Why OpenAI Bought a 17-Month-Old Podcast

We had a very interesting history with OpenAI in that we were sort of contrarian supporters when it was initially rumored that OpenAI was going to put ads in ChatGPT. The general response was, "This is bad. We don't like ads." But we are the guys who love ads, and we think ads are a really powerful business model, and we think ads can make things free. If you want to be OpenAI, you're going to need an ad-supported version of your product.

Long time ago.

We met through a mutual friend who connected us because we were in a similar stage of life. I guess we'd both started companies in Silicon Valley, raised money from actually two of the same venture capital firms, never met, but we both were living in Los Angeles, had kids, and were sort of looking for what was next.

And LA is a terrible place to live if you work in tech outside of hard tech. And we didn't see ourselves building a hard tech business. We both had some media background. I built an influencer marketing company in college. John had a big YouTube channel.

Yeah.

And we did what had never really been done before. Two guys in tech decided to start a podcast. It was revolutionary.

Um, no. So, we met, we just loved talking. We would call each other, even though we didn't even know each other that well, and just talk about different ideas and what was happening. And eventually decided to record one of the conversations. We sent it to a guy named David Senra, who has David Senra by David Senra, a popular podcast, and also the Founders Podcast, if you've listened to that.

And he's one of the strongest believers in the power of podcasting — that like we're still early, that it's not too late to start a podcast if you have a good idea, something fresh to offer. And so he was extremely encouraging, told us that we should take it deadly seriously.

Yeah, we hadn't published an episode, but we sent him a Google Drive file. He listened to it. I was shocked that he listened to it because who has time for another podcast? But he did, and he said, you know, take this 10 times as seriously. He saw something in us, and you know, we kept going from there.

And so he sort of encouraged us to treat it like a business, like a startup, not like a side project. And that's something that we quickly identified as one of our key differentiators was that we had lots of friends and we'd been in these situations before. My YouTube channel — I was running it on the side while I was also doing a business. And there's a lot of people that have investment firms that also have podcasts, or they have a company and then they start a podcast to promote that company. And there weren't that many people outside of the traditional media. There weren't that many insiders that had transitioned and become full-time creators. Dark Cesh Patel obviously in that group, David Senra, there's a few others, but it was a smaller group than I think the podcast charts let on. When you scrolled through and you saw hundreds and hundreds of podcasts, most of those folks were not spending 40 hours a week.

It was and still is low status to some degree. Like there's — I remember when we would tell our friends in the early days, because we told people with like a handful of episodes, we were like, "We're going full-time on a podcast," and like we have supportive friends, thankfully. They were like, "Yeah, like that's excited for you. That's awesome." But you know, in their head they were just thinking like, "Wow, they're down on their luck, completely lost. Couldn't come up with a single SaaS idea. What are they thinking?" And so we started taking it really seriously, and that unlocked a lot of things, that unlocked interviews with bigger names.

Well, yeah. Before we get there, we should talk about like why we initially got any traction at all is because we weren't doing interviews. We did like 150 hours with just the two of us talking before we had done a single guest. And instead of having — there were shows that we love, like Acquired is, you know, an amazing product, but it's like one topic, and we would have, you know, we would talk for 90 minutes, but we'd cover 50 topics.

Yes. 30, 50.

So we'd read the whole Wall Street Journal, the Financial Times. We'd pull from random headlines. We'd pull up tweets and posts on different social media sites. We'd print out the posts. So we'd typically have a stack of posts, and we would pull them up and flip through them. And Jordi has one there. We would discuss it. Sometimes we'd just be laughing about it. Sometimes it would take us on a tangent. We'd talk for 10 minutes, but then we would clip that reaction to that post and then quote tweet the original post. And so that was sort of a do things that don't scale moment. It was a love letter to whoever posted that original idea or that take or whatever they posted. And as you know from posting, oftentimes the numbers get really big. You get a thousand likes on something, and it all kind of just melts into the background and it becomes just noise, and you see the number go up. But if someone quotes your post with a thoughtful commentary that's filmed in 4K with cinema cameras, and the hosts are wearing suits, and they print it — they took the time to print out your post — it's just, wow, these people went a lot further to say something. I'm at least going to see what they had to say about the thing that I fired off in the shower, right? And so that just told individual people — maybe 50 people a day got a little message from us: "Hey, we exist. We liked what you said, and here's our commentary." And then they followed us. They reposted it. They quote tweeted and said, "I can't believe this happened. What is this? This is weird. This is interesting." And so that was an interesting zero-to-one moment.

