A Conversation With Ari Emanuel About Entertainment and AI

Listen to this post:

Good morning,

This week’s Stratechery Interview ended up not being an Interview, but rather a conversation with Hollywood super-agent Ari Emanuel. Emanuel is the Executive Chairman of WME Group, Executive Chair and CEO of TKO Group Holdings (the owner of UFC and WWE), and Founder of MARI, a live events and experiences company.

The occasion for this conversation née interview was the recent publication of Emanuel’s memoir, Roll the Calls, a reference to Emanuel’s relentless daily phone call routine; I got on Emanuel’s roll for the first time a couple of years ago, and quickly learned that when Emanuel wants something he gets it.

That’s a theme of his book, and of this interview: Emanuel decided he wanted to have a conversation and ask me some questions too, so that’s exactly what we did. We cover everything from the state of social networks, AI’s impact on Hollywood and the tech industry, and what skills might not just persist but become more valuable. We also discuss why Emanuel is making such a big bet on live, whether or not Hollywood is logical, and whether talking on the phone is the antidote to polarization.

As a reminder, all Stratechery content, including interviews, is available as a podcast; click the link at the top of this email to add Stratechery to your podcast player.

On to the conversation:

A Conversation With Ari Emanuel About Entertainment and AI

This interview is lightly edited for clarity and content.

Topics:

The State of Social | Roll the Calls | AI and Creativity | Buying UFC | Post-AI Tech | AI and Generational Change | MARI | Hollywood Logic

The State of Social

Ari Emanuel: Can I ask you one question? Because I saw your interview. Who was it — you did it with a guy in New York.

Patrick O’Shaughnessy?

AE: Patrick, yeah, who I love.

Yeah, I read your interview with him, it was great.

AE: I loved your point of view, and explain a little bit more to me because I love your thinking — that on a SaaS model, the only one, based on my understanding, based on your model right now, that’s got it right, OpenAI should have gone to advertising as opposed to charging, Muse has it right.

Yeah, for sure.

AE: They don’t need to charge you because they have the other revenue stream.

Right. And well, not just that, but the LLM, like, it’s actually really interesting. So when they launched Muse, they made this promise, “We won’t use your data that you use with Muse for advertising”, and no one cares. No one knows they made that promise, and no one cares, because no one cares about that stuff. And so they, in some respects, handicapped themselves by making this promise.

AE: Just like Apple, we’re not going to look at your data, major mistake when Tim [Cook] said that.

Right. But in the long run, I think it’s going to pay off, because you’re going to have this complete change in how advertising works. Right now you sort of have data on the user and you have data, broadly speaking, of what a particular product matches with and you’re matching, it’s called the two-towers approach. So you’re matching these two things together, and that’s how advertising basically works at scale, and it’s all these vector databases, all these sorts of things.

AE: Television never did it properly.

That’s right. So the way advertising is going to work in the future, and it’s already happening, is it’s going to predict what you want.

AE: Well, the funny thing, what I thought when you said that—

(laughing) I can never remember if I’ve said it before.

AE: One is, Mark is going to be able to make a ton personally and charge, he’s going to be able to use the AI to do it, and it’s going to be just more and more and more.

It’s more than that. So what you’re going to do is you’re going to get a personalized ad predictor that basically is inside your VM. They’re going to be credibly able to say it’s totally private, because it is totally private.

AE: Because it’s just directed at you.

That’s right, it’s just predicting what you want to see. Just like it’s going to be a predictor of content you want to see.

AE: Oh, I didn’t think about it that way, yeah.

The reason why there’s such massive upside is this whole agent thing — this is the real Google threat. So everyone thought when ChatGPT came out, that was a threat to Google, but the human was still in the loop. As long as the human is still in the loop, Google is okay. But all of Google Search, Amazon search—

AE: Once the human is out of the way…

Then all these search-skim ad models, which is what all these models are, they’re all bullshit models, because it’s just attacks on search.

AE: Well, are you saying the same thing about Amazon?

Yeah, 100%.

AE: I think Amazon, because of Muse — I mean, Walmart made the deal with them.

Which is a huge deal.

AE: Amazon’s going to have to figure out — I mean, a lot of stuff is going to go through.

Well, and the problem is that the problem with having a physical moat is that means you have a lot of capital tied up in that, which means you do have a forcing function that means you have to adjust.

AE: It’s what happens when you’re just saying that when you have a digital, not a personal, on the Amazon side, your $60 billion is in jeopardy on your advertising business. I mean, it’s a problem.

It’s a big problem, a huge problem for them. So the upside is the new products still need advertising, and things like live events, all the sorts of personalized stuff that the Internet makes possible. Like, one of the most important functions is going to be the generation of human desire, that is the post-AI economy.

AE: MARI! That’s exactly how I think about it. It’s like, you know, not only what Elon [Musk] said to me — it can’t be disintermediated.

Don’t waste it. We’re going to get to it. I get to talk now, you don’t get to talk, because I have to record this for the interview.

AE: I can’t believe we’re not recording now!

I am recording, but I didn’t say we hadn’t started yet, so maybe we can include this. We should make this more of a conversation.

AE: We can just go. This is the interview.

Okay, this is the interview, we started.

AE: Exactly how my life goes.

Okay, I’ll throw away my script here.

AE: The reason I love talking to you is this, I can’t articulate it as well as you can.

You have this massive need for the generation of human desire, and you’re going to have people like you who are creating the vessels in which to sort of collect that human desire. But that means that advertising has just been between 1% and 2% of GDP, so let’s say that holds.

AE: Oh, it’s going up more.

It has gone up with digital, yeah, it’s up to between 2% and 3% now, but historically, it’s been between 1% and 2%.

AE: I think it’s going to above that.

I think it will too. So this is the conservative estimate. Amazon advertising is impaired, and Google advertising is impaired, because there’s no more human in the loop, that advertising is going to flow to surfaces that are not impacted by AI.

AE: Facebook.

Facebook.

AE: I couldn’t think, after that conversation you had with Patrick, if there’s another group out there, and maybe it’s X? Because I think Elon’s going to do that, right? But I didn’t think of another one, another kind of digital player that could do that, except maybe Microsoft inside LinkedIn.

Which, by the way, is doing very well.

AE: Very well. Because actually that is a digital personal because people are involved.

I think social networks are only going to get more valuable over time, not less.

