Messy Media
01:24:51 Share

Oh, I loved We Built This City. That's a great song. Yeah, I don't think it's a, I think it's a novel song. I'm not sure it's a great song. It's in the canon. It is in the cultural canon. We can get more into Yacht Rock talk later. Yacht Rock's not in anymore. I think that's out. It's never out. It's not true, actually. I don't know how you say things like this. It's not in. It's a podcast. You just say things definitively and then move on to some other topic. I don't know what's my business.

A couple things. One, I should say, welcome to People vs. Algorism. Sometimes I forget that each week. I'm Brian Marcy. I'm joined by Troy Young. Alex Schleifer is not here. He's on vacation. Troy and I only get two weeks off a year from PVA, but Alex is European, so he gets eight weeks. But he'll be back hopefully next week. Is he trying to do like, wash drink goodbye?

We're not going to let him get away. He's got too many fans. People like Alex. Yeah, I know. So keep the Alex love coming because we need to make sure that he doesn't try to run away from the podcast. He also has an enormous ego and he loves to have it said. He loves to be pet. Speaking of that, you got pet by Dylan Byers this week on The Grill Room, a very good media podcast. He's had some really good guests on. I really enjoyed it. Yeah, I'm the shittiest guest he has.

No, you've been on a few times. I hope you're not like flirty with Dylan too. I'm whatever. It's fine. If you want to replace me with Dylan, that's fine. Dylan's a great reporter. And I enjoy the podcast when you're on it and when you're not on it. But you know, you guys, you won. He will do, he will document the death of legacy media until he falls over dead. He, I know he loves it. And he pumps out a lot of content. You guys got to that. But,

But you did fit just one note. You didn't promote this podcast. The first rule of the information space, we've gone over this a million times, is to be totally shameless. I took that note. I took that note, in fact, so much so that I incorporated it into the email this week that we're going to send out tomorrow. Okay, good.

So let's talk about, cause you, you know, I think you're like, you became like Dylan's like AI whisperer, I guess. I mean, that was, I actually took, had a bit of an issue with that because they are, they're a punny bunch over there. They like to punt it up in the headlines. Yeah.

Which to many people in the field, from what I understand, is a low form of journalism. Well, that's me. I think it's a low form of humor. Oh, really? Because journalists have this weird obsession with puns, particularly British journalists. Well, the first time I was on, I think that they titled it Troy's Story, then something. And then the second one was Troy's Story 2.

And I said, listen, Dylan, I'm happy to be on here, but you got to cool it on the Troy story thing. Like we're not going to three, even though from what I understand, toy story three was the, what was like the, the, the best in the franchise. Apparently one of the, one of the standout, I don't, it's hard to be one, but how about Troy's don't cry. I like that. Well, Troy's don't cry. I, so I said to him, don't do that anymore. It's annoying. And he said, I don't write the headlines.

Oh, yeah. I see that. There you go. So he's throwing John under the bus. That's fine. But as it turned out, they put me in the nerd corner. What did they say? What was the title? I don't know. Something about AI. Love story? AI love story or something like that? Yeah, AI love story. That's right. So I'll take that. I take that. I like AI. So, yeah. Let's talk about it because we've been using AI more with our preparation. Because I keep...

Look, Alex is not here to say the spaceship is over the White House. Okay, fine. I just keep trying to use this stuff to be more productive, you know? And let's just...

Give a little rundown of like how you are using AI with this podcast. Because we do it with this podcast. I mean, like if you use any tools like Riverside, AI is used in lots of things in the editing process. It doesn't replace the editor, you know, but it absolutely speeds it up and helps set that up. But in our preparation, we are increasingly using AI. I'll tell a couple. One is a story of frustration and failure and

And the other is, I mean, I thought was for me a little bit kind of something that changed my concept of how to work and how I would use the tools. So we'll start there. Let's call this one the Yacht Rock case study. So I'm watching on HBO, I'm watching Yacht Rock, the documentary. It's really terrific. If you want a companion, by the way, a playlist, questionnaires,

Quest Love, who's featured on the documentary prominently, has on Spotify the sort of canonicals like Yacht Rock playlist, which goes on forever. And it's tremendous. So there's something for you, Brian. And you're in Miami. You should be listening to that kind of stuff. It fits the vibe. I'm on yachts frequently. Yeah. So anyway, I'm like...

During the week, I'm kind of a disorganized person. And I'm like, I'd like to make a note of that Yacht Rock thing and put it in the file for things that I would consider for our good product TM segment at the end that you guys get me to do. And so I'm always looking for something that touched me in some way. And so it worked. I liked it.

I was like, usually I'd put this in notes on my phone. I would just type it in, you know, my list of good products. Instead, I was near my computer and I opened up a tab and I went into ChatGPT and I saw this little thing, this interface on the left-hand side called projects. I created a new project. I'm like, I'll create a good product project. You open up a project and it's like opening up a folder. And then you open up that folder and you can add stuff to it, right?

You can upload reference material. You can tell it how to speak, speak in the voice of, you know, go look at everything we've written on PVA, you know, speak in that kind of voice. And I just go in and I type, keep it up for me. I just watched the Yacht Rock documentary. Boom.

And suddenly it does what AI does and it kind of pulls all this information about the show, the important figures in the history of Yacht Rock, where the name came from. I asked it a couple of questions like Yacht Rock was widely sampled. Give me some examples of where they used it in hip hop. It gives me all that. It shows me a couple of videos. There was it turns out that there was a YouTube series called.

about Yacht Rock that was this kind of fun, spoofy thing that is actually where the name Yacht Rock came from. And then I said, take all of this and write me two paragraphs that I can read on the podcast for a good product. And maybe we'll get to that at the end. But no, maybe we don't need to. And it was pretty sensational the way it worked. Now, what was interesting about this is...

Not, you know, what I've just described, I think, is a fairly typical, you know, chat use case on, you know, on chat GPT. The difference here is had a folder, had reference materials, and then it created something called a canvas, which is like a document. And I could go in and I could edit that canvas. Right.

So I was like, wait a minute, this is really different. Suddenly I'm not just chatting with something and that conversation evaporates. I have a permanent place for a project that

That I've defined and I've added reference material and all of that, that I can come back to and I can edit it. And, you know, in the edit, there was a couple of places where I wanted it to kind of add more information. And I just said, add the information to my edit. And it kind of redraws the document and adds all that stuff into it. So I want to,

Just kind of like there's a subtlety here, right? And I can edit, I can, I go to word or Google docs or whatever is your, your, your, your creation or writing environment of choice and you do your thing and they nag you a little bit now to say, summarize this or, you know, use AI in different ways as like a co-pilot in that environment.

And I'm like, yeah, thank you. I don't really need it. I'm here to write. I know what I'm doing. I need the spell check, maybe a little bit more, but not much. But it's primarily a writing environment.

And what I realized is how powerful this kind of project notion was in ChatGPT as kind of Troy's workshop, right? And it was a combination of research, writing, editing, and it made it like so much kind of better and empowering somehow than just writing something in Word. And so I found it incredibly, incredibly valuable and it changed my mind about kind of

the augmentation of a human being in knowledge work. Like this was, this was kind of power, a powerful combination of a kind of friendly intern helper. And it, at times, by the way, it fills me with rage. Like a lot of times, like there's another use case that's, that's been a complete fail and so much so that I, so, so the yacht rock use case was, was very successful. The second use case was, it was a disaster. And to do this, by the way, I, I bought,

I paid Elon 400 bucks to get access to Grok 3 because I was desperate to get like a function that could kind of start to do this. And what you notice is there's tremendous limitations to data sets inside of AI. But what I was really trying to do is I moved our week. We have a conversation, folks, that goes on all week between Brian, Alex, and myself, and more recently, the anonymous bankers.

Who is going to be joining us at some point this morning? Yeah. And what's interesting about that conversation, first of all, I really look forward to it. And it's just a really kind of vital connection between minds and people that I really like. And there's always something good to eat there. So everybody's constantly circulating stuff that they're reading. But it's messy. It's messy. It's messy.

