was in the Bodyguard. I'm sure those people would love to talk to that. So I think AI gives.

A (22:35)

They did have to give you permission.

B (22:36)

Yeah, well, they would. Of course we wouldn't do it without it, but I think AI just gives you this opportunity.

A (22:40)

But Robin Hood's not around to sort of.

B (22:42)

No. And actually, actually the copyright laws after a certain period of time expire, so you can have a bit of free rein on fantasy characters. So there is this appetite to explore what isn't possible in the real world through the use of AI which is an interesting pocket of our user base and the way people use the platform, for sure.

A (22:59)

People who view your site and OnlyFans say you can't tell the difference between a person twin.

B (23:06)

Well, it's interesting.

A (23:07)

I mean, what would you look out.

B (23:08)

For to know our images are better? I don't know.

A (23:12)

Yours are flawless.

B (23:13)

Well, I'd like to say that humans are flawed, so you say that it's actually interesting point. So I won't name names because I shouldn't really. But some of our creators are using and selling some of their AI content that's produced from our models on their Only Fans pages, for example. Users can't tell the difference some post which they're obviously completely allowed to do content from our AI models on their social media platforms. And we often see that it gets better interaction, or at least the same interaction and engagement and people don't know. So the quality of the engagement and the interaction and the visual experience is on par or can even be surpassed. Now, on our own site, we're pretty clear that it's AI. I mean, we don't at every single sentence tell you because I think that kind of kills the buzz a bit. But they're called digital twins. It's clearly an AI generated platform and in terms of building trust with users around the use of AI and artificial intelligence, that's very important to us. But when put in an environment where you're not saying banner, it's AI like Instagram or like TikTok, engagement levels are the same, if not better. People love the content.

A (24:15)

Well, there have been some famous cases recently of people thinking someone was real at Wimbledon and elsewhere.

B (24:19)

Yeah. When they weren't 100%. And I think that. I think it's. That it's not communicating the reality which is at the core of the trust issues. Right. It's passing something off for something that it isn't, which is. Which causes an awful lot of the worries and panic around the use of AI.

A (24:33)

So what I'm slightly worried about is, you know, you're this sort of issue around people being more and more disconnected from one another. We've talked a lot on the podcast about the importance of community. And you know that people end up sort of in their rooms engaging in relationships on sites such as yours and becoming more and more isolated from each other, maybe thinking this is a real friend and, and sort of a lonely person looking to this for some sort of solace when it's not real and it in the end won't give them that. I don't think it might do. But is that something you've thought about? It worries me A bit, yeah, Definitely.

B (25:13)

Something we thought about. I think the first thing to say is, at the moment, memory is a significant limitation of AI. So in terms of being able to engage in a conversation where you and this character are probably spending 12 months nonstop chatting, it can't happen. I mean, it's impossible. At a certain level, the context window runs out and the interaction kind of resets. So I'm not sure that, product wise, we're at the stage yet where that kind of addiction or engagement is entirely possible. That doesn't mean that we shouldn't think about it, because at some point that will be something that becomes possible and becomes a threat. We do look at the flip side of it as well, though. There is sort of studies that suggest that, take ChatGPT, for example, that bizarrely, people sometimes feel more comfortable talking to and expressing emotions and feelings to a machine than they might to somebody in the real world. And so I think at the same time, these models can play a role in allowing people to express themselves, to build up confidence around interactions, to have conversations they might not in the real world. So I don't think we can ever eliminate the risk that somebody is going to enjoy our platform so much that maybe they spend more time than you might think they should interacting with.

A (26:17)

It's not for me to think whether they should or not. I mean, it's their life, not mine.

B (26:22)

Yeah, just.

A (26:23)

I was concerned about the overall sort of risks here. I was just. I'm worried that people become totally absorbed in this and lose sight or not lift their heads above the screen and see what's going on in the wider world.

B (26:38)

Yeah, I think, you know, I think there's a. A threat of that and definitely across the industry, more broadly, I don't.

A (26:45)

Yeah, yeah. I mean, that's not. Not exclusive to what you're doing, Felix. I don't want to give that impression.

B (26:49)

No.

A (26:51)

So you've raised 10 million.

B (26:53)

Yeah.

A (26:54)

Pounds or dollars?

B (26:55)

Dollars.