Yeah. Early lessons in going to zero to one was, you know, you need some element of format innovation. Like podcasts are incredibly noisy, incredibly powerful, but we came in with something that was a truly unique product because one, there was no one else that had us talking about crazy ideas for 90 minutes. We had a monopoly on that.

And then a real focus on marketing the show, treating the show like a product that we make every day, but we were incredibly fixated on marketing that product. And I think that happened through clips and, you know, the love letters concept, things like that.

Exactly.

I grew up in LA, studied economics in college, moved out to Silicon Valley as soon as I graduated, joined.

I grew up in Pasadena. I live there now, actually.

I went to Northeastern in Boston, and then I went out to Silicon Valley. Was sort of on a finance track post the great financial crisis. Was very interested in that, but then got bitten by the Y Combinator bug, reading Paul Graham's blog, reading Hacker News, wanted to move out to Silicon Valley, maybe join a tech company. Realized that a lot of the big tech companies I wouldn't be fit for the corporate ladder. Wanted to start a company. Started one company, teamed up with a Y Combinator company. The first business I started was called Soylent. It was a meal replacement shake.

Yeah. You heard about that? It was like a decade ago.

Yeah. Yeah. Not a great name for a food product, but that was the point. The point was like it was rage bait. It was rage bait. Like everyone had liquid death, that kind of thing.

Yeah, liquid death. Same thing. And so, yeah, I was interested in technology and working on software startups, but then we got pulled into this consumer packaged goods direct-to-consumer boom and grew that company a bunch, raised a bunch of money, eventually sold that company, started another company with basically the same team, but then got that company to scale and had the opportunity during COVID to sort of think about what was next. Started a YouTube channel and then eventually went over to Founders Fund to become an entrepreneur in residence, and that gave me a lot of free time to explore what was next, and that sort of set me up to be ready to go full-time on something completely new when the right pitch came my way.

We're not very good at being concise, as you can tell. We talk — we do this for three hours a day. We'll have hit the four-hour mark by the end of the day of podcasting. We really love this. But I grew up in the Bay Area. Was always obsessed with the idea of building companies, building startups. My first-ever company was a skateboard company when I was 12, because I figured out that the boards that I was buying in California for $35 would be a blank Canadian maple board I could get them made with my logo for $17.50 in the Midwest. And so I went and raised like $500 to do my first run. And so started selling skateboards. That was kind of when I caught the bug. Took me about a decade from there to build like a real business that could support my life. But I actually met a number of podcasts while in college. And this was at a time where you'd have like a podcast with an incredibly loyal fan base, tons and tons of downloads, and they'd be in the top 10 of, you know, whatever category they were in. And they'd be running like zero ads, which sounds insane, but they would just be doing it for the love of the game, and maybe they had another business. And so I went to them and I was like, "If I can bring you advertisers, can I take a cut?" And so I started doing that. And that over time turned into a company called Branded Native, which is still running today, but does like YouTube creator advertising for a bunch of different brands. And so I didn't raise money for that business. It got profitable quite quickly. And I started taking the profits and investing those in other startups. So then over the last however many years, I guess eight years, invested in like 70-odd startups. Eventually started a fintech company. I started maybe the most 2021 ZERP company ever, a company called Party Round, which was Venmo for fundraising and investing, because I would get invited to invest in a startup, and then I remember the first time somebody said like, "Well, I'll send you the docs, and then you can just wire after that," and I was like, "Wire transfer?" I'd never actually done a wire. I had never actually done a wire transfer at the time myself, like an outbound push. And so I was like, "Why is this not like Venmo?" A lot of VCs thought that was a great idea in 2021. As you can imagine, everyone and their mom was investing in startups. And had built like a fairly viral brand on X for that business. Ended up evolving it into a banking product over time that was building a stablecoin, basically a stablecoin bank. The month that we launched, FTX collapsed, which killed a lot of excitement that people had around crypto, as you can imagine. And then a few months after that, SVB collapsed, and people started really caring about how much FDIC coverage they had. We at the time had standard like 250k of coverage, and our average customer balance was like a million dollars, something like that. And so people are like — I was telling our customers like, "Yeah, there's a bank run that's happening. Like you need to be practical with your business." We ended up getting acqui-hired by a company called Ramp in New York City, and then worked there. And that's like when I met John.