It's very messy. And so I asked my colleagues to move the conversation to WhatsApp from iMessage. And there was a good reason. There was resistance among some. There was a lot of resistance. And you see, you feel bad for the sort of change agents at Accenture when they have to, you know, ask someone to use a different piece of software because they complained. And it was all like, I hate WhatsApp and it's so Euro and I don't do anything there except for like talk to my maintenance people. I don't know who said any of that stuff. I don't remember. You said that, my friend.

Anyway, and then and then. So the reason I did it is because in WhatsApp, I can export our conversations. And what I wanted to do was export our conversation, dump it into either ChatGPT or Gemini or Grok and say, take the past seven days of links and the descriptions for those links. Then go out and dig a layer deeper, read the content.

inside of each of the links, come back and give me a tidy list. Link, author, writer,

you know, what they said in WhatsApp and the description of the content and anything notable in that. And that has been incredibly frustrating. It's been really, really hard to do that, but I want to, so I'm, I'm still working away at that. I think, I think the, the secret to being good at this stuff is like, you keep hammering away at it. You got to keep challenging how you do things. You have to be super agile, all that stuff. But then, um,

Yesterday, in what seems to be this sort of unending string of innovation in the AI or the sort of frontier or LLM model war. So yesterday, I read an announcement about a new service from a company called Convergence.ai, I believe out of the UK, I think, or the company may be called Proxy, P-R-O-X-Y. And turns out that

They have something that you can get in the $200 bundle. This is happening a lot. Like, what do you pay for with DeepSeek? What do you pay for for Grok? And what do you get for the premium tier? And the premium tiers are getting really expensive at a couple hundred bucks. Anyway, I have another use case, which is, I think, really, really fundamental. And I'm going to spend a lot more time trying to understand this one. And it's that I want...

AI to be the interface to my information gathering process. So when I wake up in the morning, I want to say, give me the top headlines from these sources in these categories. And I want the, and in this format and I'll, I'll dive deeper, but I want it to go out and give me like the top stories from the wall street journal, New York times financial times and list it out for me. And then I can, and then I can go on, on my way. Yeah.

So this company proxy allows you to do what you can do on ChatGPT, I think at the $200 level, which is called operator. An operator essentially operates your computer on your behalf. It's like an agent, right? Wait, OpenAI has an operator. That's right. This all stuff is so confusing. But the OpenAI has an operator. It's part of the $200 a month package, right? So proxy, and you can find proxy at convergence.ai if you want to check it out.

allows you to set up an agent for free. And I was like, well, that's interesting because I am finding that on all of these services, particularly when I'm looking at not evergreen information, but information where we intersect. And this is the primary use case that I'm looking for. I don't want to plot the trajectory between earth and Mars and like do some scientific thing. I want, you know, current information. Um,

So you see the limits of that. Grok has obviously has the sort of corpus of Twitter or X. You know, OpenAI has done licensing deals with several companies, including Reddit. So they're better in that space. Different companies have different sources of data. So I asked Proxy to log into the New York Times and start summarizing different parts for me. And what it does is it splits the screen in two. And you can see it go in.

Go to the New York Times, take your information, authenticate, and then you can sort of, you see what it's doing on the right-hand frame, and you can ask it to go and do things.

That was extremely slow, but definitely pointed in the direction of where things have to go. And what I mean is this, and I don't understand the limitations of this technically or legally, but I want to use a service. My preferred service from a product kind of interface perspective is definitely ChatGPT. I want to be able to go in and itemize out my credentials for all the places where I pay for subscriptions and

And then I want it to be able to take from that source of information and bring stuff back to me. Can't do that yet. And it's got to go there. I suspect that, you know, the media companies are resistant because once it crawls the data, what's to prevent it from from indexing that data. But you can see where this is going. I think that.

that AI is an incredibly powerful interface layer to make you more efficient. It made me more efficient in the yacht rock example as a sort of collaborative creation environment. I want it to be a collaborative information finding tool to do that. It has to go to the sources that I cherish and the sources that, that, that,

And it's going to have to pay for access. It's going to have a toll booth, whether it's toll bid or any of the other pro rata. Right. But I already pay. Right. So I want I want they're going to have to cut in. They're going to have to cut in the publishers or they're going to get blocked.

I mean, this is not. Yeah, I'm just saying that why can't we credential ourselves at the layer of our authentication with OpenAI? This isn't a technology problem, is it? I don't think so. This is a problem of a typical problem of a lot of people who want to float on top of the people actually who are making content, take money off of it and reconfigure it.

Yeah, but there's already an economy in place of how we paywall and charge for content at the level of the individual.

Correct. But they do not get to then take that paywalled content, whether you give them your credentials or not. It's against the terms of service. And if it isn't now, it will be made explicit in that. Right. But wouldn't it be technically possible? You're paying them 200 bucks, right? No, I get it. But wouldn't it be technically possible to silo off?

my data set from their overall sort of kind of master LLM training data set. You would think that there's a possibility there so that they're not paying for it. The bottom line is they're taking copyrighted content and they are reconfiguring it and they're wringing economic value out of it and they're going to have to pay. Right.

Right. What I would have you consider is it's really hard math for someone like the Financial Times to say at any price, come in and sweep through all of our content to make it widely available to all your subscribers. That will never happen because the Financial Times wants the ability to sell directly to you and me at a premium price.

So all I want is the best of both worlds. I want it to authenticate the stuff that I pay for and to be able to be an interface layer on top of that. And it can go out and get the rest of the broader sort of evergreen corpus of content elsewhere. But I want my own instance. And that's what I'm suggesting.

So those are my use cases. What about you? What happened to you this week? Have you been using it as an environment to write in? Have you tried that? Yeah, I imported like, I guess, yeah, I imported all of my newsletters to it and I've been using it as an editor and also, you know, within a project and kind of upload all the transcripts from podcasts and whatnot. And, you know, it's useful as an editor for sure. And I like it for...

Being able to come up with, it's funny because a reader actually sent me a funny email I got yesterday. He said that he's like, yeah, I've been dealing with stuff and haven't really been catching up on all your newsletters the last year or so. So I asked him to summarize everything and he shared a Google Doc with the summarization. He's like, it doesn't replace the actual. Right. I mean, maybe it's just being nice to me.

And I think it will be fine for those kind of uses. I think internally, like I like it for, you know, it can do like an outline of like a guide, like a guide to subscriptions. Like, you know, this is like a normal thing in B2B offer guides in order to get like first party data. And this would get you like 75% there. Like I could do this. Like, whereas I'm like, oh God, I would have to like spend like, you know, a day or so doing this. I'm going to do this in like two hours. It's not going to like write it. I wouldn't, I don't,

Even if it has access to... But have you used it to make content other than a guide? Have you used it to edit any of the stuff that you post to your email? Oh, yeah. I used it to edit my newsletter. I was like, well, you have all of my content in there. What examples do you suggest I use? What are points that are missing based on my previous writing and stuff? And I think it's completely useful for that. It acts as...

you would with an editor to some degree. Is it frustrating? Yeah, but like a lot of editors are frustrating, you know, so it's at least great. One thing I do wonder with this, because it's really good at summarization, is at least in my view, I don't think it really replaces human curation exactly. You know, like you bring up like John Ellis as a good example of

you know, someone who is really good. What is his newsletter called again? Media news items, news items. Like who's really good at, I think it's curation, not aggregation. Like I think aggregation is going to get crushed by it. Just look at John stuff. You know, one of the interesting things he does is he doesn't require you to link out on everything. He'll abstract out a couple of key paragraphs from the items so that you can actually get what you need in his document, which is a nice approach. Yeah. Yeah.