A (26:55)

Dollars, yeah. So far. So do you have venture capital investors or your family or.

B (27:02)

Yeah, mostly venture capital investors.

A (27:06)

Some big investors.

B (27:07)

Yeah, we've got some big investors. I mean, we have particularly big investors in the AI and crypto space. Generally, funds that operate in that space have slightly looser vice clauses. So you probably know that a lot of traditional institutional investment in the UK will have vice clauses that protect against them. Investing in certain industries might extend to gambling, certainly extends to adult. But family offices and funds that operate in other sectors don't have quite such stringent vice clauses. And that tends to be the place that we can shop so within funds in that section. Yeah, absolutely. Some of the biggest names within that space are backers of ours and then also a collection of angels as well. We probably have seven or eight angels on our cap table who may not have been able to invest through the fund that they are managing partner of or a director of for the vice clause issue that I mentioned, but have sought to make a small personal investment themselves just because they believe in the potential of the business.

A (28:00)

Right. And you have assembled a team or in a process. You said your CMO came from ownership.

B (28:05)

Used to be global head of marketing creators there.

A (28:08)

Where else you have you recruited from? Senior level.

B (28:12)

So at a senior level we have two senior team members from pornhub. So former product lead. So live our product lead in the business. You run product there for over five years. Yeah, we are head of growth. Ryan was previously there in another life. We have a lot of tech developers from. There's a big adult business called Byborg who are operating a lot in the AI space.

A (28:34)

So how did you find all these people as a recruiter, I'm curious, who did you ask? Recruiters.

B (28:40)

We have used recruiters.

A (28:41)

There are agencies that specialize in.

B (28:44)

We have used recruiters but particularly for the. For the more senior executive level positions. And then generally what we find because we have shopped in pools where we've brought people in from big organizations that they come in, they want to fill the rest of their team and they generally know people within those businesses that they poach effectively.

A (29:00)

Is it problematic having everyone remote? Feels to me. I mean I like going into a place where people come together. Yeah, I think you're going to change that. Stick with that.

B (29:08)

We had a lot of conversations around that. I think better give and take. So what's harder? Culture's harder. Obviously communication is ten times harder. You know, the proactive work you have to put into communicating changes or vision or little day to day things throughout a remote team is far greater than if you're all operating within the same in person space. You're all hearing the same conversation, you're meeting the person at the coffee machine and talking to them. And given the communication, in my view is the lifeblood of any successful business. It's something that we struggle with. But on the flip side, like I mentioned earlier, talent density as we might describe it, is a really important thing for us. And so not being limited by a physical location, particularly on the technical side allows us to access the best talent and the best person for the job. So we've chosen to Prioritize that and try and cover some of the limitations that a remote culture or business do create. But it is harder.

A (30:08)

But it sounds to me like you feel you can really scale this up using technology with still a relatively small team.

B (30:14)

Yeah, absolutely. I really think so. Yeah.

A (30:16)

And your chief, your chief revenue will be from subscriptions imagining, is that right?

B (30:22)

Correct. Yeah. So the business model within the platform is a subscription model. So you subscribe on a monthly basis to either an individual.

A (30:29)

How much does that cost?

B (30:29)

9.99.

A (30:31)

Right.

B (30:31)

On average. So we at the moment have three pricing tiers, although we're condensing that into kind of one 9.99 tier that's on the kind of platform side. And then on the API side, which is our B2B side of the business, you effectively pay for use of so of the software. So if I'm another platform and I'm using our AI models to power content experiences, every time I want to generate a text message that gets sent to a user on my platform, that costs me one cent.

A (30:57)

Do you have much business doing that?

B (30:59)

An awful lot, yeah. Hugely growing. I think for us over the next six months that's going to be the boom area for a couple of reasons. And our business there sits in two different categories. So you have your existing AI businesses. So people that are doing similar ish things to us or trying to use AI to create content, and the reason they want to use our models is if they feel what we have is better than what they have and that might be quality, it might be stability, it's often speed because speed directly correlates to price. So existing AI businesses might use our AI models instead of their own because all of those different considerations.

A (31:31)

But to do what?