AE: I agree.

If anything, we’re almost at the nadir of social network value.

AE: So name the social networks in the United States. Facebook. Instagram. X. And I said LinkedIn. Did I miss one?

No, I don’t think so. So that’s the brilliant thing about Muse, it’s not only this phenomenal opportunity, it’s also a direct heat missile at Meta’s competitors, to the extent it lands, and that money’s going to flow back into Meta.

AE: And I think it actually — depending on what [Amazon CEO] Andy [Jassy] does, who’s very smart, he’s got to make a deal. Over time, he’s got to figure out that $60 billion, which is one of the biggest problems.

Here’s one thing that’s interesting, and I’m curious what you’re thinking about it. So in the long run, of course, AI is very highly deflationary but I think there’s an aspect that it is short-to-medium-term inflationary, in that ads are a subsidy to the people who are not impacted by the ads. Stuff is cheaper, Amazon is a good example — the prices Amazon charges are definitionally less than they would be absent advertising. And we know this because their advertising revenue is more than their e-commerce revenue — or their profit, I should say.

AE: And the advertising does subsidize all that.

That’s right. So if the advertising goes away and they have to charge more, that is actually inflationary to a certain extent, because prices have to go up. And I think you’re going to see this — like, for example, what’s a very easy Muse use case? Find these subscriptions that are lingering, what subscriptions do I have that I should cancel? There’s a lot of advertising businesses that are supported by people forgetting to cancel their subscription.

AE: This is the same argument I’ve been making with people. The cable business — ESPN was elevated not because so many people were watching it, there was that select group of people that had to have it, and then they could get up to $10 for it.

And everyone paid it.

AE: And once that broke with Netflix, etc., it just went away. The cable operators and cable were broken, it just fell to the ground.

Yep, and you have to pay directly for sports, so sports fans were the biggest welfare recipients in history throughout the entire cable era.

AE: And now they’re going to pay for it, but advertisers are going to pay premiums for it, because it’s a very high-end group of people that they want, on whatever platform they have.

All the video advertising, streaming advertising — particularly to your point, if they actually finally do it right. And which, by the way, Amazon is probably the furthest ahead there, generally speaking.

AE: I think [Skydance CEO] David Ellison will catch up. He hired the right CTO out of Facebook, and he’s going to build a pretty good model. The one place — would you define YouTube as a social network? It’s in between, right?

What is a social network? So is Instagram even a social network anymore? It is, to an extent. But the big shift, and the reason why TikTok sort of blindsided Meta, is TikTok is classified as a social network, but it’s not a social network at all, it’s a user-generated content network.

AE: Well, so is Reels.

That’s right, and Reels is a direct response to realizing, “Holy cow, we missed it”.

AE: And the issue is that, unlike Threads, Reels works.

There’s this famous Paul Krugman article that’s saying the Internet is not going to have any more economic impact than a fax machine.

AE: He said that?

Yeah. Maybe it was a long time ago, I wrote about 2014, he must have said it in the late ’90s, early 2000s or something. And the reason is because most people don’t have anything interesting to say and the funny thing is he was actually half right and you saw this with what happened to Facebook. As long as your content was limited to people you knew, it was fun.

AE: Everybody cared!

They came to not care, because what happened was TikTok realized, if you’re limiting what people see to only the people you know, they’re not seeing the best stuff. What the TikTok algorithm does is it basically finds the best content across the entire service, and feeds it to you, such that it’s the best content in the world — even though the percentage of good content to bad content is hilariously small. The vast majority of content sucks, but the absolute value of good content is very large, because the total amount of content is so great, even if you get a very small number on a massive denominator. And so they’re like, “We’ll just show you the best content”.

AE: And that’s what their algorithm did, they could figure that out.

That’s right, and that was Meta’s blind spot, because they killed Snapchat by leveraging their network. You already had friends — you could see a story on Instagram, and it was instantly useful to you, because you were already following people on Instagram.

AE: But why — I want to ask you a question, and this brings up Snapchat. Why hasn’t he [Snap CEO Evan Spiegel], as opposed to doing glasses, figured out he is one of the last social networks? And it’s still good. He’s got a billion users.

(laughing) There’s a lot of good questions that can be asked about Evan Spiegel‘s corporate governance, most of which are not very flattering.

AE: I mean, he is one of the last social networks, we’re talking about social networks — he’s one of the last ones, and he actually still has a great user base.

He does, and it’s still young.

AE: Very young, and they use it to talk to people.

The engagement is off the charts. If you use it, you use it all the time.

AE: And if he did it right with AI, as opposed to the glasses

I mean, this is the eternal Snapchat question. That’s why they are still such a lure to investors, despite the fact they pay out gazillions in stock-based compensation, they never seem to really make money, blindsided by Apple a few years ago, and all these sorts of things, is because it’s so clearly a tremendously valuable asset.

AE: And completely utilized by a generation.

Yeah, absolutely. So, yes, Snapchat is always a question.

AE: So now let’s get to the advertising piece on live. Because my opinion is the following, let’s just say you were right, 1% or 2%.

By the way, this is a perfect framing for this interview, where you just sort of took it over. I think we’re actually conveying, at a meta level, the experience — the Ari Emanuel experience. So anyway, continue.

AE: (laughing) Well, it’s kind of the experience, I’ve been reading your newsletter, and I just call you out of the blue, and we start talking.

Roll the Calls

Well, your book is Roll the Calls.

AE: You got one of them! I rolled you!

You did. It kind of made me scared for this interview when I was actually reading the book, I’m like, “do you actually stop for an hour and not talk on the phone?”.

AE: Yeah, because it’s an interesting conversation, I’m going to learn something.

I did remember the first time you called me.

AE: You were in Asia, weren’t you?

No, I was in Seattle the first time you called me, but I was still living in Asia. And the second time I was in Asia, I missed your call in the middle of the night. But what was striking is your assistant emailed me, I found the email asking for my phone number which says, “Ari wants to call you”. I replied, and I think my phone rang within 30 seconds, which was stunning to me.

AE: (laughing) Were you sleeping, and you had woken up and got there?

No, it was Microsoft Build, it was in May, so I was there. I think I was going to interview [Microsoft CEO] Satya [Nadella]. I was in the lounge at the Hyatt downtown getting breakfast.