I mean, look, all of these things, you know, there's a fine line between, and this always, in the aggregation era, it was always this, you know, there's a fine line between what's like, you know, reader-friendly and what's like, you know. Business-friendly. Yeah. Links. Yeah. Yeah, everyone wants to get that. But anyway, like, I don't think it necessarily replaces, you know, human curation because, but the curation that has to happen is going to be

I think a little bit more idiosyncratic. It has to be surprising. You know, it has to have a lens. I think a lot of, you know, we've talked about like about Feed Me a lot. And I think a lot of what that is, is, you know, Emily has a lens on a bunch of different things that intertwine into the

It's hard to really pinpoint what Feed Me is exactly about, but she has a lens on it and a lot of her stuff is curation, right? I mean, there's some, yeah, okay, Balthazar is new. Yeah, I mean, you got to realize that if you look at the human mind as a machine, it's

Emily spent her whole life first in Long Island, then in New York City, sort of pressed up against the glass of New York society and culture, and has an extremely good sort of media sense, and with that is able to kind of make those fine determinations between what's interesting and what's not. And

You know, like AI just doesn't have those receptors. It's not got that kind of elegant, you know, very, very specific training. And so, yeah, that's not going away. You know, the human curator part is, or even the writing part, like AI ain't never going to write like Tina Brown does.

And so, you know, I think that that's safe territory for a very long time. I did an audience survey for the rebooting recently and I got like a few hundred responses and I asked about like new products. And what was really interesting to me was that the one that scored the lowest was like daily news coverage. Like do people not want that?

So curation actually scored was one of the highest of things that they would like to see. Well, also, because I think you've made the point that where we used to get curated items from is sort of polluted now. And that was Twitter. Well, X is a disaster. It's good for some things, but it's a disaster for that kind of curation. I've like, you know, dealt with it. And I think a lot of people, it has a different use case. Well, the people like you left, they were run out of town.

They're on blue sky. I'm not on blue sky. I'll go on blue sky, yes. No, actually, I won't. But, I mean, X is interesting for that because they've gone so against that and they adopted TikTok, the algorithm approach of TikTok,

And they're now developing their own stars. Do you know Anthony Pompliano? I mean, yeah, you always run in. He's like a creature of the internet. He's around. He's a creature, I think, of X, right? Like he exists. For those who don't know Anthony Pompliano, you know, he's...

He's into crypto. He's a Miami guy, ex-military. He published a book recently of those motivational sayings. He once hawked Bitcoin pizza. There's this viral clip of him promoting a crypto company that ended up going to zero. He's a big podcaster, right? He's got a big podcast. He's a podcaster. He's MAGA adjacent. And-

X is now making a business show, a daily business show with Pump. He's got like 2 million followers on X.

And I think this is like the way they're going to go. Like they have their, every platform network needs to incubate their like homegrown stars. I mean, we see obviously YouTube, there's a ton of YouTubers that exist. Substack is, is going in this direction. It's one of the reasons I kind of left. I don't know if that's a mistake or not, but I just didn't want to be a substacker and they have a roster of, um,

I mean, I would argue that, I know you had said this on the podcast with Dylan. I sort of didn't totally agree. Like you had said, Emily was like really great at like social media, but I think she's really great at Substack. And I think Substack is developing its own quote unquote stars online.

because it has its own mechanics. Like if you go over to Substack Notes, for instance, it's a totally different experience than going on X or even I presume like Blue Sky. It's very earnest. It's earnest promotion. Everyone's hawking their wares, which is fine, but it's wonkier. And what does well on Notes is very different than what does well on X. And I think

you see Substack developing

its own ecosystem. And I think that is part of the lock-in. That you can't, not only do you need the recommendations and the credit cards on file, but for a certain type of person, and they need to keep the whales in there, right? For a certain type of person, they become creatures of Substack who can't necessarily exist in other parts. Yeah, I mean, I think it's true. Maybe, I don't know if it's a recent truth, but platforms develop cultures and people that populate them and

and a kind of comfort with a certain format type and personality that works with that format type. And you see that for sure in Substack. Emily has a great post about this, by the way, which is,

you know, about what happened to the sort of written content with the advent of Substack. It's called, I think the post is called, it's really good. It's called The Machine is in the Garden. It references another book title of the same name. And it's about what happened to Substack when, you know, with this kind of new platform that encouraged people to charge people for their writing. I guess Twitter always had a culture. It's just that, you know, Elon changed it.

Right. And he disingenuously said that that culture was artificially put in place when actually it just grew organically in my experience having been on Twitter since pretty much the beginning. That's just not true, but whatever. We're in this era where like everyone just...

you know, throws something out there that is clearly not true and they know it is not true, but it's all part of some sort of negotiation of like trying to move, you know, some semblance of the truth more towards like where it's beneficial to the person who's throwing that out there. I think he knows better, but whatever.

One of the things I've noticed is some of the big media sort of refugees who have really become good at that kind of format. Like, I think Paul Krugman is far better on Substack than he was on The New York Times. And, you know, one of the reasons he left The New York Times was that they –

They killed his blog and then he went over to Substack to put his wonky and his blog was like a lot of like the wonky political stuff because after the financial crisis, there was this explosion of like econ bloggers and he wanted to be in the mix on the daily mix of like, and yeah,

He was writing, it's interesting because when he went over to the New York Times, I remember thinking, oh, this is great. We're going to get a lot of in-depth economics writing and stuff. And then he sort of went into, you get pulled into anything. He sort of went completely into politics. He was very, very vocal during the George W. Bush presidency on politics.

But the blog allowed him to really go into where his specialty was, which is economics. And they started to cut down on that. They gave him a newsletter as a sop because he opened up a Substack and they were freaked out. And this is back when the New York Times was terrified of Substack, I guess.

And now he's on Substack. He left because they basically were like, yeah, you're writing too much. And they were trying to sand down his opinions and whatnot. And I think he's more interesting. I think Nate Silver is way more interesting on Substack than FiveThirtyEight was as a brand with him as the leader of it. And I think, like you said, Tina Brown...

She's good at this. And that's how you can tell good writers to me. Because there's a lot of writers who are only good at one particular medium, right? But when you can flex across different mediums, that's when you're really good to me. And that's something I've noticed is that kind of adaptability is rare, but something to be appreciated. Did you get through the entire Incredibly Long Murdoch's story? Well, there was two of them. I read the one in The Atlantic.

I find, I think, you know, I was a... How many words was that? Was it 30K? I don't know. But like, I grew up as a really passionate magazine reader. And my wife's father, one of the things I like when I met her is he subscribed to everything. So I would go to their house and just sit and, you know, go through. I mean, they had everything. They had, you know, they had The Atlantic, they had The New Yorker, they had The Economist, they had...

They had everything. And so I love that type of journalism. And I do think that there's a type of storyteller, you know, that thrives in long form magazine journalism that, you know, I kind of miss from that format. So I love that piece. I read all of it. To me, it's perfect. It's kind of perfect for my attention span. It's like whatever you said, 30,000 words or not. But to me, it's like a short book.

And I'd rather read that than the whole novel. Yeah. It's interesting to see how Puck is trying to reinvent the magazine format. You know, like it's messier, which I think is the hallmark of a lot of media in the information space. It's just messier because, you know, authenticity is probably overvalued, but, you know, it is highly valued right now. And no better way to

To show your authenticity. It's like the old kind of, you know, the reliable notion of understanding mode. And there's a time when I want just the facts, ma'am. I want, you know, I want bullets. I want efficiency. I want to get through the news. I don't want to spend a lot of time on, you know, the idiocy of Trump's, you know, what Trump said to Zelensky or something like that. I just want the facts.