B (31:32)

For instance, so to power content. So for example, all of our digital twins, so Jordan, Karma, Electra, some of the ones I mentioned earlier, they can all through our API live on all of our clients platforms as well. So actually let's say you're a content site like OnlyFans, of which we are, we are speaking and are working with some already our AI characters can automatically live on that other platform, interact with fans there as they can on our platform, create content for fans on their platform as they do on our platform. So that's one way of interacting with us directly, taking our characters and the ability for users to interact with them and then there's also being able to do the same, but for them. So if I'm a real world content platform like an OnlyFans, for example, and we work with a Lot of these types of businesses, we can create AI versions of their existing real world characters that can effectively then populate their own site. They might just want to create images using the image model, they might just want to use our text model to have. To allow users to have conversations. So it's a whole different format and way of using the software. But I think for us, we really see over the next six months that this is a huge boom opportunity because everybody's looking to use AI and explore how they can integrate it within their business.

A (32:41)

Sure.

B (32:41)

Particularly on the content side, for some it probably won't work, but I think they're keen to do the experiment. For others, I think this will completely change their business models.

A (32:48)

So that 999 that the customer's paying per month, if you were to divide it up, you know, like a pie, how much of that goes to the creator, how much of that goes to the platform, how much does that go to other API?

B (33:02)

What's the sort of 80, 20 split? So creator gets 80% of the net revenue, we take 20% the net revenue, meaning we take the cost of generation. So every time that the user sends or wants to receive a text message, it costs us money basically, relatively.

A (33:18)

That's because you're using the chat GT.

B (33:21)

That'S because it's our own models. But every time you want to create a text, we effectively have to pay a GPU, which these models to use some, some generations like GPU.

A (33:30)

So how much of that 999 would that typically consume?

B (33:33)

Depends on the usage, but a relatively nominal amount, I mean, you know, for. You might be looking at.

A (33:38)

So people are burning through that?

B (33:40)

No, no, I mean, and again, at scale, the cost should be relatively cheap. So I mean you're looking at sort of 0.01 or 0.02 cents for a text message, for example.

A (33:49)

Right. So for the 999, what you, you would, what would you allow in your sort of planning?

B (33:53)

I think in a good, I think for a good, relatively engaged user, they might use, you know, right now a couple of dollars of the 999 in generation. Right.

A (34:01)

Cost, and then the rest would be 799.80 20 split.

B (34:06)

Yeah. And, and actually, you know, some of our whale users will obviously use an awful lot more of that. But like whale users, that's what we call them, the whale users. Exactly.

A (34:14)

Generally our top five, are they using it quite a lot?

B (34:17)

I mean today in terms of most pop. Most highest paying users, you know, they might be spending, you know, 700800 sort of quid so far since we launched. So that works out at what kind.

A (34:29)

Of 9.99 is PER.

B (34:34)

So two forms. So you can get. You can. Sinbad the sailor and all of his friends. So all of our original characters, of which there are hundreds, male and female, for a flat fee of 9.99. Or you can pay per digital twin, much like you would on OnlyFans. On OnlyFans.

A (34:47)

So if you want. So you get Sinbad and friends for 990.

B (34:50)

Yeah.

A (34:50)

But if you wanted to upgrade to.

B (34:51)

Get Jordan, you have to pay 999 for her as well.

A (34:54)

You'd have to pay that again.

B (34:55)

Yeah, exactly.

A (34:56)

Oh, so some people might be spending 100amonth.

B (34:58)

Yeah, yeah, exactly.

A (34:58)

If they're different.

B (34:59)

Yeah. I mean, the whale. Whale users would be most.

A (35:03)

So you already have whale users?

B (35:05)

Yeah, as we turn them. Yeah, right, yeah.

A (35:07)

I'm just wondering what sort of people are.

B (35:09)

So.

A (35:09)

So they might be on there for hours and hours.

B (35:12)

Yeah, yeah, absolutely, yeah.

A (35:14)

Do you know what sort of people they are? Mainly men, I imagine. Are they mainly older men, younger men?

B (35:19)

So generally the demographics a little bit, we're getting. We're getting more data. I mean, one of the challenging things is in terms of some profile elements, like location, for example. Google Analytics will tell us quite a lot of it, but there are some profile functions.

A (35:33)

So.

B (35:34)

Yeah. Where are these people? Well, us would be 60% of our users generally. Tier 1 English. So this, the product is speaking, do you know? I don't know. I have to come back to you on that.