Ready for this. We had been like — both of us had gone through the sort of like Silicon Valley factory, had some wins, had some losses, had been like, I would say, very humbled by work and life, right? Like I went from being like, you know, running a hot startup to basically like having my business blown up by like, you know, the market dynamics. And then I'm sitting there, I've got like a one-year-old son at the time, I've got another on the way. Like that's like deeply humbling. And so I think that based on like what we do and what we talk about, they expect us to be like, you know, maybe cocky or something like that. But I think we came at TBPN from a very — we had been humbled by Silicon Valley, been humbled by startups, and I think that has informed our coverage.

Right. Right. Like I think we give people — we always would give people the benefit of the doubt. We took a strategy early on which was called golden retriever mode, which is an idea that, you know, John has described many times on the show, but like being in a golden retriever mindset and trying to act, show up in the world, and at least that was our role with TBPN is showing up as like golden retrievers. You're like happy, friendly, and dumb. Yeah. You know, not trying to be, oh, I got this person, I'm smarter than them, I'm the smartest one in the room, always. Being, yeah, a little humble in some ways.

I introduced them. I introduced the founders of Ziotics. I love them.

Yeah. Connected the team.

Yeah. Yeah. Yeah. I connected the team. Stephen, one of the co-founders, went to high school with me. I absolutely love that product. We ordered a bunch.

Yeah. So it's a couple things. So one, because I started this company, Branded Native, and I had literally done thousands of like influencer podcast, YouTube deals. So I was obsessed with the commercial side. And early on with that business, if I would connect an advertiser with a content creator and the first few ads went well, I would immediately book out like an annual deal because I thought it was in the interest of both companies. I was like, the advertiser wants to allocate budget effectively. The content creator wants predictable revenue. And so I would do that when I could. And it worked well. And so for this business, like, you know, we had a small team, and knowing that we were having to show up and be live for three hours a day, I knew that we weren't going to have time to be doing a lot of ad sales. So I basically wanted to — we only sold advertising on an annual basis, which was good for us and good for the companies. So for last year, you know, we had kind of projections. Since this was 2025, our first like full year in business, we had projections and we would show those to advertisers and say like, "Hey, this is what we think we can do for the year." They would sign on, but they'd be on a fixed rate. So they would be committing to annual spend with us, but they'd be on a fixed rate. So if we grew beyond what we were projecting, it was just like, you know, free incremental impressions or reach. And so that was super important. Also because we wanted the ability to know how much revenue we were going to have in the month of March and April and May and things like that because we wanted to hire a team. We wanted to invest in our space. We have what I would say one of the most unique podcast studios in the world. It's certainly like sub a million dollars, but you know, we invested a ton of money into equipment last year, and part of that was like having this predictable revenue stream. And then the other approach that we took was, you know, I was pitching advertising with us as like sponsoring a Formula 1 team. So, you know, it wasn't like we were going to — we would run ads for every company every day, but then we would put their logo on like, you know, merch that we would make. We would put their logo on clips. If we did, you know, anything IRL, we would show up for them there. And so it was like, you know, actually like, you know, a big tech company sponsoring a Formula 1 team. They know they're going to be on the car. They know they're going to be here. They know they're going to be everywhere. And so by having like, you know, basically one contract, and then we were very open and upfront, and it was part of the show that we loved advertising. Like we truly think it's amazing. I was like, I love that somebody anywhere in the world can listen to this content for free that we put so much effort into making. And it doesn't matter who you are, it's just free. Like it's truly an incredible business model. It's very aligned. And so we were very — I would say a lot of podcasts in tech, not you, I think you're very commercial. And that's a compliment. But a lot of podcasts in tech were like above ads. They were like doing ads would be low status because I have a venture fund and I get management fees and things like that. And then the other thing is like we had a focus early on, and it's still our focus, which is we believe that there's 200,000 people in the world max that we make content for. And these people are running businesses. They're investing in the companies of the founders that we talk to. They invest tens of billions of dollars a year. They can, you know, run companies that spend billions of dollars a year on everything from advertising to software to cloud, etc. And we always believed that that was a big enough audience for what we were trying to do because in this business we go live for three hours every day. We can't afford to talk about things that we aren't interested in. We can't afford to be explaining to people like what AWS is. You know, we don't have the time nor do we want to take the time to explain SpaceX's business in the year 2026 because we assume that our audience has been following SpaceX and understanding SpaceX for a decade plus now.