And then there's a time when, you know, I want to go in and kind of read as entertainment. And I just think that the modes are different and you're probably a different time. It's probably a different time of the day. And it reminds me of the sort of all, you know, the existential challenge between...

you know, media companies that wanted to be commerce companies. It's like the shopping mode is different than the media consumption mode. And you have to be careful how you combine them. If you're going out to kind of get something, to buy something, to evaluate products, to scratch that itch, right? Like, I just think you're thinking really differently about

what, you know, the time is set aside, you know, to do, you know, like if you look at something like a content commerce play in the food space, you know, you might be reading about food more than likely you're reading a recipe or, you know, planning a meal and that versus buying a set of knives is just a really, you know, a really different kind of allocation of mind space and time. So, you know, I, I think that that's why I really want my, my,

AI use case in morning news aggregation to work for me. I'm willing to pay individually for all those services, but there, I want, you know, there, there's a time that I want that interface to be completely flexible to go out and get what I want, bring it back to me. And I want to kind of have this investigative back and forth with that, that body of content. And then I'll go in and read later. And I think that's a different time when I'm, you know, kind of feet up sitting back and reading.

Yeah. Did you see that the New York Times is getting its staffers to use AI internally to produce their content? You know, not to write the content, you know, with all the safeguards around the use of AI. Well, this is the reason that I... It does get shit wrong. One of the reasons I spend so much time with this right now is it just like...

I just can't see how, you know, this doesn't change. This technology just doesn't change every industry. And that doesn't necessarily mean the New York Times product that we consume, although it will, has to change. But certainly the process of making it has to change.

And we talked about this on the Grow Room episode. It's like, of course you want to use AI in putting together stories. Of course the creation tools are going to change. Of course it's going to change the workforce. You know, there's a really good podcast, I think, that kind of sadly named Invest Like the Best, which is to me not a- I like that podcast. What do you mean? Well, I love the podcast. Patrick O'Shaughnessy?

Patrick is, I think, is really, really talented. Anyway, he had this guy on Ravi Gupta. The name of the episode is called AI or Die. And he talked, he said something the other day that really struck me. And it's a good episode. He's a really thoughtful guy. He's an investor. But he said that the pace of change is so unrelenting.

That the number one thing, you know, he talked about the process of integrating AI into companies by literally going through everybody on your payroll and saying, how could we do this differently to deliver against a kind of value proposition for the customer?

So look at every individual. Do we need this person? How do they contribute to the value the customer receives? Could it be done better, either replaced by or augmented by technology? And he said the number one challenge for any large organization right now with regards to AI is agility and flexibility.

And I just think, you know, that sounds like kind of McKinsey speak, but it's absolutely true. If you were going to change at the New York Times or, you know, at any media company, how you make things, the roles that people have, how people collaborate, the lines between different disciplines inside of your company, you need to have an organization that's by definition massively agile.

And so I would encourage rather than, you know, I mean, I just, I think that the inflexibility of the company union relationship is a big problem at this point of massive change. Oh, a hundred percent. But I think it goes beyond that. Like, cause I wrote a little bit about this today. I don't know if I totally nailed it, but basically about how, you know, it's not enough just to be a specialist. You

these days, like you have to be able to flex beyond your like specialization. Specialization is like now like table stakes is like a lot of cliches, but you have to be able to go beyond that. As a brand, you have to. And I think one of the good things about

newsletters and podcasts is that it rewards an idiosyncratic and messy, messier approach because it, you know, with contradictions and with revisiting topics, that was something Krugman, is it Krugman? That was something that he, you know, mentioned was like, they were like, well, you've already written about this. And like, the reality is like,

everyone's just still working it out and you're going to revisit topics like all the time. it's by necessity a layered process for sure. Just like layers and layers and layers. And so I think that there's going to be more media of that. But also I think within organizations and I think I zeroed it out on media is, you know, with all the more with less, et cetera. It's,

The not my job thing doesn't work. And I think a lot of, you know, the union, the, you know, the unions put pressure to have, you know, okay, this is the job. We're going to define it. You're not going to expand it. And the reality is in a volatile environment, particularly these tools changing so much, you know, you have to be able to flex beyond what your specialization is. I couldn't agree more. I just bring that home in a slightly different way.

I think that the big fuzzy line that's about to appear or to sort of semi disintegrate is between kind of operating functions and organizations and technical ones.

And, you know, as you know, in any organization, it's like, yeah, we'll get the engineer, we'll get the tech guys to do that, right? Or, oh my God, we want to build this, but, you know, they'll never, you know, get it done in time. Or, you know, first you have to rigorously define the requirements and then feed them through this product function into the technology organization. And I think that, like, literally in the next couple of years, you know,

you know, the controls of technology at a development level will be put in the hands of a new constituency. That means that there are going to be essential functions that, you know, certainly the technology crowd performs to manage, you know, a larger ecosystem of technology, but business people and operators are going to be able to make their own applications.

And, and my son and I talk about this a lot because he's, you know, an artist and a musician. And, and I'm like, dude, he's like, dad, I'm going to learn to program, or I'm certainly going to learn to ask the AI to program and then know where to put the files and how to set stuff up and know how to troubleshoot. And so he's got like Python instances set up and he's using it for, you know, exploration of, of sort of creative ideas in art and in music and stuff like that. But like, I, I,

I said, you know, we've agreed that the most important thing for him to do right now is to like get super technical.

But getting super technical doesn't require really getting that technical. It requires more of like a hacker mindset that I feel like there's obviously a generational advantage to this. And I think that is going to be an issue for a lot of people in the sort of middle plus parts of their careers in that they sort of grew up in a system where you were rewarded for

basically being a generalist like you know middle manager you sort of like expand and you got away from the

the metal. And now we're seeing that middle management is getting the brunt of a lot of these cuts. I'm continually amazed at this time where the headline number of unemployment is so low, yet the headlines are always about like, it's like, "Wake up." It's like, "Okay, Meta's going to whack thousands." And by the way, they're going to say that, "Hey, these are low performers because they want to talk to Wall Street, not to

their employee base because daddy's home at Meta and other companies. And

I think that's just a reality that it's really dangerous, particularly for people, you know, in that middle to later part of their careers to narrowly define their roles and be like, well, I don't do that. You know, we kick that over. It's like, you can't do that anymore. No, the other thing that struck me there is how the definition of layoff of low performers is kind of evil, right?

Right. I mean, why don't you just lay off some of the staff that you don't need or you've made a determination that you could be more efficient? Because, because, because. No, I get it. For Wall Street, like a one, right? It's just like, but suddenly you've got 3000. Oh, you were part of that last wave of low performers. Yeah.

Yeah, we don't really need anyone right now. You know what I mean? It's like, you're labeled. This guy on LinkedIn was like, went on this like, you know, long, I mean, he got laid off and was like, we are not low performers. And like, all this is like, you know, and that's,

But that is part of this call it a vibe shift or whatever with companies really cracking down to a degree that I don't know if you saw, Jamie Dimon went on a tirade about work from home. It was recorded, of course, because everyone records these things these days. And he claims he's in the office seven days a week, right? Yeah.

But I think this is going to have incredible downstream impacts because you shared that article about how Gen Z is not interested in clawing their way to middle management. And I think this is part of a big shift where you have...

A new generation that's coming of age that sees capitalism as a scam, largely, but instead of trying to tear it down, wants to hack it. That's why derivatives trading has exploded. I think that's totally true, actually. I think that back to the Pompliano thing, I think that...

crypto isn't an alternative financial system. It's a cultural thing, right? And it's get rich quick and it's betting and it's

you know, playing it's gamified. It's, it's much, it's, it's hacking the system. And, and, you know, like those videos of, of actually refreshing videos from barstool Portnoy, he was like, these are mean, what do you mean? Mean coins? You're sad that you got the rug. Like that's the game. And so it's, yeah, the game is to pile in and try to get out before it goes down, but not too early. And yeah,

It's taking the game elements that are inherent in our financial system, but we call them traders. And then basically taking away all the quote unquote fundamentals, which by the way, are not never maybe an aggregate. I guess Jeff Bezos says the stock market is a weighing machine, not a voting machine.

But, you know, what are the fundamentals on NVIDIA? Like what are the fundamentals are kind of made up. And I feel like a lot of like younger people have seen through a lot of the made up things. Everyone makes up this stuff. That's a terrific. That's a great point. So you got this sort of pop finance thing.

slash gaming thing on the crypto side. And then you've got like broad participation in markets. And someone, this young woman that was staying at our house right now said to me last night, my grandmother bought me some stock in NVIDIA. And do you think I should sell it?