A (35:44)

That's interesting.

B (35:45)

UK 25%.

A (35:47)

Right. Which part of the UK?

B (35:48)

15 all over. Pretty even spread, I think. Age demographic, 35 to 55, 60, generally. In terms of paying customers, in terms of users, younger demographics. So anything from 18 to 35, we see high user numbers, high traction coming to the platform, high traction using it for free. But generally they're slower to get their wallets out, which isn't really a surprise because disposable income is lower and that sort of thing. So the user correlations from that perspective, I wouldn't say are massively different to most sort of B2C businesses in terms of where the value lies.

A (36:20)

Yeah. So you worried, I mean, you're going to get regulated because, you know, AI is a new. You know, you're in those frontier pioneering, potentially.

B (36:30)

Yeah.

A (36:31)

Bureaucrats will catch up with you.

B (36:33)

Definitely. I mean. And let's look at age verification as a prime example. Right. So as part of the UK's Online Safety act, they obviously recently extended that for all producers of adult content or over 18 content, they had to do formal age verification processes and either that comes through the user having to upload their ID to the site to verify their age or through an approved facial recognition provider. So Ofcom have a couple of approved ones that UK businesses can use. That's a massive challenge for.

A (37:00)

I imagine a lot of people don't want to necessarily upload their id.

B (37:03)

They don't know. So we're fortunate. So who does that affect? So I think cheap sites. So these are sites like pornhub who make their money through advertising, for example. So all of the content's free, you can come on. The first thing you see is over 18. That's a challenge for them because in order the rule stipulates at the point of viewing that over 18 content, you need to have performed an age verification process. So for them that's a, that's a huge barrier to users coming onto the platform. Because who wants to upload their ID to a site like that, really? I think a lot of people worry about that, data protection and all that.

A (37:33)

Where's it go after that? Yeah, correct.

B (37:35)

For us, I think it's slightly different because nobody can access over 18 content on our site until they have an active paying subscription. So we're able to put age verification procedures in after the point of payment, before the point of seeing the content. And generally what that means is we can significantly lower the churn at that stage. Because I've just paid 9.99, I'm not going to not, you know, go through the final steps of the age.

A (37:57)

But if you find out this person's 15, you'd have to refund them, wouldn't you?

B (38:00)

You would, yeah, absolutely.

A (38:01)

And you do that.

B (38:01)

And you do that. Yeah. Through the payment processes. Yeah.

A (38:03)

Right. So. So, so when you get to the age verification, how do you do that?

B (38:08)

So we use a facial recognition provider. So it's actually called Yoti Plug for them.

A (38:13)

Yeah.

B (38:13)

Approved by Ofcom. Approved. So effectively you press a button, you take a scan of your face.

A (38:19)

Right.

B (38:20)

And then it will estimate your age and then allow you to proceed from there. And that's a UK specific thing. So in terms of your reg point, UK from an age verification specific.

A (38:29)

Still put your face into it.

B (38:31)

Yeah.

A (38:31)

They've got a picture of you sort of thing. Yeah.

B (38:33)

Although generally, I think behavior wise, particularly amongst young people, that's quite a. That's quite a well trodden path.

A (38:39)

People are worried about that.

B (38:40)

So like banking, for example, for our business, banking App. If I have to add a new payee to our business banking app, I would have to scan my face.

A (38:48)

Yes.

B (38:48)

To be able to. To, to be able to get that approved and uploaded. I do the same when I log into my iPhone. So I think people are quite.

A (38:54)

But I imagine you have to keep that data so you can say, well, this person is.

B (38:57)

So we. Yeah. So we don't keep it ourselves. We don't store any of that data. Yeah. The verification provider would store that. They'd have to store it for a period of time. And after a period of time it's deleted.

A (39:07)

Right. Okay.

B (39:08)

So it's not forever.

A (39:09)

And then I understand in some countries, some jurisdictions, I think Denmark is one where you, you have copyright over your own appearance. I suppose.

B (39:19)

Yeah.

A (39:20)

And. And that people can't earn money from that independently of you. Would that affect your business model? On.

B (39:27)

Don't think it would affect it. I generally.

A (39:28)

I think it's given by permission.

B (39:30)

Correct. I generally think it's a net positive.

A (39:32)

Yes.