AE: Oh, wow. I don’t remember what the conversation was about.

It was some Article I wrote, I don’t remember what it was.

AE: I didn’t understand it completely, and I wanted some answers.

Do you still talk on the phone as much as you did back in the ’80s?

AE: Oh my god, yeah. I don’t like this whole thing. Ron Meyer once said to me, we were talking about this — and he still reads papers, and I read papers online — and he said, “You know, if texting existed before telephones did, and then all of a sudden telephones came out, people would be calling people as opposed to texting”, because the reverse happened, texting’s cool, and people just do that. So yeah, I’d rather have a conversation.

The bandwidth is just higher.

AE: Especially for me — one, being dyslexic. Two, the misinterpretation, potentially. Larry David can write funny texting, I can’t do that.

Well, I think what’s interesting is, I mean, a lot of the book is — and you talk about this — just sort of your not giving up, I think, is a generous way to put it. Asking a million times, you mentioned the dyslexic angle, not being embarrassed, people can keep saying no — it doesn’t bother you. But I think there’s an aspect of that relentlessness combined with, in a phone call, because it’s so much higher bandwidth, you get the shape of the problem or the shape of the person much more clearly and so that pressure you’re putting on is also much more finely targeted.

AE: When I get on a phone call, like I just hung up with Marc Andreessen about something, and I didn’t understand something he was asking me, I was asking him a question. I just want the answer to that one thing. I’m not sitting here — it could be 14 calls, because I might have multiple calls, but in each situation I don’t have a long conversation. I just don’t understand something, I want an answer, I’m going to the people that I think have an answer.

That answer, for whatever my life is going on, answers some question that I want answered, to kind of continue a conversation or understand something so I can think about what we’re doing, how we’re thinking about this. This is one of the reasons I wanted to do this — because one, I love your newsletter, I read it, as you know, consistently, but I was wondering, AI, live — your point of view on it.

I have a point of view — one, that people are going to have more free time, they’re going to come to it. But also, I think there’s going to be the ability to put it on social and I think advertising in arena is going to go pretty high, because you’re concentrated and I think that’s true for both TKO and I think that time that people have now is going to be so much more precious, they’re going to have more money — we can talk about that — and it’s just like where the world’s going.

So is it “Mari”? Is that how you say it?

AE: Do you know how that name came about?

I do not.

AE: I have a Monday morning weekly call with all the executives at TKO — WWE, UFC, everybody — and we were going over, I don’t know, an advertising issue or a site fee issue and the executive goes, “Mari!”, I go, because Mark Shapiro was talking — I go, “What did you just say?”.

Oh, it was Mark and Ari.

AE: I lost my shit on him, it was hysterical, then I had shirts made with “Mari” — Mark and my face combined, it was really funny, so that’s how it came.

You mentioned, as far as the thesis, people are going to have more free time, live events are going to become more valuable.

AE: And what were you just saying about your point of view about people’s time and events? You were just talking about.

I agree. And I think there’s a second layer, which is I can relate to — you just have a question, you want to get answered. You came up in like the age of the world talking on the phone so of course you’re going to call someone to get the answer.

AE: Or somebody that I think has thought about it maybe differently than me, that I could — would make me change an opinion or add to my opinion.

Right. Well, but in the ’80s, you’re an agent, you needed to know everything that was going on and that was the original purpose of rolling the calls. So I obviously am younger than you.

AE: It’s not obvious that you’re younger.

(laughing) I know, you do look great. These are your barbells and your famous walking treadmill. So Google came out when I was in college, it was unbelievable, it was so much better than everything else. But even before then, you go back to me being a kid, and like having the encyclopedia at home, or having the dictionary. Every time you encounter a new word, go look it up, every time you encounter a concept, look it up. This aspect of just always, if anything comes up that I don’t know, “Get the answer!”. AI obviously takes this to the extreme — do you ever ask AI for any of your questions?

AE: All the time. I was having this conversation — maybe this is way too political. I just didn’t understand why people, both on the far left and the far right, were using the phrase “genocide” as it relates to Israel and I said, “How the F did that work?”, there was that book about how words were created, I forgot the name of the book, long time ago I read it.

So I used Grok, and it said, “This guy, 1944, a lawyer, family went through the Holocaust, etc”, and I’m like, they’re trying to take a word that was created because of a genocide, by a Jew, and make it their own — the whole thing blew me away, we all constantly use it.

But that’s a great topic where it’s like, are you going to have like an unloaded conversation about that?

AE: Every time I ask somebody, “Do you know where that came from?”, I just did it with Bill Maher — nobody knows. This guy in 1944 was adopted by the UN in ’48, and they defined the term in two different — a Latin word and a Greek word, and it was crazy.

There’s this aspect though, when you can ask AI anything — tech is really fundamentally atomizing as it is, you all have your individual feed, I could be sitting here on TikTok, you’d be on TikTok, we’re watching totally different things, even though we’re sitting next to each other. Although our TikToks would probably be relatively somewhat similar.

AE: You’re looking at golf videos?

I go through phases.

AE: (laughing) I go golf and surfing videos.

Yeah, we’re getting into skiing season, all very aspirational. Not reflected in real life at all, at least for me. But one area where we still would talk to people all the time is asking for help or asking questions, “How do I do this?”, “I’m trying to figure this out”. And one of the amazing unlocks for AI is you can do anything. My assistant was just here and, like, he’s creating all these AI apps, and he didn’t know what GitHub was. He had to do all these sorts of things, but it’s fine — he could just take screenshots, and the AI would walk him through it. So on one hand, that’s phenomenal, that’s amazing, anyone can do anything if they have volition. On the other hand, in one fell swoop, we eliminated an entire vector of human contact.

AE: No question. I now no longer — if you really want to get to it — except for how to train, you can put videos of your golf swing into the AI and say—

I have a friend that does this!

AE: “How does Mac O’Grady think about P2 and P3 and P5?”, and it shows you exactly.

Yep, and he’s like, “My golf game has progressed more in the last three months because of that”.

AE: A hundred percent.

But the point of that is, I think this fundamental desire for human connection is going—

AE: Going to skyrocket.

That’s right. And so that’s the benefit of actual social media, not just UGC social media, but actual people that you know. It’s to the benefit of live and live events.