And I'm like, I don't know. You know, there is no way, there's no way that she's ever going to do the research to make a sort of sound financial determination on whether to stay long on NVIDIA. It's just a game. It's just like, okay, maybe. What are the vibes saying?

Yeah. Like, I don't think, yeah. I mean, like, but I think maybe it changes. Like, these things, these people mature, they all change. Remember, like, millennials were just about experiences and not owning things. Then all of a sudden, all the millennials have pitchforks in their hands because they don't own houses. So, like, things do change as you age. But I do sense that there is a rejection of a lot of these things.

paths that have been laid out. Look, not everyone has the privilege to reject that, right? But the idea of getting a job, clawing your way into middle management, building your 401k so you can retire at some point in your late 60s and probably won't be Social Security, et cetera, is kind of... I think people are looking at it and be like, wait a second, there's nobody who gets rich doing this. Like,

Nobody. And the only way you get rich is you have capital. We don't have capital. We got to get it. And we're going to try to use our wiles in order to find the creases in the system. It's great. It's growth hacking. I don't know how it's going to end, but that's what you get if you have a winner-take-all society. You're going to get people who take gambles. And that can be literally sports gambling. That can be GameStop gambling.

It can be all kinds of things. One of the other things I noticed, I shared this earlier, I think when we talk about the casualization of everything, and particularly as

there's so much, I think finances is interesting right now because it is going through a transformation where retail is really becoming a major part of it in a totally different way. When you look at the meme coins and that is splitting over into, you know, meme stocks really. And Robinhood had their earnings call and it looked, it was pointed out by Katy Perry, not that Katy Perry, different Katy Perry who heads investor relations at Public Banking.

which is basically a Robin Hood competitor, that it looked like a post-game press conference from like a football game or something. They were sitting up behind like, you know, a backdrop of logos and whatnot. And I think we're going to see more of those kinds of ways. Well, really, what you're saying, I mean, we really have two very different kind of worlds where we have,

Elon Musk a half an hour late to a late night debut of Grok 3 with three sort of H1B geniuses and

you know showing real-time unrehearsed demos of totally messy super messy grok making a game or grok plotting the trajectory between Mars and Earth or whatever it is and then you see the Apple launch of the the new iPhone yesterday that was like literally the the epitome of high production but that's in the uncanny valley I feel like you know at like the the the Apple one because this becomes the norm right

Right. And the very polished presentation, like, I mean, great that Jensen can, can get up at CES and, and, and, and do, you know, remind people of the Steve Jobs kind of thing, but like it or not, I think at least at this moment, that kind of Elon Musk approach with the late night grok is probably more effective. I would think. Yeah. Yeah. Is anonymous banker coming? Yeah.

Anonymous Banker is going to log in in two minutes. So let's talk about Anonymous Banker. They are coming on. We can't gender Anonymous Banker. Although I don't know how anonymous Anonymous Banker is because I got an email recently

I'm someone who identified anonymous banker. Yeah, well, people can speculate. I don't know that they were right. And, you know, one would never confirm or deny. But I didn't. I didn't. An anonymous banker is of uncertain gender.

And we're going to Charlie's Angels Anonymous Bankers voice. We're going to voice cloak and we'll see how it goes. But one of the things that Anonymous Banker I think is going to talk about is just what's going on in the funding marketplace around AI companies. And I find this stuff crazy. I find the idea that who is the ex-OpenAI employee that's starting SafeAI, SaferAI? Is that Ilya Hukamov?

Yeah. But then the other one, Mira just started her own. I'm like, I'm not using first names. Like, I mean, I just find it, you know, against the backdrop of crazy, you know, competition on the LM side. Most recently from the one that we talked about a minute ago, you have, you know, companies without products with a roadmap and a genius raising money at $20 billion, $20 billion. That's a lot of money, man.

We're talking about raising money at the market caps of like a lot of big companies, you know, based around like really leveraged off of the, you know, I guess the brilliance of a single person. Is that right? Am I getting that right? Yeah. And a lot of, look, look, a lot of this is going to look awful in, in particular looking back, right? Like, I mean, there's going to be tons of cases of lighting a lot of money on fire that are going to be laughable, right? The AI pin this week, God,

Got sold. HP bought it. I guess they have some patents or something. It basically didn't exist. I don't even think Alex had it. And all we got out of that entire journey was a pretty hilarious launch video with the guy who was putting people in a trance. It was over $20 billion, by the way. Is he going to be like an EVP at HP now? He doesn't seem like it.

He's fit for that. He's not ready for middle management at HP. Someone put it as like that it was a deserving fate for him to be like put in charge of like digital printers. Hey, man, there's downsides. Let me check on our friend, the anonymous banker. Oh, there's anonymous banker. They are here. It doesn't have to be a secret.

We're trying not to gender you. We're going to keep the anonymous thing going for a little bit anyway. It's more fun. All right, we've got Anonymous Banker in here. Anonymous Banker, thank you for joining us from your undisclosed location. Yeah, thanks for having me. So let's talk about this X valuation. You had sent me a note on this, and Alex is not here, so this is a safe space. Has Elon pulled this off? Yeah, I mean, so there's all these different narratives out there, and I think what's interesting is

there is knowable information, right? There are banks that are shopping this debt. And so there are debts with detail. And so I would go off, there's people that have been reporting it from sources that say the debt's trading close to par, which means they're basically able to sell it at what they issued at, which is a good indication that Twitter's doing just fine. So maybe blue chip advertisers have pulled, but they're still generating enough programmatic advertising to make the people that hold the debt comfortable at the

business is going to be able to repay them. We've seen some reporting banker that revenues down materially, but the cost base is so much different than profit is actually up. I mean, the best thing that debt holders are one of the best ways to get an indication of the health of the business, right? So it's typically equity prices are tied to the health, but debt is actually the best way to understand the business is doing okay.

And so if they're trading at par, which is what we've heard in terms of the larger banks selling the debt to other people, to me, the business is doing just fine. So it may not be what Meta is able to do and grow their revenue 25% year over year, but it's still a healthy business. Yeah, I thought what, but what about the XAI? I mean, some of this is...

Part of the valuation is because the spillover valuation from the holdings in the AI company. X owns 20% of XAI. And is this just a backdoor way to get exposure to that? Kind of like Yahoo's valuation was always propped up by its Alibaba stake. It wasn't really about what Yahoo was worth. I like the reporting on the price of the debt because they don't have any equity upside. All those holders care about is cash flow.

uh and getting paid back on the debt so the nuance with the twitter debt is it's a floating rate interest rate so even companies that have highly rated debt right now where they have a fixed interest rate that they issued pre sort of the market destruction a couple years ago is trading off of par because the interest rate's lower than what you can get in today's market so one of the things that's helping twitter's debt trade at par is that it's floating so it keeps changing it keeps going up based on the market

But that's what I like about the reporting around the debt is that they're only looking at Twitter as a business. There was some things that said, like, you know, they're potentially moving revenue around. But as long as the debt holders are getting paid back, it's a viable business. Yeah, I mean, they're getting, what, $200 million a year from XAI for access to the XAI?

Well, that's where that investment hits the P&L and improves cash flow. I mean, who's to say that that is, is that the right number for, you know, I mean, it's not like they're going to license that data to other AI companies. I mean, it seems like that's a lot compared to the other AI deals already.