B (39:33)

And it's one of the reasons we, we Creators and ownership. One of the previous issues with other tube sites or adult sites has historically been the creators don't own their own content and they tend to find that their likeness is basically being abused beyond them. When they stop creating content, they move on to a different part of their life. Their content is still monetized by the platform and there are loose laws around ownership. So that's actually one of the reasons we wanted to operate the model in the way that we do, which is the creator. It's a licensing agreement. They license their likeness. It's all consensual. They own that twin, they want to come off the platform, the ability to create content on the platform stops and they can take it down whenever they like. So for us, I think it's. It just plays into the business model we've already created. But for others that might be trying to approach the same industry in a more nefarious way. Yeah, I think that's a massive challenge.

A (40:18)

So some of the creators earning good money out of this already starting to.

B (40:21)

Yeah. And not onlyfans and money yet, but people and onlyfans who are earning absurd money. But certainly for the scale that we're at.

A (40:28)

And that's all about scale in the end.

B (40:30)

I think so, yeah, it's all about scale. And I think what I spoke about, allowing creators to create their twin without human intervention, that's a huge opener up of scale for us. Because at the moment I'm trying to think of a way that we likened it. What we're doing at the moment is it's almost as if Apple had made every app that exists on the app store themselves. Right. Like it would never have blown up in the way that it would have done. It's not a scalable business model. So I think we're about to experience that scale. But some of our creators are making really good, really good money, kind of high four figures a month in the first few months of, of them being on the platform. And I think it'll only go up from there.

A (41:04)

Right, right. So what, what do you see as challenges? What, what's on you? What worries you? You know, you're moving fast, you're growing, what worries? The future is quite unpredictable. Yeah. What's on your mind?

B (41:18)

I think, I think for us, AI is such a hyper competitive space. We don't own up, we don't. We haven't built our own foundational models as I mentioned, they're our own fine tune on top of existing foundational models. So what that means is that we're relatively at the mercy of what's called the open source community. So when sort of AI houses open source and release details of their models, we need to be the fastest and most effective to fine tune, train and deploy the next best thing to ensure that our infrastructure stays ahead of us.

A (41:44)

So which ones are you using?

B (41:46)

So we would use on the LLM side a model called Llama.

A (41:51)

Right.

B (41:52)

We would use on the image side a model called Flux. Our voice model is actually third party. We have an enterprise grant from a third party. And then our video model, which is probably our next big development, would be using a model called one, basically.

A (42:05)

And these are all open source.

B (42:06)

These are all open source models. Exactly, yeah. So staying ahead of the competition for us and ensuring we're always ready to test and potentially adopt the next best model is of critical importance to us for any business that's essentially venture funded. Runway deployment of capital, how you balance Runway with sort of maximum value extraction, that for me probably is a day to day consideration when it comes to thinking about team budgets, marketing, spend, all that kind of thing. Because we want to operate in a business where we are raising enough to be able to get us through a certain period of time, give us enough time to move to the next level. But you're at your most vulnerable in the early stages in terms of giving away your equity and the value of that. And so also trying to marry that security with what's realistically the shortest time that we can try and prove the Most amount to investors, to anyone else, either through driving to profitability or driving to significant value enhancement to get us to the next round. How can we play with that line? And so I think that's always a consideration that, that we think about day to day.

A (43:12)

And what are you doing in terms of marketing this to make people aware of it, to build your audience?

B (43:17)

Well, the creators are a huge thing, as I mentioned. So creating content that they can, that they can push and post on their own social media platforms and to their users is, is a big thing. Affiliates in our space, affiliate marketers. So these are effectively people who are often work independently and they get paid or remunerated each time they bring a paying customer and use it to us. And there'd be millions of those around.

A (43:37)

The world that would like agents introducing people.

B (43:41)

It's sort of like that. Exactly, yeah. And they would do so through driving traffic from other sites, for example. And so that, for a business in our space, is of massive, massive importance. But also always social media content where we can through meta, through the existing tools.

A (43:57)

So do you have a content team?

B (43:58)

We do, yeah. We have a video team. They're already actually for a lot of the video they're making because it's all AI. Right. So we have to use an awful lot of AI tools to be able to create videos of our digital twins and of our characters. So we have a content team, social media team in house, growth team, and then external support around things like SEO and that sort of thing. But I think the creators are and will always be our number one route to growth because they bring the audience, generally, they bring an engaged and a vetted audience. And if we can encourage and incentivize them to push content through those channels, that still and always is our maximum sort of bang for our buck in terms of users.