AI and Creativity

AE: But don’t you think AI is going to help user-generated? It’s going to unlock more creativity? They’re going to be able to do more because of it, and create more interesting user — with a little bit of AI in it? And I don’t think that is going to be detrimental to interaction.

Oh, I’m not saying UGC is going away, this is an “and”.

AE: I agree with you.

One thing that has come up in reading your book and thinking of your history, very sort of clear, you came up big in TV when it wasn’t super popular.

AE: You know why?

Because the scripts were short.

AE: 39 pages.

See, I actually read the book.

AE: It was so fucking great when I saw that I stopped sweating when people used to give me 120, 140-page feature scripts. I used to, when I got them at CAA, I used to go to Warner Brothers, and I would have the person that was doing coverage get me the coverage.

So if there was a modern screen-reading software where the AI could just read it to you, would you have been a movie guy?

AE: Oh my god, I do it now. I read a bunch of scripts, but one of the great softwares, I use Speechify — it’s incredible how it reads the script to me, and I can get them done now, understand them all in an hour. It’s fantastic.

But then you see streaming coming, that’s sort of a big shift and I think sort of being attuned to technological changes and their inevitability. I actually have a question about that, I want to ask about the streaming shift specifically. But where do you see AI then in terms of entertainment?

AE: I think it’s going to be an incredible unlock for creativity. There’s going to be some disintermediation for people inside the system that now exist because you can imagine, very soon, you’re just shooting — either it’s a green screen or light stage — you get the shot, you put it into the AI, say, “I now want this”, or you do the makeup once, and then the guy or the woman never has to put on makeup. There’s going to be some difficult conversations.

But for the likes of great filmmakers like Baz Luhrmann or Alejandro, I think it’s only going to be one of the great creative unlocks of all time, and David Ellison talks about it, a lot of people talk about it. And I think the cost of production is going to come down in a good way.

That means there’s going to be more content and I think there’s going to be more time for the consumer, and therefore a lot more content being made because David Ellison’s going to pay $30, I think Netflix is starting to pay more, Disney, Amazon, that’s only a good thing. There’s going to be more content being made, more creative, I think you’re going to find also a lot more creative people that we’re not thinking about now that are going to use AI to create an incredible storyline.

But isn’t this ultimately a YouTube bull case? Why is that good for Hollywood?

AE: No, I still think you need a Ben Affleck, a Marty Scorsese, a Baz Luhrmann — you need real, you know, Jim Brooks, Larry David — you need great creative to make great product. I don’t think it is like, “Oh, anybody can make a great movie, a great TV show”, I don’t think the AI is going to step into that breach, I do not believe that.

Is it going to be, though, much more — you’re going to find the next great creators because they did get started on their own?

AE: No, no, no. I think, first of all, if you are a creator — whether it be a writer, director, actor, whatever — this town finds you.

Has that changed? How has that changed in the last 40 years?

AE: It hasn’t. An agent’s going to find you, whether that be on YouTube or wherever, they’re going to find you and they’re going to help you and the great thing about being an agent is there’s constant conflict — budget, actor, distribution, storylines.

Our job is to, one, get the best economics for you, find you the money. It’s going to be easier, because I think it’s going to be less money and then make sure the distribution, the marketing, and all that works properly. There’s going to be other ways to distribute it, that maybe you can create models that, depending on the economics, you can use the system on social, YouTube, etc., but the job of an agent to find those people — one, that are creatively incredible, and to match them and resolve the conflicts that occur — that is a skill set that I don’t think is going away anytime soon.

Is Hollywood agents? Because I just asked about Hollywood, and you just gave me the bull case for agents, are they one and the same thing?

AE: Well, you need us to make sure that whole assembly line of when products get made, to happen and sometimes it happens at the high end, which is a Warner, a Netflix, an Amazon, a Disney, and sometimes it happens at a smaller scale, and then you have to navigate how to get bigger people with better economics to buy it and that’s the job of an agent.

Buying UFC

One way to think about your career is you’re an agent, you’re hustling, start your own firm, firm needs to get bigger, you have two transformative deals — William Morris and then IMG. But then there’s the, “What? They just bought what? UFC?”

AE: Well, what we did was, you know, this was Egon Durban and myself and when we were just William Morris, I was running around all over the world trying to raise money, as Teddy [Theodore Forstmann] got sick. In IMG, back in the day, I went to see Eike Batista in Brazil, who was at the time at $30 billion, one of the fifth richest men in the world, I saw Martin Sorrell when he was running WPP, Khaldoon [Al Mubarak] in Abu Dhabi, just to, “Okay, I think I’m going to need this amount of money”, and then I had been in business, you know, been talking through Raine, Marc Andreessen and Egon Durban.

And Egon and I — when he had Skype, and then he had sold it, he wanted to put content on Skype, he had thought about that way before anybody was thinking about that. I think that was 2009 or ’10, so we had been talking about how to do that, he wanted Raine to invest. And he saw the world early, before anybody in PE had seen it, where it was going with content and distribution and that and when we had built the company, we were global, we had live events, we had sports production, sports distribution — we had negotiated all these deals on sports and movie and television rights.

And we said, “We should now own something”, so we just had this combination where, when we represent something, we realized, “Oh, they’re losing value, and we could change it, then we could buy it”, we had the licensing business, all these things, we had sponsorship, if you put something on that you owned, you could really reap the benefits, and you do all the things anyway — you’re just collecting 10% and that was a conversation we had. The first one actually was PBR, and we made that work.

And then Lorenzo and Frank Fertitta decided to sell and our nervousness was it was on Fox, the deal I had made — they could have stepped in easily, and then Jeff Bewkes at Warner wanted it, or his team wanted it. Jeff was trying to downsize, to sell it, and so he didn’t want to do it and that just left us against the Chinese, who was the other bidder. And Dana [White] didn’t want to go with them — he wanted to come, I’d known him for 20-plus years, and it was an easy transition.

You just articulated it, because you moved from this somewhat unstable, like, why did you have to get big, people could leave you with clients, or agents could take their clients with them, into this much more stable owning IP.

AE: Owning an asset.

Owning an asset. But is there an aspect where, with AI, as it becomes more and more pervasive, as it creates more and more options in terms of content, that actually being an agent is going to be a better place to be precisely because of that human component that made it feel so unstable?

AE: Yeah, I mean, I think there’s going to be an incredible place for being an agent. Incredible. Across everything we just talked about, sports.

