2.8 million downloads. #1 podcast in the UK. Fully AI-generated Adam Levy used Claude to untangle the Epstein files. Now he’s proving it’s a repeatable business model 🎙️ New to streaming or looking to level up? Check out StreamYard and get $10 discount! 😍 https://streamyard.com/pal/d/4884228748673024
The YouTube video features a conversation between Matt Medved, founder of Now Media, and Adam Levy, founder of the Neural Broadcast Network (NBN). The discussion centers on the success of the AI-generated podcast "The Epstein Files," its impact on the media landscape, and the use of artificial intelligence in content creation.
"There's no media organization or newsroom on Earth that could have done what you did in the time that you did it." - Matt Medved
"I want to create information that is source-rich and authoritative." - Adam Levy
The conversation highlights a pivotal moment in media, where AI-generated content is reshaping how information is consumed. Adam Levy's work with The Epstein Files serves as a case study for the potential of synthetic media to fulfill gaps in traditional journalism, pointing towards a future where AI plays an integral role in content creation. The success of NBN could redefine the relationship between technology and media, leading to a new era of trust-driven, data-centric storytelling.
Okay, we are live. What's up everyone? Matt Medved here, founder of Now Media. Um really excited about this conversation with Adam Levy, uh founder of the Neural Broadcast Network. Uh this is just a super interesting story at the at you know, the intersection of media and frontier tech on the AI side. Um Adam uh launched a podcast called the Epstein Files. Basically trained Claude uh pointed Claude at uh the the files uh like 48 hours after it was after after it dropped and created uh a podcast that went number one in the UK, has more than 2.8 million downloads, and clearly has struck a nerve um and is turning I think kind of changing minds around the value of synthetic media and what uh an AI-generated podcast can be. And so, Adam, super happy to have you join the stream. >> Thanks for having me, Matt. Yeah, I'm glad to be here. Love NFT Now, been a follower since the beginning. So, thanks for having me. >> Absolutely. I know we've been we I've been on your podcast a few times. So, it's great to it's great to be able to to share you know, return the favor uh share the love. >> Yeah, TVT to Mint, to the good old NFT days. >> Yes, yes. We've got some great memories of Mint. Um we've got around 30 minutes. So, I there's a lot to cover. So, I want to dive in. Um let's let's real quick just for those who may not be familiar. I've been following your you know, what the trajectory of the Epstein Files and NBN um since the beginning. But, for those who are tuning in now and wondering what this is all about, give us a quick background on just how you decided to create a fully AI-generated podcast around the Epstein Files. >> Sure. So, my whole background has been in building owned audiences. And um that kind of like came to fruition with starting the Mint podcast. I grew my newsletter through that and then eventually merged to doing stuff on chain, to which uh you have a lot of experience in as well. Um I've been selling my podcast on chain. I've sold a ton of episodes and I guess like as crypto winter was hitting, I was curious to see the timeline switch uh into AI Twitter or Cloud Twitter around December and even a little prior, summer 2025. And yeah, I just started messing around. I mean, I think folks like you or folks like myself, like we have so many ideas, but historically we've been confined by our technical abilities to execute those ideas. And now with these tools, we can create our imagination. Like anything we kind of imagine, we could put to paper. And I was experimenting a lot during the holidays in December of 2025. And I initially built myself this app called Distill. And I'll try to pull up on my phone, but it's basically a PWA app. And the goal of Distill is to basically take a bunch of sources that I'm interested to learn more about, whether be TikTok videos or whatever may be, distill that information, fact-check it, and then create me an audio podcast, basically an audio file around it to listen to. I'm just like super ADD, dyslexic, like I'm like I consume best with listening on like two, two and a half X. And um when the DOJ dropped those files shortly after, I was like, "Oh, I could just use the info that I built for Distill, scale it a bit more, welcome more data, and then create this comprehensive outline on what the documents show, the court filings, the flight logs, um the emails, and kind of like avoid the talking head perspective between what the left and right are saying, what the conspiracy conspiracy conspiracy theorists are saying, and just stick to the data. >> >> And yeah, I I I'm just a person who constantly tries to put my ideas and bring them to life. And this was just another one. Was not expecting it to blow up the way it did. And um been really grateful and amazed by the traction that it's gotten. Um it's the first audio product to top the charts. That's a synthetic product to