A (44:35)

Yeah, I mean, this seems to be a sort of area of British expertise. Is that fair? Because OnlyFans, obviously UK. Yeah. I mean, is there something going on here that makes this a good place to start a business like this?

B (44:51)

Yeah, I do.

A (44:52)

I'm curious.

B (44:53)

Yeah, it's a very good question. I don't know.

A (44:56)

There's a sort of cluster of expertise.

B (44:58)

I think there is. I mean, I think OnlyFans kind of started that. I think OnlyFans started the appreciation of the creator and the power of a creator. And what that means is that not just in the uk, but all around the world, including over here, we have, you know, millions and millions of people, hundreds of thousands of people who are pursuing content creation as A as a path, a career path of their own. And so I think perhaps being British based only fans starts a natural network effect of that domestically and in the uk. But it exists around the world. But probably a coincidence, but it's a curious one. I'd agree.

A (45:29)

I'm not sure it is a coincidence. I mean I think there's a lot of creativity here. I mean there are two ways of looking at it.

B (45:35)

Yeah.

A (45:36)

And as a sort of giving people the opportunity to be self employed and entrepreneurs in their own right. I mean that's huge. And brilliant's always been encouraging.

B (45:45)

Yeah. And encourage creativity and encourage pursuing your own thing and building a business. That's definitely true.

A (45:50)

But this, this one sort of troubles people in some aspects. But yeah, I still think that those, those the consistent aspects that identifiably British in many ways.

B (46:01)

Yeah. Well, I think identify British in terms of looking for opportunities. So like if we go back to the genesis story of how this happened.

A (46:08)

I didn't ask you that. How did this happen?

B (46:10)

Yeah, well, I always, I always wanted to always be. Wanted to be an entrepreneur and do my own thing. My dad was an entrepreneur and so I've always done my own thing and so it wasn't a great love of God. I've always wanted to create a business in this space. Why don't we go and do it? It was a pretty strategic decision. So my business partner and I, we were both in other jobs when the lockdown hit and Covid began and I was on furlough. He had actually left.

A (46:32)

What were you doing? What was your job?

B (46:34)

So I had a student housing business, my own business for about three years which we ended up closing. I was working for a cosmetics business which was backed by a previous investor of mine from another business. Really doing something different, working for somebody else. But I think re energizing and thinking what's next? And I wasn't enjoying it, particularly the lockdown hit. And my business partner, who was in between jobs and kind of stranded a bit because he'd left a PR company was moving to a big firm, lockdown hit, the big firm pulled the offer and so he was kind of left in a market where finding a job wasn't particularly easy and there was no safety net of furlough pay or whatever it might be from existing employment. So he put a LinkedIn post out basically saying do any other marketeers or brand people want to get together with me and do some pro bono work for charities or businesses fighting the COVID effort? And that got 100,000 views in 24 hours. He had 5 or 6,000 people sign up and he called me within about 12 to 24 hours and said, listen Felix, this thing's kind of taken off more than I thought it would. Would you like to build it together? And so I said, sure, absolutely. I'm on furlough, I'm not one to lie around much necessarily. So we did it. So for six months we built an amazing pro bono business helping, you know, charities in domestic violence space, NHS charities, COVID vaccine, raising money for them, publicizing them, all of this kind of stuff. And we had a network of volunteers from comms director at Disney volunteered some of her time right down to people coming out of university who are looking for something to put on the cv. And after about six months, bigger businesses came to us and said, you know, we love the story, we love the work. If you guys wanted to do this for profit, give us a shout. So we did. So we set up our marketing agency, look after group and that was a profitable business, grew relatively well. We had some pretty good high profile clients both here in the UK and abroad. But I don't think it was, it never felt like the end game for us. We really had the appetite and the desire to move into a high growth tech business that we felt had truly global scale. And we had a lot of businesses, as you can imagine in a company like that come through our doors asking if we can pitch for work or we're struggling with this. And our vision for going into the marketing business was this gives us an excuse to be in any room anywhere in the world to learn about different sectors, to learn about the challenges and the issues, what drives them, what doesn't. And from there potentially that big high growth idea will come. And we both shared this distinctive view that when something chaotic like a global pandemic happens, the whole narrative and shape of things shift. And my feeling going into it was if I can't come out of this period back on the path to the big high growth business idea that I think I can and want to build, then I've done something wrong, it has to be now or never. So we built the agency. We did that for about two or three years and then we just got chatting and we said, you know, we've done this for a while, we've thought about different spaces. If we were going to do something, where would we do it? We knew it would be AI. We looked at adult and we felt that if you look historically, adult has always been the trailblazer in the adoption of New technology. So the first ever online payment on the Internet was taken for an adult business since probably the advent of the tube sites. OnlyFans was probably one of the only people to really innovate and shake up the space. But they struggled massively with regulatory issues and challenges. And so I think that we felt that if we could combine AI crypto with some of the infrastructure elements of how the business can be run and monetized and trying to navigate challenges with traditional payment providers and stuff, that through that we could overcome some of the barriers to entry. And then we had a blue ocean of opportunity in terms of an industry which has some of the highest user numbers in the world and one that is relatively under innovated in. And so we took a calculated step to explore something in that industry and the way we did it was my business partner nick. He spent 12 months just researching, immersing himself, talking to people within the industry. And during that 12 months I continued to run the marketing agency and effectively fund this initial research, then development of an MVP and then building of the actual business. So it was a very gradual process until we got to which stage at which we kind of said listen, this is really exciting and this is the time to jump. And I jumped ship over to join them and did this.