Post-AI Tech

You’ve articulated — which I completely agree with — the extent to which AI supercharges particularly exceptional individuals. Presumably, those super-exceptional individuals would benefit from having an agent, I think that would be your take.

AE: If they’re as creative as those super-exceptional people on the business side of their life in the entertainment business — creative as it relates to getting financing, structuring the deals, resolving conflict, marketing, bringing them new IP that they’re not thinking about — I think there’s going to be incredible value.

Should people in technology have agents? If you have this sort of idea that — Keith Rabois has had this analogy of barrels versus bullets that companies only have a limited number of barrels, and a company’s capacity is constrained by how many barrels they have. And that most people are bullets — they are pointed and directed by the barrel, as it were.

AE: You know, Mike Milken said to me once — he thought teachers one day would have agents. They’re probably doing the lecture business, and probably in the book business because he said universities are really going to be vying for the top.

Well, you have this situation now. People are getting paid, AI researchers are getting millions of dollars.

AE: And I guess you could say that’s an extension of what Mike said probably about nine years ago, ten years ago. I don’t see the value, personally, that I could bring but I’d probably — with probably some VCs, that probably could happen.

Is the VC model even going to make sense, if actually it’s just going to be a world like the superstar world of Hollywood, actually going to be the model for lots of other industries?

AE: And I have a really good relationship — I think Mark thinks that VCs are going to exist.

Well, I mean, he’s very invested in making sure that’s the case.

AE: (laughing) Do you think it is?

It’s already changed tremendously. And one way that it’s changed is the first — I think — massive change was AWS, where you used to have to, your initial raising of money was literally just to buy the equipment, so you could actually express the idea and you had a shift to you could actually just build the idea, and then you could rent off of that, and as that’s progressed, right now, a lot of VC money is really just about distribution. On the consumer side, that could literally just be buying ads on the — but more traditionally, with like SaaS, it’s like sales teams, all these things. To the extent that’s the case, then it probably still has a role, because distribution is the hard thing, you’re going to have to pay for distribution. And so we’re basically — we’re already there.

AE: They are actually agents, but for companies.

And so I think actually the shift in VC already happened.

AE: They’re the agents, they’re just not for the individual AI guys, but for companies.

It’s also just a particular form of capital, the bigger threat question for VC is, “How many companies are there?”. If the models eat everything, that’s a problem, if the models end up being this commoditized layer on which you can build lots of companies and services, then it’s going to be a greater and greater.

AE: Do you think software completely goes away? I just had Ryan Smith on the phone, you know, his firm, and he’s in the software business, and he’s saying they don’t go away. Definitely in his interest, he’s got to defend his.

No, but I think the pricing pressure is going to be pretty severe.

AE: And they bought at the high, all these guys.

The credible alternative is rolling your own and it’s not just that rolling your own is going to be viable, but it’s going to arguably be desirable, because you can build the exact solution that you want. Now, the reality is, building software is fine, maintaining it is hard. I think a strength of American business generally is an overall allergy to doing stuff that’s not your differentiator.

AE: I agree. I see this in us.

I remember I went to Europe a couple years ago, and suddenly I realized why one of the reasons they’re so bananas about this data protection stuff is because, everywhere you go in Europe, they all have their own customized software that is asking for way too much information. Like, I’m going to a museum, you don’t need to know how many children I have with food allergies. Whereas everyone in the U.S. is like, “Oh, we’re not going to do that, we’re going to just sort of buy off the shelf from someone else”.

AE: Because they have so many regs. I don’t know what it is. It’s crazy how many regs they have. It’s nuts.

But this is sort of like a life philosophy thing I have, which is you will always benefit from doubling down your strengths, as opposed to trying to ameliorate your weaknesses — or fix your weaknesses. You paper over them to the extent that it’s possible, you can double down your strengths. And I think that U.S. business in general does a very good job of that and that is actually the optimistic case for software.

AE: That is the engine of U.S. capitalism.

Right. The U.S. tolerates more downside in exchange for more upside and that’s like an articulation of that. So given that, even though people could write their own software, it’s probably not going to be that smart to actually do it but there will be some cases where it just works so much better. I think the bigger problem for the software companies, at least in the short to medium term, is vastly increased competition for their customers.

AE: Well, there’s going to be people that are going to do it. And then all of a sudden, one’s going to be incredible, because they’ve written this incredible software that competes with an existing — and then there’s going to be new competition, pricing pressure.

If you had your lane and everyone had their lane, if AI actually supercharges productivity for every company, they’re going to want to expand horizontally, which means there’s going to be a lot more competition in other software companies. Which again is going to compete down margins. I think a lot of margin pressure, I think there will still be a software market.

AE: A lot of these PE firms that invested at the high end are going to get crushed.

Oh yeah. Yeah, like if you had money in, in 2021 or so.

AI and Generational Change

AE: Now, here’s the other question I have for you, what inning are we in on AI?

It depends on what framework you’re looking at.

AE: I’m just asking you, your framework right now.

Very early. There’s a sequence of games, if you want to use the inning metaphor. By the way, I’m glad I’m in L.A. and know that you are not a Dodgers fan.

AE: I’m a Cubs fan.

I know. I’m a Brewers fan. I don’t like the Cubs either

AE: You don’t like the Cubs? Yeah well, you’ve killed us. So shut up.

(laughing) Back to innings. The game right now is the devouring of everything digital. And that is going to play out for a while. Any company that is digital is threatened. They have opportunities, but they’re also very threatened. There’s, I think, the next game is going into the physical world.

AE: What do you mean by that?

Robots. Like actual, like manifesting into — and I think that game is a lot further away than people think.

AE: Because, let me just tell you something. I went up and I saw the hand that Elon’s building, which is incredible. It’s still difficult, very difficult to manufacture, very difficult to train. Chinese are not doing it any better. It’s hard.

No, especially when it comes to fine motor movements. Really, really hard.

AE: Yeah, it’s really, really hard. Will they get it? I think they’ll get it. I don’t know if it’s in two years.

Not just that, but once you’re in the physical world — I mean, you can already see companies struggling with inference having marginal cost, when they’re used to compute basically being treated as free. Now, if you think about costs, it sort of throws a wrench in things. When you’re dealing with actual physical robots that will break down, and maintenance, and all these sorts of things, the cost structures are just totally foreign as well. So I think that’s quite a ways away.