surpass legacy media. Um, it it did wonders. I think we're now at like 3 and 1/2 million downloads, something like that. Every week between all the series, we get about 70,000 downloads. Um, so yeah, I love this whole synthetic media angle. I feel like it fits perfectly with my love for data, building software, content creation, and media. So, yeah, excited to see where this journey takes me. >> Absolutely. You know, it it's interesting. Uh, I think there's there's understandably a lot of uh fear and skepticism around AI-generated media, synthetic media. Um, but I think The Epstein Files is a really interesting example and a bit of an illustrative example that I've used to help change perceptions and minds around both what the value of these of of this the format can be because, you know, in my eyes, it's like, you know, there's no media organization or newsroom on Earth that could have done what you did in the time that you did it and draw uh draw all of those connections between millions and millions of of documents and files uh and then presented it in that format if it were fully, you know, done with humans and studios and and researchers. It just would it could not have happened in that time frame. And so, it's a really it's really it's success is indicative to me that listeners are responding to something that they were not able to get from traditional media. And I'm curious like how you see that as, you know, the driving force behind its success and also as a blueprint for NBN. >> Yeah, I think people have lost trust in media. I think we've been saying that for a minute now. And um I'm I'm just building for myself, honestly. Like I like to consume information that has no spin, that is fact-driven, data-driven, and can tell me things for what they are. And I constantly find myself lacking that in the market. So, The Epstein files was like the perfect example of what I wanted to see, but nobody was doing. And same with War Desk. War Desk is documenting the day-to-day developments of Operation Epic Fury in Iran between the US, Israel, and the IRGC. And like I want to have my finger on the pulse on that. I want to understand what's going on. Um and that that that inspiration kind of spans across any any topic that I'm personally cur- interested in or, you know what, people at large are interested in. Like I guess a good a good um preview is like DOJ recently dropped Department of War dropped the UFO files, right? So, it's like that's already in the process of being developed. I'm going to be releasing that soon enough. So, these are like things that I I care about. Like I'm a I'm a data-first kind of person and I want my media and the information that I consume to reflect that. So, I think that is I think in a world where there's so much um like AI fluff and like content, I think it's important to create information that is source-rich and authority authoritative in that you can reference data as part of the underlying conversation, the sources, and do it in a in a in a way where you can trust what you're listening to. And that's really the the premise of like how I lead myself in creating these series um and just kind of building for myself through that lens. >> I love that. Um one thing I'd love to I'd love to kind of take take us back to after you released the Epstein files, at what point did you realize this is now taking off or like you this is this spark is happening, which like as a as a media entrepreneur, like when you have those those moments where that are feel like inflection points of attention, um it's very exciting, but it's also pretty stressful, too, cuz you're like I need to maximize this in the right way at the time that I in the way that I you know at the the time in the way that I can while I have it. So, what was that like for you when you realized that like the the Epstein Files was taking off? And how did that inform how you approached, you know, continuing to put out episodes, launching new series, et cetera? >> Yeah, I cited a lot of what people were reading, and to see the traction that it got, or not reading, but people were commenting um on every episode. And it was a first of its kind product in the market. Like, I wasn't the first AI Epstein podcast. There was a guy who was posting like three to four times more than me, had a deeper archive, was posting more consistently, but the quality was just subpar of what it could have been. Um and I wasn't even like doing a competitor analysis in that regard of like, where can I insert myself in the market? It was more just like, let me create a product that I wanted for myself. And when I saw it starting to take off, like within 7 days it hit 100,000 downloads. And that metric within itself, I mean, I I was hosting a human podcast for years. Like, I brought I brought I published hundreds of episodes, like long-form and short-form episodes. I never I never reached that scale, you know? So, to see it being done in a synthetic version was really eye-opening, and it's like, wow, not only have I been burnt out of being a human creating content. Like, at my peak, Matt, I was publishing 40 to 60 pieces of content a week across different form across