I think what's unique about Twitter is licensing that to other people, right? I know a venture firm that's paying $5 million to get access to that knowledge graph. And so I think it's a fair business deal. Obviously, Elon has very sophisticated investors in his deal with him. So it's interesting where a lot of these journalists are becoming kind of pundits in the sense that they're abstracting their thoughts on Elon and

and X when there's very simple business cases of why Twitter's doing it is that knowledge is really valuable. The data is really valuable to train AI. And I think one of the nuances with Twitter is it's some of the best information to provide the next layer for kinetic action. And so what I mean by that, and one of the reasons why other people are going to want to license that data is it's positive signaling for these models to figure out what's going to happen next. So when someone tweets about an event, like can

People joke around and say one of the first places when they went after Osama bin Laden, some guy posted on Twitter that there was noise in his backyard. And so I think going forward, you may see more and more licensing deals. Obviously, Elon's not going to let Sam Altman have access to that data, but I think you'll see accessible and paid for by other people who wanted that information. Do you think that Elon's going to send a drone after Sam Altman? Is that what you're saying? Exactly. Exactly.

Let's talk about these AI valuations. We were talking earlier about how some kind of Mensa engineer can split off from open AI and have no product and have a $20 billion valuation. Is this math or is this magic? When people come to me and ask how much money should they raise and at what valuation, there's sort of two ways to

you can go about it as an advisor. One is you look at the multiples. And so in the last go around when there were SaaS businesses and they had a run rate ARR, you would do a multiple of that. You typically like the series A. But there's another way to do the math, which is when you raise a round of capital, you're looking for 20 to 30% dilution.

And 12 to 18 to 24 months of runway, which gets you to your next valuation milestone. And so when you reference like the Ilya fundraise, right, his seed was a billion dollars of capital at a post money of 6 billion. And so it's effectively the math of the, you know, giving away 20% of his business. He's now done a subsequent fundraise, but I would say yes, in today's market, I asked a friend who's a prolific investor right now in AI and I asked him how he's getting to his valuations and he laughed and he

And he said, most of these people don't yet have products. It's the team and, and, and Ilya, but Ilya, I think still doesn't have anything. They're just building, they're, they're raising money on what they're going to build. That must be nice. Yeah, I think so. But here's what's changing. So after deep seek, there's been a shift, right? So over the last couple of weeks, I think there's less focus on funding foundational models. And now, uh,

And this is not something new, but I think where everyone's focused now is like, okay, we have the foundation layer. We have the engine. Let's fund and build the companies that are going to go after the verticalization. So there's a business that's been around for a while called Harvey. It's been around for about a year and a half, two years. Harvey is effectively- Yeah, it's legal. Yeah. And they continue to raise significant rounds of capital. And there's a lot of things in the legal profession that can be easily adapted to

to AI language models. And so I think that's where you start to see some of the news and very valuable companies created is more on the vertical side.

like the next layer of what AI is going to be able to power. And the other thing I would love to get your take on is you had mentioned the slow ventures approach. This is on the totally different end. Let's go from like, you know, billions to like, you know, low single digit millions, which is, you know, venture was not a good fit for a lot of digital media companies. I think we can say that, right? I mean, Troy, you raised the series a

a series G, I think, at say media at one point. Not a lot of home runs there. Obviously, media business just is not a tech business, right? But you still need capital to grow a business. What was your take on how Slow Ventures is looking to come up with a different way to solve that problem that people need capital to grow, right? I mean, you can do it on cash flow, but if you're going to grow a lot bigger, you need capital. It's going to take forever.

What was your take on that? Because it's kind of an interesting approach, but I don't know if it's totally aligned. Yeah. I think the question is still out there and has not been answered is what's the best way to align your interest with these creators? Only taking 10% of their business and not really having any control typically is not a great thing for investors. You saw that with the crypto companies where people made fun of them not having board seats and not having any real corporate

corporate governance because they just didn't know where the money was being spent. I think there are very rare cases of these creators being able to create additional valuable businesses. Everyone references the Mr. Beast example, right? And I think what's interesting is you'll actually probably see Mr. Beast do an IPO at some point later this year or next, where he takes his whole empire and brings it public.

But I think that's very rare. And so the slow ventures approach, even though they titled it as new, the Turner Group's been doing it successfully with some really interesting creators in gardening and auto. So I think it makes sense with the right YouTube personalities. But to me, you need more structure and focus on what comes next beyond just advertising revenue that you generate on YouTube.

So question and comment here on one side, we've seen, you know, lots of companies do sort of creator or media infrastructure plays where they provide the operating, the sales and some, you know, the, the, the, the sort of infrastructure to, to scale those businesses and set a whole bunch of people inside of a kind of roll up like structure. It seems to me that the, the,

The approach on the slow venture side, if you want to look at it a little bit differently, is don't really think of those as much as media companies, but as sort of vibrant IP companies. And if you're investing with that $60 million in, I don't know, 20 or 30 different, you know, kind of up and coming, you know,

that are sort of in and of themselves IP, then there's going to be business extensions, whether it's, you know, frozen pizza or chocolate bars or events or, you know, television shows that have, you know, kind of, you know,

you know, discrete value that you'll own a piece of. And really, you know, if you had 20 of them, you only really need one to work at that size of a fund. I think the problem is the best ones don't want to take just money. Right. So, and that's where, and not, not to give solid ventures advice, but basically instead of just saying, we're going to give $2 million and have no control. It's like, we're going to have $2 million. And guess what? We have the person that

that built the Feastables business. We have the person that built these three businesses on the team who are going to help you because we know people that are really successful on these different platforms and they don't need dumb, like a $2 million check. And so I think I agree with you, Troy, that there is an opportunity for a portfolio approach or because there's a lot of signaling data, right? You can tell when someone's going to get traction way before they do. I just think the best people are not going to take the money. There's a, there's a guy, this guy

This guy, Caleb Harrison, that does financial breakdowns. When he actually sold 10% of his business to somebody right at the beginning, right when he was going up. And there's an episode where someone was interviewing him about his company and things like that. And he was like, it was a dumb mistake. Like, I can't believe I did that. And luckily I can buy that guy out because I don't need the money. I don't want someone here. So I think...

There is an opportunity to have a portfolio of creators where you own 10 to 20% of the business, but I think you have to add real value. I think you have to know, you have to have that Walmart connection. You have to know how to build products. You have to know how to get to Netflix. Like those are the things that I think these creators value and that's what they're going to be willing to trade for some equity in their, in their company. Yeah.

Hey, Troy, you're talking with Emily at some event, right? Ask her if she would trade a million dollars for 10% of Feed Me. She would say, fuck no. She would say today, no. Yeah, we're going to talk about that in an hour. In fact, the Uber is on its way, so I have to jump in a minute. You guys can carry on. Brian, you can maybe, or an anonymous banker can select a... I already highlighted what I thought were two good products this week. One was...

The new OpenAI project feature through ChatGPT and the second was the Yacht Rock documentary, which I'd encourage you to watch. I will. I need to bone up on. I thought Yacht Rock was out. It's never out. Okay. Ciao, guys. All right, cool. Bye. Anonymous banker. So where are valuations in media companies like now? Because is there any activity whatsoever because...

Obviously, the sector has been hit incredibly hard. What is investable on the money side? Traditional and digital media companies, what you see is people cleaning up their portfolios. I would say most large media brands that you would recognize, like the logos you would recognize today, have portfolios.

for sale sign on a lot of their smaller titles. And so there are... Who are the buyers? Is it just like, you know, people who are going to do a roll up and ring out like a lot of costs? Yeah. So, and like, but I think there's like a negative bias where it's, oh, they're just running crappy content and doing programmatic advertising. But what's interesting is that's happening everywhere, right? You look at a Hearst magazine today and they have a staff of like three people and a bunch of contributors. So I would push back and say, yes, it's like,

bit more a different way to look at these operators that are buying these smaller assets for no money and actually making them profitable are not operating them as if a lot of people would want to you know over historically right with like full-time riders and things like that but I think that's where the whole market is going in terms of valuations you know you can look at

Like as if Davis or a future and their multiples in terms of EBITDA are low, they're down there five, six, seven times per year EBITDA, which pre the last market disruption, they were up eight, nine, 10 times. Future was above 14 to 15 times.