Yeah. The other thing with the audience we found — yeah, we just found ways to get ad impressions to people that normally don't see ads. I think you've talked about this too, about how certain powerful people basically don't experience advertising. We figured out how to get ads to them. One of the ways, during our interview — before our interview with Mark Zuckerberg, I read him a Ramp ad directly because our ads are integrated.

Put the headphones on and John started — because I have to do an ad read before the interview starts, and with what we're doing right now, I'm sure the ads will be sliced in later. Although shout out Ziotics, they did get a freebie in here. But our ads are integrated. So there might be a Fortune 500 CEO coming on, and right as they're coming on, they hear an ad read. Throughout the interview, they're looking at our screen which has all the corporate logos including the Ramp sponsorship right there. So there's ad integrations and logo placement there. And then after they hop off, they might still hear us read the next ad. And then every time we share one of their clips — a lot of media companies and shows and podcasts have a clipping strategy, but they don't monetize the clips very well. So all of our clips have brands on them, logos, and then at the end of every one-minute clip is a full ad read. And so, you know, of course, some people scroll to the next video by then, but a lot of people do stick around and they do see that ad impression. And so the actual ad impressions go much further. And these clips, if it's some AI for legal company that's on there, it's going to get sent to the competitor company. Even if it's a very small audience, it's going to get sent around that small community. And so that clip might only get a few thousand views, but it's going to be with the right people because it's essentially trade media.

Yeah. The other thing that we did, we realized early on that if we were going to be doing live podcast-style ads, we weren't going to be able to do a 90-second ad. Like people would just immediately bail. And so we pitched advertisers on instead of giving you a smaller number of 90-second ads, we'll give you, you know, 250 20-second ads. And really selling that like repetition and frequency. So I've memorized all the ad reads. We have buttons that we can pull up and they take over the screen to show you the graphic really quickly, but it's not long enough that you want to pull out your phone and tap tap tap tap tap.

And just to give you an idea of how the advertisers get like sort of strange incremental reach, we made a bunch of merch early on that people were really excited about. And we never sold merch because it's not our core business, and it wouldn't really move the needle from a revenue standpoint. But there were websites that were making fake TBPN merch. They're still out there and putting our advertisers all over the merch in the way that we did. So they were building a business selling our advertisers, you know, basically like impressions for our advertisers. And so, yeah, and I think the TBPN story is as much around like marketing innovation and business model innovation and format innovation as it is timing, right? The timing of launching a show like this during one of the most historic technology cycles of our lifetimes certainly plays a part as well.

No. And you're instantly losing when you do that because if you're advertising enterprise products, one customer can pay for the entire ad placement, the entire deal, right? And so that was something we recognized early on with Ramp. I was like, yeah, we're going to get you or at least help you close a public company, and if that is the only customer that we get you for the entire year, job's finished. Everything else is free. Everything else is free. And so, yeah, we had a couple of consumer partners early on, but we realized that like the people in our audience can actually go and buy $5 million a year of this SaaS company, or they can spend $20 million on cloud, or they can spend $50 million on inference, but they can only buy like one Eight Sleep, right? Or maybe two. Right. And so we realized like the best use of our ad inventory should be entirely — and so our friends with consumer companies, we'll just support them. Like John's drinking Andrew Huberman's yerba mate company, and we don't even charge money for that because, yeah, we just like them.