different mediums. And I'm like, at some point I'm probably going to end up replacing myself. Like, let me just see if I can put this together, and and just learn from what people are responding in in commenting and in embody that, and see how I can take that to the next series. And that's what War Desk was. And then Rex the Gilgo Beach architect, not too long ago there was a Rex who was the Long Island serial killer. Like, he went on trial. I did a series around him. People that that's built a following. Um in fact, the one that I find really interesting, the one that you and I were talking about in the green room, like neurodivergent, like that's my favorite series, but it's the least uh least uh performing series out of the catalog so far. And there's different reasons why I think that is and it needs to optimize it. But, um I'm still learning, you know, what makes a hit a hit versus what makes something um kind of like die down and and lose its traction. So, I'm not going to come on here and say I have the perfect formula. It's definitely a work in progress and I'm learning as I'm shipping. >> Makes sense. Well, that aren't we all, you know, all of us playing with the AI tools and and doing that. I'm I'm curious what what happened some of those insights to you in terms of what like what has worked that surprised you or like what are some things that like maybe you thought this is going to be a hit and then it was a dud and what what is that learning process kind of showing you? >> Yeah, I think um the biggest surprise factor is the fact that people loved that it's AI. >> Mhm. >> Like they loved the synthetic voices and they loved the delivery of the content because it's kind of like a split of the sea. Like there was a camp of people that loved it because they could trust it and there was references of the files and there's people who hated it because they felt like I was taking opportunity away from humans. And um I'm kind of like in this camp of like let me just keep doubling down on creating the synthetic version because at some point that's going to be the norm and these models are only going to get better and better and better and people are going to learn that the media they consume on television, on radio, on their phones, whatever device we have in the future will end up being synthetic, like just by default. Um so, I kind of see it as like um going head-to-head with what people believe like their their thought process to what should be the norm of consuming content and who should be presenting that content. And I've even gotten those criticisms from legacy media companies. Like I did an interview with The Times. I was on their podcast. And they were grilling me as if I'm taking opportunity away from journalists and whether or not I consider myself a journalist and and I kind of just explained to the host what I explained to you. It's like I just want to consume media the way I want to consume it and that product falls short in the market today. So, I'm still learning as we go and a lot of the insights have been predicated on like, okay, I should keep doubling down on creating synthetic content because I do think that's the future and if I could do it in a in a trustless manner similar to how smart contracts operate like very inspired by the crypto days of like building crypto products. Like I want to build a trustless media engine. Like that's the vision. Where when people see an NBN production, they know it's rooted in fact and data and they could trust it by default and like that's going to be their go-to source of trust. Um and I think we're we're not too far away from that being the reality. >> Mhm. No, I I agree. I agree and I It is interesting to think about how this becomes more of the norm. Um as you know, as you've gotten this uh you know, kind of explosion in attention and also um you know, building now what sounds like a you know, there's a consistent number of of uh of downloads per week, per month, etc. How have you thought about monetization? Uh and and what what does that look like for NBN? >> I think um I'm not trying to innovate too much on monetization. I'm going to stick true to the ad model for now and I'm going to innovate more on the product that I build that ends up producing these series. So, not trying to get any like any too creative. Like there's already a tried and true model of monetizing media and podcasts and just going to piggyback off of that. >> Sure, that makes sense. Are you are you currently monetizing now or are you uh and and how how how did you kind of approach that strategy uh to like kind of ease into it if so? >> Yeah, my strategy is basically looking at what's hot and building a series around that and finding a sponsor or leveraging the existing ad networks like Megaphone for example that pair you with these ad pre-rolls mid-rolls and leveraging that to to yeah, ship ads and collect your your cents or dollars per download for for what it's worth, yeah. >> Yeah, it makes sense. Uh what what was the um you know, the kind of like the the driving force behind uh some of the like like the series like Neurodiversity I thought was like a really interesting choice because uh