And so I don't think you're going to see that many large transactions. There's been a few rumors of like Time Magazine potentially selling, but that's again, $100 to $200 million deal. I think that's the only thing about media is that these deals are pretty inconsequential in terms of size compared to what's happening in tech.

10 SaaS companies that sell each week that are more valuable than a time deal, but you're going to see the time news written up 20 times more. Yeah. I mean, this is something that we've discussed. I mean, these businesses typically, and they've always been, they've been bigger, way bigger brands than businesses. I mean, like I think Troy made this point on the grill room. It

Is that like, what is the New Yorker as a business? I mean, it's not that big of a business. I mean, compared to like, you know, it's just like a run of the mill SaaS company size. Like it's not even a run of the mill. A lot of-

a lot of the direct advertising and things like that have dwindled.

I think what's the thing that scares me from an M&A perspective is like the replacement of it are not necessarily sellable businesses or people that want to sell them. So like everyone that's creating a sub stack, I don't know. One, I tell them don't sell their business because it makes more sense probably to stay independent. But even if, you know, I was to help someone sell their business to Puck, it's not going to, it's not going to be like 10 years ago where I'm selling something that's a couple hundred million dollars of value like Puck.

can afford a couple million dollar M&A purchase, which is not interesting from an M&A perspective. So that's what kind of is concerning, I think, to a lot of people in the space of like the replacement of the companies of the past to what's being created new is not of the same company. Yes, they're generating... Yeah, no, they'll generate cash flow. Like you can generate like a lot of cash flow with these businesses, but... And that should be probably the goal, really. Like because...

You mentioned Mr. Beast IPO and this is like go-to example. They're like one FBI rate away from having like zero value, right? I mean, does Mr. Beast, okay, fine, you've got long-term value of the library. Imagine the risk factor section of that. I mean, this is one individual and what are you going to like substitute in like a new, meet the new Mr. Beast, right?

Like, Jimmy's gone, but we got this other person. You can't do that. Yeah, I think there's a lot of... Maybe, but... Yeah, no, I agree that there's risk. I look at it a little bit differently. I think Jimmy's a unique... The problem is, like, sometimes to understand an industry, you have to take away the outliers because it's, like, not the perfect...

way to think about it. But yeah, I mean, Jimmy has built acid value as a studios. He has these different brands, but if I think if you take him away, you lose two things because beyond him just being the person on the YouTube channel, he's like maniacal when it comes to building businesses. Like I don't think he having talked to people that work for him, he doesn't sleep. Like he's,

just as engaged as an entrepreneur and operator as a Steve Jobs or a Bill Gates when they were building their companies. He's like that significant to that business. So I guess you could ask this. This is a weird question to kind of ask back to you, but it's like if Bill Gates left Microsoft, you know, three years into it, would Microsoft

be in the same place. No, same thing with Steve jobs. So I know it's a little bit different and it's not. Yeah, but it's that, but it's more right. Like I, to me it's like, yeah, obviously you get a key man from with a lot of, of, of these businesses, but at the same time as a media business, when you don't have, at least I've always seen this, it's like, there's so much upside in the market right now to a personal brand versus an institutional brand. In most, in most cases, um,

that of course you go with it like in the short term, but the sacrifice, there's always a trade-off and the trade-off is,

You're not going to have a lot of long-term equity value as opposed to an institutional brand because everything is tied up in a person, a person. You know, the brand is tied up. There's a couple of categories where you have to keep the personality like cooking, right? Where that's actually a long-term value play. I agree with you. In certain instances, people like should be like a 404 media. Like it started with a person or like,

Some of these subsects are moving into more of a brand and there's more sustainable value created there. I think one of the nuances is in this new media world, because I was at CES with a friend at the Admin booth. And basically I was like, oh, why don't you

to these three things like you're interviewing why don't you post it to these five places and he like looked at me like i was crazy like if i did that i would have to loop in three other people and do you know five things and get five sign offs and so i think to your point of like why like media is evolving to a place where it's around these creators but the creators exist and are doing well because they've they're the only people right now that are

to this current media environment where if you write a subsec newsletter, you can go interview that person that was in the news the same day and then do a live stream or do all these things that consumers want. And so I think that, I think some of these people that have good brands are going to evolve as a personality will evolve them into brands and then be able to hopefully transact. I don't know.

Yeah, that'll be interesting to see because I think you run a lot of risk of diluting the tie you have that got you there. It's better for you as an individual. I think then one of the big challenges of any of the quote unquote creators is, man, there's so much that falls on the creator themselves and you're

I think it's kind of overdone with the burnout thing, but yeah, it's a constant treadmill to be on.

And for the creator itself, way better to de-risk, be sure you have equity value. But then you recognize that, oh, wait, maybe the value is all just tied up into me, myself, versus the quote-unquote brand. And I can't necessarily finesse that. But it'll be an interesting transition. Yeah.

Let me try to frame it a little bit differently. Yes, as the creator, even you have a parasympathetic relationship with your audience, right? People want to read the rebooting because it's Brian talking to me, not some other journalist. I think the way that you create additional value is not trying to expand the relationship you have directly. So say if you have email as your channel, you don't necessarily need to change it. It's about adding additional layer things around it. Oh, yeah.

meaning like your events and things like that and so i think that's the way that smart creators will build a ton of value like all these sub stack people it's like they don't need to make more sub stack like basically it's like can you make a movie or can you depending on what genre you're in and that's where it's almost like what you know alina dunham she kind of did in reverse but it's like the value where she probably created the most value is on the tv the hbo show right like that's

millions of dollars, whereas the newsletter that she had, it was in the infancy, but it probably made no money. And so I think that's a better way to frame it is not thinking about how they're going to expand their email or whatever their initial connection is, but what can you take that parasympathetic relationship? How can you basically exploit it? Right. Yeah. Well, let's not talk about exploitation exactly. Augmentation.

Yeah, no, because I mean, look, media is moving into, or maybe it's already moved in. And you mentioned time they got their first forges there originally is, I always say it's a front business for a better business, right? And so you need to, you need to develop us at what starts as a side hustle, but becomes like actually where the main value lies, at least in equity value, because otherwise, you

I don't know. It's hard for me to see how there's a ton of long-term value in a personal newsletter, maybe. I think it's access and influence, basically, and then you have to leverage it for other things. Because at some point, I don't know about you, but I'm filled up on my Substack subscriptions. I have like five. I take 50 bucks a month now with all these different

Yeah. And they'll bundle and stuff like this, but it's like, it's, it's yeah, you have to find that other business. And it's funny because we've talked about this in this podcast and it's sort of like a basic thing, but it comes up all the time is how you always have to be building the next business while you're operating the current business. And, you know, the next business for a lot of these, you know, write about this, you know, for a lot of the, you know, newsletters or podcasts has to be different. It's not just creating,

creating more newsletters, more podcasts. It has to be a different business. Because you see it, you're out there competing with a lot of people who use media as the front end for influence for all kinds of different things, not to directly monetize it. I mean, all in doesn't care necessarily about ads.

ads on, I mean, yeah, they have like, you know, an event, but they're after something different. And there's, that's just one example. There's so many of them out there that if you're a media business that is like, okay, I'm going to sell ads, I'm going to sell subs. It's like, you're competing with people who are like, no, I'm going to do it differently. And, and there's only so much attention to go around.

One nuance is for some of these businesses where you have influence and the right audience, brands will pay for you at an exorbitant amount that's not tied to a CPM. Right. And so like the last couple

whatever, 15, 10 to 15 years ago when fashion bloggers, when it was like a big thing to have a flash fashion blog, I know people that were making $500,000 a year from the Luxottica to eyewear company just to wear glasses and a couple of shots a month. And so I, I think that there are places of value that aren't always discernible to the outside

people, and obviously you're not going to get a Lugsonica deal writing about B2B stuff, but I think some of these consumer-focused writers and creators and influencers can earn

like they can start to level up with deals that are not based on performance where it's just like, we want to be associated with this person. And you see it today, like with actresses and athletes where they're associated with non-endemic brands to whatever they're doing, like LVMH sponsoring these different people. And those deals, when they're showing up to fashion shows and things like that, they're not just doing it for free. And so I think it's not...