AE: Well, I say to everybody, Elon was talking about copilot or self-driving 12 years ago, 8 years ago, something like that. It’s now incredible, the car, even though he said probably a lot sooner, it’s now incredible.

This is the Elon experience in a nutshell — what he says, the timing’s not always right. But what he says is right.

AE: It’s going to happen. I just think the timing is a lot — you know, is it a decade? Is it five years? We’re a little bit off. But is that coming? And when that comes, it’s going to be an incredible — GDP is going to go crazy.

Right. Well, but I think you can see even in the digital area, where everything in digital is threatened right now — and yet, not that much has changed, even though the AI’s capabilities are very high, because change takes time and one of my theses about SaaS companies, to sort of go back, is SaaS is actually a demographic story.

AE: SaaS is a demographic story?

That’s right.

AE: What the fuck do you mean by that?

So of course, you had to have the technology for SaaS to be possible — ubiquitous broadband, browsers with sort of capabilities to do web apps. But you also needed a generation entering the workforce that was used to clicking around on web pages to do stuff and that wasn’t a foreign concept.

AE: Yeah, and that took a little bit of time.

It took a new generation. And where, of course, “There’s a web app to do this sort of thing”, is the default assumption. You go back to the ’80s.

AE: What was the first SaaS company?

Depends how you define it, but I would say Microsoft, in that they really pioneered subscription pricing, it took a while to actually get there, that’s a business model question. I think from startups, I think Box is one of them that was very early. Poor Aaron Levie, he just got killed on his — he had such a vision for SaaS, but it was so early, it was so hard for him to raise money, that by the time he finally got that company to market, he barely owns a tiny slice of it even though he sort of helped write—

AE: He’s really smart.

Their prospectus is actually really interesting, they went out the first time, and total rejection by the market so they withdrew their S-1 and I actually talked to him about this at the time. And what came up with for the second S-1 was cohort analysis. This idea that — you need to not look at the company as a whole, because it’s going to be unprofitable, you look at a cohort, and it grows in profitability over time. And once you see that, you realize you actually want your SaaS company to be losing money, because the way monetization works, the more money you’re losing, the more your upside is actually increasing. Which ended up being — that was every S-1 that came after that had a cohort analysis in it.

AE: That’s a little history.

Yeah, a little history. But you go back to the ’80s, you had people bringing PCs to work — like toy PCs, like Commodores or Apple IIs, or even Apple I before that and that’s how the whole IBM/Windows story happens. IT administrators are like, “We want them to buy IBM, can you please make something for them?”. But even then, people were using it as a glorified typewriter. There’s just the generational component of — how are most people using AI today? As glorified search. That is the analogy to the glorified typewriter.

AE: A little bit of topspin on it, it will take a generation. I mean, my wife now — she’s designing, for her fashion line, almost everything, then giving it to designers, designers are tweaking it, it’s going back and forth, that will just become like one more iteration, “It is what it is”. But at the top end, you still need the high-end designer to make sure the tweaks are human.

The problem is the AI is sort of — it’s a probability distribution, you’re sort of definitionally getting middlebrow.

AE: You’re getting — what did George Gilder say? You’re getting the greatest average.

Right. But the other thing with AI is, if you just ask a general question, that’s what you get. This is part of my pushback on a lot of the safety stuff — AI is still a responsive entity, it responds to a prompt.

AE: As I like to say, it’s like watching television, you’re still leaning back.

The thing is, the AI will give you what you put in. And so if you talk to the AI about an agent-related question, the quality of analysis that you would get out of the AI would probably blow you away, this is really, really good. If I ask the AI an agent question, it’s going to be the most Wikipedia-esque, basic thing. I’ve talked about this, there was a friend — there was some court case that happened, I think it was something App Store related, it was relatively early on. I’m like, “It was unbelievable how good the AI analysis was”, because I was going back and forth with it, and he’s like, “Oh really? I wasn’t impressed at all”, and I looked — he had asked the AI, “What do I think about this case?”, and it was the most generic answer, it’s such a general question. Whereas I already had an angle on it, because I knew a lot about the area. But I also asked a very specific legal question about it, and then it was like very in-depth.

AE: That’s why I think there’s going to be an incredible creative unlock in our business. When Baz Luhrmann is inputting what he wants as a scene, as a shot, etc., and he’s giving it exactly what you’re saying — what he needs — or Ben Affleck, or Pete Berg, or whatever — the conversation is completely different from you doing it.

Right — I have no idea what to ask for.

AE: Yeah, and I just think there is that 5% that I do not think the AI, because it’s the greatest average, is going to be able to kind of take the next step. I don’t believe it gets past that point. I still think the human condition is needed, and that’s where the greatest value comes from.

Right. Also, for stuff that is sort of net new, and taste — taste is not a verifiable function.

AE: It’s the black box.

MARI

That’s right. So one thing that was interesting, you mentioned the politics angle before, I want to ask you another question about your phone call thing. You mentioned you called one of your most famous people you represented — who was obviously Donald Trump, who you fired in 2016.

AE: I didn’t fire him! The President shouldn’t be making money. I can’t make money, it’s okay.

You related a phone call that you had in 2024, after he was reelected, and I love the anecdote, it was just like a very sort of back-and-forth conversation. Is he the one person in the world who likes to talk on the phone more than you?

AE: He likes to talk on the phone, he always picks up, he’s incredible on the phone. He’ll pick up, talk to you about anything, and you can say anything you want to him. Tell him your point of view on it, he’ll push back or not, he’ll take it in, he’ll say what things are that are on his mind, and you can push, have a conversation, and then he’s off the phone.

Right. Well, because you mentioned this in the context of, obviously, Dana White spoke at the Republican Convention, Elon Musk was part of the administration, you’re close to both for obvious reasons. And you said in the book, like, “Yeah, I lost the election, but I’m not going to not be friends with half the world”.

AE: Yeah, I’m not going to not talk to you.

I completely agree with, I think that’s a great sentiment. How much of that is actually intertwined with you being the phone guy? Like, how much of the feeling, like, “I can’t even relate to or connect to people on the other side”, is because people are living in text?

AE: I don’t think you can get a sense of a person unless you’re either Zooming or FaceTiming with them, or talking to them. I said it at the beginning — I don’t know if you’re joking or being serious on a text, it’s just too hard, I’m not that sophisticated, and it gets completely misinterpreted.