obviously Epstein files huge moment news moment where there's a ton of search traffic around it. Um same with, you know, the Iran war uh and I thought that that was interesting because it felt like Neurodivergent felt a little bit different because it wasn't like tied to a specific news event. It's more of this evergreen thing that affects so many people. You know, as I said, I have ADHD. So when I saw that, I was like I you know, respect. You know, I I want to check this out too. Uh I think a lot of people uh who deal with, you know, kind of uh these different um diagnoses but that can be can be classified as disorders but can also be classified as superpowers, right? Like um I just have have a bit of like a personal connection there. So I'm curious if, you know, that represents a bit of a shift in strategy and if there were any uh the response to that indicated anything about that. >> Yeah, I'm ADHD. I have dyslexia. Like sucked in school but somehow made it through and um when I saw the whole meme around Neurodivergent on the Twitter timeline, particularly rooted in Alex Karp from Palantir, um I thought there was a lot of interesting um Yeah, I thought it was I thought it was interesting to be able to look from a bird's eye view all the Neurodivergent people that ended up changing the world as we know it and understanding their decision-making in their career from the neurodivergent lens. And it's like the same gut feeling that I have with building the Epstein files or War Desk. Like it's not Like I see signals around me and then I kind of like marinate with those signals and use my own creative process to come up with what I think should be the next series and at this point I'm firing from the hip in many ways. You know, it's like I'm it's like I built this product that I could point into any category and it will produce a series and um programmatically like without my intervention for what it's worth. So, it's um it's like one of those things where I see signals around me and I embrace those signals. I sit with them, I let them marinate and then I kind of churn and produce what I think may be interesting and kind of from a place of what I would think is interesting, what I want to see. Um There's a There's other data that I that I I I built product around, you know, making more informed decisions based off what people care about on the internet as well that help kind of like guide my creative chaos. Um but I think that I don't know. I I think it's like one of those things where I have an idea, I embrace the signals and I fire from the hip >> Mhm. >> and just see what happens. Yeah. Like I'm so disconnected emotionally from the process that it in many ways it's a numbers game for me. And just running with that mindset. >> I love that. Uh I know you've you've mentioned it sounds like, you know, you're productizing it in a in an interesting way and and you're thinking about um you know, the the workflow that you use to create uh these series, um being able to um you know, create a business out of that. Uh I'm I'm curious you know, how you how you're thinking about that and also you know, it sounds like you've had some interesting reactions from the legacy media world, you know, the New York Times pod you know, the New York Times interview it sounds like there was a lot of skepticism around it and you had to kind of defend yourself. Has there been a lot of like interest like and and optimistic interest from from media companies as well and do you feel like there are media companies who are looking to do what you're doing as well. >> Yeah, my DMs have been flooded with a bunch of different comments and reactions from the camp of this is so dumb. We can't trust this like there's no room for a product like this in the market to this the most innovative thing we've seen in media in the last two decades um to everything that kind of falls in between those two those two brackets. Um Yeah, I feel like um everyone's wondering how they can optimize their their business with AI and I think from the camp of people who are scared of it those that argue that I'm stealing jobs from the art of being a journalist like you can now argue to some capacity I'm a journalist it to some to some degree and um like like the ocean is wide open, you know, it's like you either you either replace yourself or get replaced. Like that's like that's a mindset. There's nothing else to it. There's nothing personal about it. And um so like on the other on the on that bracket like I could be the victim and you know, complain that oh no, AI is replacing me and like I want to be sensitive to that cuz there's actually really like real examples of like people actually not being able to replace themselves with AI because AI is just taking jobs and that's a really that that's a whole different conversation and I'm totally aware of that, but I guess strictly focusing on the lens of audio and podcasting, frankly, I think if my goal is to create a product that entertains and that informs and if there's a better way to do that, why not fully embrace that? And I think that's the lens that the other bracket has is like, "Wow, this is really innovative. Like this is very