Again, these are not businesses. They're more like it's a one-person thing. Okay, cool. Any other topics we should discuss, Rigo? I think the interesting one, well, we could talk about this another time, but it's basically when Substack will allow advertising in the affiliate. Oh, I love that. I love that. I mean, what is Substack already allows advertising? There's tons of advertising. I ran lots of ads on Substack. They said fine. They didn't care.

They don't do it themselves. They're starting to... They'll do it. Of course they will. They have to. Yeah, yeah. I think the biggest... The interesting thing with Substack, like one of the guys, the founders, I think the CEO, who calls himself like the head writer... Hamish? Hamish McKenzie? Yeah, yeah. So Hamish got asked like who is his customer. And I think he had like...

You didn't have a perfect answer because you wanted to say the writers and also the audience is the customer. They switched on it. They used to say that the writers were the customer. And then all of a sudden, they were like, well, it's an ecosystem. And it's like, okay, I see which direction you're going.

And look, when you take all that venture capital, it's not to be like a better ESP for a bunch of like independent writers. Like it is to build a network is to build like a YouTube. There's trade-offs to it. We'll see if they keep the whales, you know, on Substack. And they have a lot of lock-in. Yeah, I think the reason why they can keep the whales is because everyone has churn.

And so it becomes, it continues to become a discovery platform. So there are very few people that can exist off of it, but as much as people want to pretend to,

like they can grow and then leave. I think it just becomes really addictive for those new subs and they have staying power. I think the one thing that they have to change is they have a perception that they only need to do subscriptions. Yes, they allow writers to run advertising, but Substack only becomes a real business when they actually start taking a cut of the advertising because Substack has all the audience data, not

So, sure, an advertiser can show up to XYZ newsletter with 50,000 subs and hope they're getting the right demo. But Substack can easily make 20, 30% of that ad buy by making sure it gets in front of the right people. So, I think it's just a matter of time before they turn that on and actually grow the revenue. And call it something other than advertising, just like, you know, which inevitably they will. I mean, only journalists actually care about the hypocrisy of companies, but whatever. Yeah.

Yeah. And the other thing is they'll turn on performance affiliate links because so many of these newsletters link off. And you can see a future where sub-stackers are actually getting paid by the New York Times to drive subs. Like, it's an interesting ecosystem where it sort of all feeds on itself. Wait, how would that work? So if...

pretend I had a subsec and I was writing about something and I was saying, this is a great article in the New York Times. And someone clicked on that link and then signed up for the New York Times. I should get some of the value because I just exposed a reader to that website. And so I just think as the internet, as things become more traceable on the internet and you can like tokenize it. And the big thing is subsec needs to actually lean in here, but you can see more instances of value coming back to the people that are referring to

folks to sign up and buy. They really haven't evolved the commercial product like much at all. Like I've been very sorry. I mean, they've added a ton of tools. Like I left Substack like a year ago for Ghost and they've added a ton of consumer tools. I mean, I still go back to it. I'm on like, I'm still on Substack, but I don't use it. But I see like, you know, and they haven't like really done much

much to evolve the ways people can make money beyond just the regular subscribe to the newsletter. Yeah, I think they're in growth mode right now. The surprising thing I've seen on my feed is it went from like really interesting information to now like all like personal improvement. I feel like there's like 20 different Tim Ferriss, it's not Tim Ferriss, but it's like articles like that on the feed. I think you'll see they'll have to move to the monetization because Beehive

It's going to give them like a run for their money. Like Behav's, you know, leaning into programmatic advertising. They're doing certain things. Yes. Behav's my one angel investment. Anonymous banker. I can become a banker. It's my only angel investment. And it's doing tremendous. And I said to Tyler, he was like, well, we're going to do analytics much better. I'm like, whatever. Just get to the ad network. Analytics are fine. Ad network's better. Yeah. But giving up 10% of your...

I mean, think about most businesses. That's the easiest way to say why would someone leave Substack is most companies that are consumer facing spend 20% of their revenue on marketing, right? And so if you're only giving up 10%, that's the split, right? Yeah. Yeah. It's a fair trade.

Yeah, no, the discovery engine is the best. The recommendations is, and the thing that I think that is sort of underrated with Substack is you simply convert way more subscribers because they have the credit card on file. Like people don't like entering their information. I know there's lots of ways to make it easier, but the reality is you're just going to convert way more people on Substack because of that reason. And then that's like a tremendous benefit for Substackers. Yeah.

I don't think I'm going back, though. I don't want to build a... You should. I should? I never... I can't get your... Oh, God. This is a non-podcast, but your stuff still goes to my studio every fucking day. And I literally call tech support, like, so much. But I don't know. It was just fun. All right. An honest banker, thank you for your appearance there on the PVA. And that is... That's all we have this week. We need to do our good products. Oh, we didn't do good products. Good products.

I don't have a good product. Troy left this on me. I mean, Troy's good product was apparently this Yacht Rock documentary. I have to say it because I thought Yacht Rock was out. Do you have a good product, Anonymous Banker? It's a book. It's the new SoftBank book. It's actually really interesting.

What's the guy's name? What's the Masha? Masa? Masa, yeah. So yeah, so my recommendation, there's a new book out on Masa that's really good. I think it came out last month. I would read that. I mean, he's an interesting character, I guess. He grew up in this very... Well, he grew up like what? He grew up like Korean, but in Japan, and that's a strange way to... I'm sure that affected...

Yeah. And like his dad was one of these owned a bunch of gambling halls. He came to the US. Yeah. He came to the US really early. And I think that we get really conditioned in the US to have like a very, you know, our, our entrepreneur, most of our entrepreneurs and, and billionaires today follow this pretty normal path of like dropping out of Harvard to do that. But I would say that his path is like much more interesting going back and forth between the US and Japan.

and really faking a lot of stuff before he made it. And some people probably say he's still faking it today, but it's just like a much more interesting story than a lot of the entrepreneurs here in the US. Yeah, I guess that's what attracted him to Adam Neumann. He was like, oh yeah, I know this. Fake it till you make it. Totally, yeah. The problem with fake it till you make it, if you play it out too far, you get into like, you get fraud adjacent. That's the problem with fake it till you make it. Yeah.

But as much as we can make fun of him for that, it's like the ARM investment, the Yahoo investment, the NVIDIA investment. He's created, in terms of the value, he's created. So yes, he's kind of had some wrong currents there. But overall, I think his track record is pretty good. Yeah, you got to get shots on goal. And then nobody's going to...

Whatever, all the cliches, you know, you miss 100% of the shots you don't take, etc. And whatever, you know, he's also been investing probably at an ideal time for his style of investing, you know? Because there have been so many massive hits that you can have a bunch of flame outs. I think sometimes the problem with that is, you know,

You miss, there is still a judgment factor, right? There's people that invested in 10 companies that were in the internet and they have no money to show for it. And so I think there is still some judgment when you're in these different industries, even when everything's going. Hey, as an investor with an impeccable record as of now, considering where Beehive has, Tyler has really done an amazing job with that company. I mean, I understand it's hard.

Have you sold any in secondaries? No, I'm going all the way to IPO. Tyler's an animal. He's going to do great with that. That's the funny thing with Jason Calacanis is he definitely sold his Uber, I think, in secondaries before.

That sounds right. That sounds right. All right, cool. AB, thanks. That's it for this episode of People vs. Algorithms, where each week we uncover patterns shaping media, culture, and technology. Big thanks, as always, to our producer, Vanya Arsinov. She always makes us a little clearer and more understandable, and we appreciate her very, very much.

If you're enjoying these conversations, we'd love for you to leave us a review. It helps us get the word out and keeps our community growing. Remember, you can find People vs. Algorithms on Apple Podcasts, on Spotify, and now on YouTube. Thanks for listening, and we'll see you again next week.