I think they saw what we were doing online and wanted to partner with us in the way that our brands had. So we had a very interesting history with OpenAI in that we were sort of contrarian supporters. So going back to the fact that the show is very pro-advertising, when it was initially rumored that OpenAI was going to put ads in ChatGPT, the general response was, "This is bad. We don't like ads." But we're the guys who love ads, and we think ads are a really powerful business model, and we think ads can make things free. And it was like, if you want to be OpenAI, you're going to need an ad-supported version of your product. Otherwise, like Silicon Valley, we were like looking around being like, "You guys think that everyone in the world can spend $20 a month on software?" And you don't realize that like most American households will cancel a $10-a-month streaming product if they don't have a show that they love in that very moment. We were very pro-ad, and so we were celebrating milestones along that path in a very sort of — it was like half ironic, sort of funny, but also we did believe it was a real model. And we had done a number of stunts on the internet. We ran a Super Bowl ad that did shoutouts for everyone in our community. We had launched a product called Claude with ads that was sort of making fun of the Anthropic ad that was targeted at — it was punching at OpenAI, and OpenAI couldn't really punch back. And so we were laughing about that. And so we'd done a couple of these like viral marketing campaigns, and to be able to work with us, this was the deal that made the most sense to help on the marketing side, and that includes things from billboards to the next Super Bowl ad. I hope we can have a really strong voice in.

Yeah, we had been very loud about, you know, the need to — it felt like before...

AI was like real and before it was something that we used every single day, the technology industry and the labs like needed to use fear-based marketing to raise the necessary capital to build a lot of these models. And it basically by like the midpoint of last year we were seeing that like this sort of like fear-based approach to selling the product is going to just go really, really, really badly, and it has, right? Like, you know, look at public opinion around AI. People are scared of it even though they find it very useful in their day-to-day life. And so we had been unofficially giving OpenAI advice every single day through the show, being like, 'Hey, like, where's our Steve Jobs?' Like, we just going to be like, and I'm not saying that's just talk about like just talk about what the product can do for you. Because there'll be a model release and they'll talk about the stats and the benchmarks instead of just being like, 'Well, now you can go and ask for a custom piece of furniture and it will just give you exactly the image of the piece of furniture that you want. You can just go get it made for cheaper than what you would get somewhere else,' or something like that. Like there's all these different use cases that come up and they always come with like a blog post and metrics and it's not just like, 'Hey, you can go use this thing for this particular problem now.'

Yeah, and I'll give you another example. So like when there was rumor that OpenAI was going to do an adult mode, and again, like this was in Q3, Q4 of last year.

No, that was a rumor. He announced that they were planning on working on it or rolling it out under a certain umbrella that would be maybe age-gated or something, but it never launched. But there was...

And that's the kind of thing that, like, I was very... I was like, there's no way to, there's no, in my view, there was no way to defend that kind of action in the same way that I don't think there's a way you can defend Zuckerberg launching, you know, prediction markets, you know, integrating gambling into it.

You think Meta will launch a gambling product that includes prediction product that includes cash and money? Because there was some rumors in the reporting that it might be all for like social points, and I think that's a much more nuanced discussion. Like, I'm not pro-gambling, but also I've played role-playing games where I've been like, 'I want my score to be really high.' And I don't know that trying to get the highest score on Tetris is the same thing as being at a Vegas casino until 4 a.m. with real money.

Yeah, I don't know. I don't know. I just look at it as like if he really cared about shareholder value, he would be running what is the greatest business in the world, which is Meta Platforms.

Yeah. Yeah. I think Meta would... I personally think Meta is like a $3 trillion company and is going to permanently have a discount on it because of the metaverse, because of AI.

No, no, I know. But I'm just saying like as a shareholder, like at any moment could disappear.

I do always wonder about the feedback loop of like you identified a bunch of problems with the platforms. And I do wonder like sometimes there is a flywheel there where, you know, kids get addicted to social media, the next generation doesn't let their kids use the platforms, they don't get addicted, then that destroys shareholder value. And so if you're thinking long term, you're suggesting that...

That wasn't my claim. My claim was that he has a long-term economic incentive.

He might be ignoring that incentive to the terminal, but the terminal value 25 years out is pretty low, right?

Yeah. Yeah. Yeah.

One more thing, guys, I wanted to close the loop on. I think the simple explanation is like it is culturally in Meta's DNA to launch a version of the hot thing regardless of what it is. And so it doesn't matter that it's gambling and that it is objectively harmful. It's the new hot thing in consumer, so we're going to do it. And if it's photos, it's photos. If it's videos, it's videos. You acquire or you build internally.