If I pick up the phone, you can say whatever you want to me, you can be aggressive, not aggressive, I can be aggressive, not aggressive. You know exactly who you’re talking to. Or you FaceTime, or you’re doing a Zoom — that’s human interaction. I mean, it’s a little bit distant if it’s a Zoom or a FaceTime, but it’s, you know, during COVID we had to do that but that’s how people can truly understand people and get across their opinion. A lot of times with Elon, I’m texting.

But you have a foundation of talking to him in person.

AE: A long time. But at least there’s a conversation happening, and you believe it’s familial — that you’re humans and you’re talking.

Right, it gives the benefit of the doubt. To the extent that AI comes along and wipes out digital, because everything digital is under threat, and AI is actually going to do everything digital better — things like MARI are more interesting and compelling. Not just because there’s more free time.

AE: I think the multiples on MARI are — not that they go up like AI, but are on the same path.

I hate that people have framed it as the anti-AI bet. It’s a pro-AI bet, you guys are going to consume all this other stuff.

AE: When I say — I’ve said the phrase “anti,” meaning the opposite bet of AI.

It’s like the bizarro-AI bet. Yeah.

AE: Yeah, it’s like — what’s going to happen in the world? Well, this is going to have to happen, because we’re human, we’re social animals, I think the multiples are going to go up. Not maybe in comparison, because it’s just the economics between them — but they’re going to go up in live.

Are you going to actually solve political polarization? Because digital — how much is that actually a driver of this? And people are going to get back to the real world, talking to people.

AE: Here’s my opinion about it. I think when you have live events, or live sports, or live music like Michael Rapino, it brings people from all social elements together.

I mean, one of the things my brother Rahm Emanuel is saying — he wants to bring back a requirement that people do service for the country — because it’s the way it used to be, where somebody from X neighborhood and somebody from Y neighborhood that would have never interfaced, do service to the country for 18 months and that we all kind of understand each other. When I go to a baseball game — I was at a baseball game, sitting watching the Cubs win — there was a guy next to me, he’s definitely a Republican.

It was obviously not a Brewers game, because the Cubs won.

AE: No, it wasn’t a Brewers game. We didn’t care that he was Republican and I was Democrat. We were just cheering for them, and it kind of broke all barriers and that’s hopefully the two businesses that I’m in heavily. That’s the great thing about it.

For sure and that’s very optimistic. I guess the question, though, is if you think about the MARI business model, and you go to like the Miami Open or something like that — is there also an underlying thesis that actually the rich are just very under-monetized?

AE: I think high-end experiences are only getting better, because there’s a great deal of rich people that want singular experiences. But we create environments for all. And, you know, there’s inside the tennis tournament, there’s XYZ, and outside there’s another 10,000 people watching it on all the screens. And Larry Ellison and David Ellison do it, and their event — we do it in Madrid also — where anybody can come and have an experience so I think we’re trying to optimize for all.

And what overall — is there a thesis with MARI that you can sort of systematize this delivery of live experiences, and do that at scale?

AE: So we have it broken down. There’s the sports vertical, which is our tennis tournaments, our participatory sports. We have our collectible business, which is Frieze, Barrett-Jackson, Collect-A-Con, which is our Pokémon experience. Then we have, like, a holiday — Winter Wonderland, our food festivals — all the stuff in that. And then we just executed against the theater business. So the West End and Broadway — we have 70 theaters in this company called ATG.

We think that is a kind of incredible — there hasn’t been one person on the live event side that exists. In the music space, it’s Live Nation and AEG. In live events, there are onesies and twosies, and so we’re trying to capture that. And what we do is, between taking things international when it’s a domestic thing — like Frieze, it was in London and New York. Now there’s two events in London, one in New York, now there’s two events in New York, one in Chicago, one in L.A., one in South Korea. And now we’re just — I think next month, we’re doing one in Abu Dhabi, and there’s other areas.

We’re going to do the same thing with Collect-A-Con on the Pokémon business, you saw how we’ve expanded when we had UFC — performance institutes and our fights all over the globe now. Then we do unbelievable sponsorship, site fees now, that’s what we do, so we’ve kind of got it down to a system of how we monetize the assets.

So what is next? Is it just acquiring more events, then, so you can plug into the system?

AE: It’s really taking our events, like Frieze, and we’re looking at different art fairs all over the place. Barrett-Jackson — can we expand it internationally? Other car auction businesses we can get into? Are there other things like Collect-A-Con — that’s Pokémon, are there other ones that we could do? And so we just look at them. When ATG came available, it’s the perfect kind of vertical that we didn’t have, that we wanted to have.

When you think about these events — but also do it like TKO and UFC and WWE — what do you think, from a business model perspective? Is it ticketing? Is it advertising? What’s the balance between those?

AE: It first starts with broadcast globally. Then from broadcast, you’re the sponsorship. Then where you take the events internationally, can you get site fees? Then you’re adding — there’s a customer that wants really high end, and will pay for really high end. And then there’s the general public, making sure their experience is incredible. And so those four or five stacks are how we think about the business, to maximize economics.

How does that align with the way you think about UFC and WWE?

AE: Those are all the things that it started. This concept started with the UFC and then we brought our sponsorship business, our site fee business to the WWE, our premium ticketing business, our ability — Mark Shapiro and myself — to make sure our broadcast deals, domestically, internationally, all of that.

For those broadcast rights business — are those, in the long run, going to be about driving subscribers or driving ads?

AE: I think for each company it’s different, but they’re all moving to an SVOD and AVOD service.

One of the things that we saw early on at the UFC, we knew that they were going to have to get to an AVOD — from a SaaS model to an advertising-based model and one of the greatest drivers there is sports, because it’s the greatest reality show. The Monday Night game — you don’t know what’s going to happen. Overtime, you think the Rams are going to win, the Bills won, and it’s the greatest model. It’s the same thing with our sports so it’s one of the great drivers and you now have four big competitors.

Do you have any concern about ending up in sort of monopsony situation where there’s not enough buyers?

AE: No. They’re all going to need it.

And there’s going to be a bunch of them?

AE: You’re now going to see what happens with Comcast and Apple. You have YouTube out there, I think all these models are going to change. One of the fastest things growing for David Ellison at Skydance is Pluto.