high quality. Imagine if there's a way to mix the old with the new, what would that hybrid product look like?" Um So, yeah, it's a blue ocean. It's kind of like the rule book is written in real time, frankly, and um I'm getting feedback from all sorts of perspectives and what people think is right direction, but again, like all that stuff is noise for what it's worth. Like I'm so focused on where I'm headed. I've never felt more aligned with where I need to do with where I need to be and what I'm building and um I just have such a clear vision of where I'm taking it. And um people can yap and yap and yap and do whatever they want, but I don't know. I'm I'm so fired up. Like this is like the greatest moment in media, I think, ever. >> Hm. I I love that. How how would you describe your vision for it? >> The goal is to build a trustless media engine. Right now, there's an editorial approach approach in building these series and in the future, I want it to run itself. Hm. And um my approach is being product led. It's a product led media vision. NBN is a technology-first media company. I think there is um there is the There's a natural inclination to see what TVPN did, see the opening eye headline, and rush towards building a live stream because that's what you think the market has validated to be the direction for media, but I think that even below that meta, there's a even sub meta of doing product-first media production and building a technology first media company um that a lot of these creators actually are not well equipped to do, which is why they um yeah, like you're yourself, like you you understand systems. Like you're building these products internally. Other creators are kind of like struggling with how to even use AI and how to build agentic workflows and agentic products and let alone have that influence their current business operations, you know, whether it be media, consulting, whatever it may be. So, I think like the underlying meta is rooted in doing a product first approach to media and my vision is to build like this trustless autonomous media network um that essentially operates itself, runs itself. >> Mhm. >> And historically, if you look at media, it's been bottlenecked by humans. And now you can program that. Like it's a code first approach to to truth seeking, essentially. And um never have we ever been able to find that and do that. And when you think about the why now opportunity, like that's the why now. Like we've never been more better equipped to execute and ship a product first media strategy in the market that's rooted in a truth seeking approach. >> Mhm. Absolutely. You know, it's interesting because I I mentioned at the beginning it I feel like the success of the Epstein files and and some of the other NBN series like turned a bit on its head some kind of preconceptions around the podcast format. I know, you know, a lot of people say that, you know, podcasts succeed based on the hosts, right? People are drawn to the hosts and it's that personality. But I think what what this has demonstrated is that that may be true for certain categories, but the Epstein like you know, the Epstein files fulfilled like need. It fulfilled a space in the market for education and understanding what was an incredibly deep trough of documents that no one had time to like really sift through on their own. And I I when I first listened to it, I was like, I can tell this is AI, but it doesn't matter to me. I don't mind it, you know, and I think what you said, some people are even like, hell yeah, that's great that it's AI. So, I'm curious, like as you think about expanding NBN and building this network, what are those areas where you think that you think are are particularly well-suited for like an AI-generated series? >> Um I think AI is really good at at consolidating data. And um I think AI is good at Actually, AI is not good at truth-seeking. You have to build guardrails for it to be truth-seeking. Frankly, most LLMs are left-leaning and have a bias to them. Um and depending on which party you're in, you know, you might like that, you might not like that. I don't really care for that. Like, that's not what my vision is, you know, like I want to I want to be truth-seeking. I want to be as truth-seeking as possible and then be very vocal where the truth falls short with what I have at my disposal. And um I think AI is uniquely positioned to outperform on cost, cadence, and context. And with that, there's an interesting intersection of finding new series and creating new series that humans couldn't possibly do prior. That is just uniquely positioned for AI. >> Yeah, I love that. >> Yeah. What about you? What do What do you imagine being like the intersection of that? You're a media guy. You You guys have done a great job building a catalog of of articles and podcasts and video. Like, what what's What's your perspective on that? >> Yeah, I mean, I I I think like for example, I don't know that an AI-generated like sex and relationships podcast would necessarily do so well because those to me, like the personality element of the host for that is so important. That's what people are kind of drawn to. What But where