A (50:48)

So does that marketing agency LookAfter still exists?

B (50:50)

It doesn't. We'd love it to. But like any business, maybe you were, I'm sure you work with agencies that read, you know, we were sort of peaked at 11, 12 people, so relatively small but with some great clients. But fundamentally service businesses are about people and we were very honest with a lot of our clients to say, listen, we've got this other business idea, we're going to go and pursue it. What are your thoughts about that? And they were very sort of complimentary and encouraged us to do it. But I think most of them felt that, listen, if you guys aren't there to run the day to day and build this business, that's the reason we work with you in the first place. It's not quite the same offering, sure, maybe we had, but I'm pleased to.

A (51:26)

Say that you shared this idea with them, did you?

B (51:28)

We did, yeah.

A (51:28)

And they were encouraged.

B (51:29)

Some introduced us to investors. Yeah.

A (51:30)

Did they?

B (51:31)

They did. And actually I'm pleased to say that some of our team, three or four of them, ended up going on and bringing some of our existing clients on there as either freelancers or to new agencies that they went to. So continuity existed, even if we ended up stopping it.

A (51:43)

No, that's fascinating. Is there anything you, you feel you haven't shared with me that you should. The listeners want to know about your business or why you're doing it?

B (51:51)

I mean, I think one thing that I would say is what's in it for users? Right. So we've spoken a lot about creators and how they can, how they can be given an asset that can help make them money and help them scale their businesses. But why are users going to use this instead of a lot of the content that they can already consume on a site like OnlyFans? So price is a consideration. You know, those that haven't used onlyfans won't know, but those that have know that creators can charge relatively extortionate prices for access to individual pieces of bespoke content. From a pricing perspective, ours is incredibly competitive. And as I mentioned earlier, the content quality, the experience is not just the same. We actually allow you to create anything that you want served to you within 45 seconds in real time. So you can really have the content experience and engagement with the creator in the way that you've always dreamed. And that's a bit of a limiting factor of OnlyFans. So the ability to be served something instantly, the ability to be able to create whatever you want of that dream creator, of that person that you look up to or like, and the ability to do so at a competitive price is certainly something that's of interest to users. And then if we just look at, we mentioned Jordan, you know, Jordan doesn't exist anymore, but through the use of AI, we can bring back those content experiences that existed 20 years ago that don't anymore. So there's a lot that we're doing both in terms of enhancing the existing content experience that users might have enjoyed on a site like OnlyFans, as well as using AI to do something that isn't possible in the real world. I e. Re. Bring sort of bring Jordan back to life. That I think makes it an exciting proposition for users as well as for.

A (53:14)

Creators offering a sort of weird sort of immortality to eternal youth. Kind of weird that, that you can do that and, and, and, and that conceptually that could continue long after a real person has left, I think so life entirely.