like I found that like I'm really drawn to the podcast format is education. Uh and and and when it's educational, it almost doesn't matter who's delivering it as long as it's being delivered in a in a format where I can easily digest it, understand it, and you know, and it's it's also efficient, uh you know, it in terms of conveying that information. And so, to me it's like that's an interesting frame to look at it because it's like okay, well, what are those areas that are not currently being served by the market because um the reality of journalism is everybody is doing more with less, right? Like newsroom, you know, newsroom staffs have been slashed, uh editorial content budgets are not what they were, AI is cannibalizing search, you know, it's just there's we're going through a real shift in media, which is why I like when you mentioned that some people have been giving you um grief over the fact that you're, you know, quote unquote like taking jobs away from journalists. I don't know that those jobs are really even there. That Do you know what I mean? Like I don't know that there was I don't know what there is a media company that would have been willing to put up the capital and the and the human resources to do what you did in the time frame you did it for the Epstein files. Do you know what I mean? And so, that's what's interesting to me is thinking about, you know, what are some of the areas that are not being served by the market, but that there is an attentive listenership. And if you were to serve it and be able to produce it, um you know, at scale, uh there there would be an audience and for it and value created. It's something worth thinking about. You know, we we we may launch an AI-generated podcast soon, too. And you know, I think that's certainly not going to be our our main our main format, but it's something that I think is really interesting if we see like an an interesting niche where we're like there's there's a listenership for this, but there's no one really equipped in the market to provide it right now. Um and so, I I that's why I do think I I've I think this is I'm I'm a I'm obviously an unabashed techno-optimist, you know, I I believe technology it it always wins. And to me this what what this is actually a lot of value creation because it enables you to tell stories and and and serve audience subsets that there may just not be a viable market for with with human resources and the costs and and labor that are associated with that. >> Yeah. That's That's a great point. I wonder if you've seen an opportunity in creating these in the crypto space or in the digital art space. Um maybe a discussion for a different time, but curious how you see that. Yeah. >> Yeah, I I think there is an opportunity. I mean, to me what I think is also kind of critical is a lot of these spaces move really fast, right? So, you and you know, what I what is interesting about the Epstein files was that you know, obviously there there have been many releases of documents. You you have a lot to work through and there's enough that you've been you've been you've you've continued to put out new episodes um at a regular cadence, you know, over the past over the past few months. Um but there is kind of like this evergreen element of of like this is the history of what's happening and this is what happened. Whereas like the crypto space moves on a dime, right? It's like, you know, coins rise and fall and uh and some of these like really and I think it also has such a short attention span >> Yeah. >> that it's a little bit tougher to like capture the the zeitgeist. Um but I suppose if if there's any way to do it, it is with AI tooling that can quickly produce something. So, it's not a it's not a it's it it I think it's just really about finding the right opportunities there. But I I do think they exist. And then I think in the digital art space there's a lot of evergreen storytelling that can be done because, you know, and there just aren't that many people telling those stories. >> Totally. Yeah, totally. Yeah, I mean, it's a fair perspective for sure. Yeah. >> I I know we're kind of we're coming up towards the end of of the time, but can you tell us anything about um what what's in store for NBN? What what are your big priorities coming forward for those who want to follow along and get involved? Like, what should they be looking out for? >> Yeah, I'm hyper-focused on shipping new series. So, if you want to follow along, you can go to nbn.fm. Um NBN stands for Neural Broadcast Network. Um you can see all the up-and-coming series, the live series. And my one ask is that you subscribe to the newsletter because that's how you're going to be informed of upcoming series. So, nbn.fm/newsletter, or if you go to the site, you'll get a pop-up within the first few seconds of landing. >> >> And um yeah, follow along, and thanks uh thanks in advance for the support. >> Yeah. Uh this has been a great conversation, so uh Adam, really enjoy it. Uh look forward to to checking back in and seeing how how everything's going, and hopefully having some more conversations in the future. >> Thank you so much. Thank you. >> All right. Take care, man.