B (53:32)

I, I think so. I mean that, that gets into very dangerous territory. Yeah, but it is interesting.

A (53:37)

Well, I used to study moral philosophy and it, it's sort of, that's a, an interesting question in that space. And I remember thinking with, with in vitro fertilization, you know, someone who's dead could conceive A child. And that became a famous case.

B (53:51)

Yeah.

A (53:52)

Diane Blood case subsequently. So, yeah, this is an interesting, this is for me, particularly interesting because it's all new.

B (53:59)

Yeah.

A (53:59)

I mean, you're pioneering new ground and it'll have, it will raise all sorts of other questions and experiences and developments that we cannot foresee. I'm sure of that.

B (54:11)

I think so. And you know, despite all the chatter around the world being AI, AI AI, we really are at the start in terms of genuine adoption and the scale and power that these models can create in terms of experiences. I mean, we haven't even mentioned video in relation to our business. It's something we'll be releasing in the next couple of months time, but video just adds an entirely new dimension to these experiences. And I think very soon, within a couple of years, you'll get to a stage where people will be able to create, you know, two, three, four minute long videos of anything that they want. And that level of personalized content experience. We talk about personalization a lot within our business. That's unprecedented compared to what simply an algorithm to filter you the right content can do. It's an entirely new layer of being of a bespoke experience. So I really do think we're just at the start of what's possible. And as AI continues to grow and we continue to innovate the way users can engage with creators and create content, it's going to be, it's going to be crazy. And it's going to be hyper personalized and served to you now in instant gratification.

A (55:07)

It's going to be crazy. It's going to be hyper personalized, it's going to be served to you now. That's those key messages. I think that that's a good point to conclude our conversation because there's a lot coming down the track, a huge amount of change. And I'm so glad you came in to talk this through with me because I've learned a lot.

B (55:27)

Well, thank you.

A (55:28)

And I'm gonna follow you and your business with great interest because I. I'm sure it's going to develop in all sorts of interesting ways.

B (55:34)

Well, thank you.

A (55:35)

I wish you all the best.

B (55:36)

It's been a pleasure.

A (55:40)

I asked two questions at the end. I ask all my guests. The first is, is Felix, what is it that gets you up on a Monday morning?

B (55:49)

I think the feeling that we're doing something that can genuinely change the world and can genuinely revolutionize an industry or the space that we're working in. And with this business, I really feel that I'm honest to some friends when they asked me, this is your third business, really what's different about this one to others that you've built before? And not only have we already brought it to a scale far beyond what I got to with my previous businesses, this one I genuinely feel, and no, we're holding the ball right now. We really can be the ones to completely revolutionize and disrupt this industry. And it's a multi billion dollar industry and I've not felt that feeling in my gut every day like I do now with any of my previous ventures. So I think you've got to wake up with a yearning, with a genuine excitement to open your laptop, to sit down, to begin going again and, and I have that. And since we've been starting this eventual thinking about it almost two and a half years ago or so from that day, I've never lost it. And it's just an inside motivation and guttural feeling that makes all the difference, I think.

A (56:48)

Okay, well the next question, which is from my book why you 101 interview questions you'll never fear again fits perfectly because it's where do you see yourself in five years time?

B (56:57)

I think from our space. I really believe that we'll have done to the kind of AI content and creator space what OnlyFans did to the creator and they're not safe for workspace generally. I believe that we will have revolutionized the scale of innovation that is possible through AI and how we apply these models to create content experiences which are something unique and that the world's never seen before and provided creators with an asset to be able to make the absolute most out of the value and the IP that they've made worked so hard to create through a tool that gives users a way to interact with them in a way that currently isn't possible in the real world. And I think we can very much be a household name in doing that and maybe in the way that OnlyFans has been.

A (57:39)

Well, I look forward to watching this space and I hope you'll come back and talk to me in five years time and we can see where.

B (57:45)

Definitely.

A (57:45)

Thank you Felix. Thanks so much for coming in.

B (57:47)

Thank you.

A (57:47)

Enjoyed that conversation. Thank you Felix for joining me on All About Business. I'm your host, James Reed, chairman and CEO of Reed, a family run recruitment and philanthropy company. If you'd like to find out more about Reid, Felix and O, all links are in the show Notes. See you next time.