Right. Fox put together Roku and Tubi, which I thought was a great deal.

AE: So you’re going to have a lot of bidders. The one thing that’s great — and I think about this all the time — if the Lakers are $12.5 billion. Fantastic. Love the franchise.

Good story in your book about buying tickets.

AE: That’s a team. And that’s not the league. If a team is worth that, what are my four leagues worth? Because we have boxing, WWE.

Is that part of the allure of the UFC and WWE — you kind of get the whole thing?

AE: Yeah, running a league is different than running a team, they’re both very difficult to do. You know, they both have their personnel issues and management issues and etc. But, you know, Mark and I —

The difficulty is the moat.

AE: Mark and I run leagues. Those are hard things to do, as opposed to just a team. You’re thinking about different things, and if a team is worth that — what are the four leagues worth? I have a different conversation than the stock market is telling us right at this moment.

Is the sporting asset, given your whole model, that you are the most jealous of, Formula One?

AE: No! I think it’s an incredible sport.

Just all the things you’re talking about — the combination of the in-person events plus the rights, and all those sorts of things.

AE: I mean, think about it. First of all, I love golf, I love snooker. I mean, there’s a lot of sports that I love, that doesn’t mean that I want to be in business with them, or run them.

Hollywood Logic

Is there a world — just to go back historically a little bit — where you see streaming coming, and streaming shows up, and it’s actually this big boom, because it’s an extra window?

AE: At the time, remember, you had network syndication, so on the RSNs, you then had a cable window, it was incredible. And then all of a sudden Netflix came, you had a third window.

And is there a world where that actually persists?

AE: Well, I don’t think the syndication model — first of all, most shows are happening on streaming. You don’t get to 100 episodes, which was required.

But the reason they’re happening on streaming is because everyone decided they need to compete with Netflix. Instead, there’s a world where everyone accepts Netflix as the permanent third window, and actually makes more money all along instead of shooting themselves in the foot.

AE: That’s a logical conversation. But let me say the other side of this conversation. Do I think the network business goes to zero in cable? No, I don’t. So there is going to be a play. Remember, the original — when I was in the business, you needed to get to 100 episodes to get into syndication, now you’re doing 40 episodes.

That’s a lot these days.

AE: That’s like four seasons. So what’s the syndication window to that? It’s different. And Netflix is owning them, HBO is owning them, you’re just negotiating for the profit participants.

Is Hollywood logical?

AE: Yeah. That’s the only way I know — that you’ve been able to create over 100 years of great entertainment on the global scale, one of our great exports in the United States. Don’t you think?

Oh, for sure. Massive soft-power instrument.

AE: And I think it’s probably on the export basis — I might be wrong about this — probably our fifth or sixth most important export economically.

I write about media along with tech and Hollywood to a certain extent and my hit rate feels lower than it is, sort of, with just more tech companies and I hate the explanation that it’s because Hollywood acts more irrational.

AE: Not on every individual movie or television show. Because what do they say — if you’re batting .300 in baseball, you’re in the Hall of Fame. Same thing in Hollywood. So that’s, you know, there’s 66% — right — you’re screwing up.

Well, you mentioned David Ellison several times. Why is what he’s doing make sense?

AE: He believes if you change the tech — he believes that there’s going to be more consumption, economics for creating products is going to come down, that he’s going to be able to create unbelievable value in the SaaS model, and also take this direct-to-consumer business with the ad-based model that he has now, and also create vertical content on-site that he can kind of capture greater value than it currently exists. He’s the only one that will own his cloud, own his tech, etc. Netflix has AWS, I think Disney’s trying to change its tech right now.

Why do you need to own your own cloud?

AE: Well, I think he thinks — I think by owning that data, he’s going to get more understanding of the consumer and etc. We’ll see, that’s his point of view.

Can I ask you a question? Did you like the book?

I did. I was plowing through — no one cares about my personal schedule, but it’s been very fraught, but I was listening to it while driving, and I discovered on the plane I could listen at 3.5x speed and read at the same time, and it was unbelievable, I’ve actually never done this before.

AE: I do it all the time.

Both at the same time.

AE: I don’t look at the book. When I’m listening, I usually do a 3, 3.2x — you fly.

But if you’re listening and reading, you enter such a zone. I was flying — I flew here this morning from Chicago to L.A., I was shocked when we were landing. But it helped. I thought the book was very entertaining, I enjoyed it. How much did you write versus—

AE: You know, what we did is, when we started in 2016, [J.R. Moehringer] just kept on asking me questions. Then I answered all the questions. We met once a month, and every other week, every other month, and then he first met all my assistants.

I was going to ask you, how many assistants have you had?

AE: A lot. Some of them had kept journals, it was the craziest thing, then they met family members, friends and so he just compiled it, and then he and I went back as he was piecing it together. Then we were supposed to end when we first went public, but we didn’t go public, and then we just continued until it was 10 years.

What would you add if the book had come out a year later? What’s missing from it?

AE: I don’t know. The response has been kind of crazy from parents that have dyslexic kids, people that have dyslexia, people loving — like, re-understanding that being aggressive and rage, and taking agency for yourself, is a good thing, not a bad thing, because they’ve grown up in this place where you have to be delicate. It’s been crazy. So all in all, I’m completely happy to where it’s at right now.

Yeah. The bit about you know, you frame it as rage, but really just agency — forcing resolution.

AE: Yeah. I think that’s important to get to the end. I was having a conversation today about it. It’s like, why do I want to play this stupid game? Like — I know what you’re doing, you know that I know, let’s just have the conversation. You have to do this for this reason, I know that, just get to it already. I cannot stand this whole gray area. I can say I don’t like you, or I like you, and here’s what you did to me, and here’s how I feel about it, and go fuck yourself, and it’s ok.

Isn’t that, like, very — particularly on the West Coast — West Coast always strikes me as much more indirect, passive-aggressive.

AE: Oh you don’t think Wall Street or Silicon Valley is indirect?

No, Silicon Valley is. I think Wall Street’s much more direct.

AE: Oh you do?

I was thinking about the East Coast versus West Coast, you’re from Chicago.

AE: We’re pretty direct in Chicago.

Yeah. Is that the real secret?

AE: Maybe. I mean, it’s the only way I know how to do it, maybe that’s because I grew up with my mom and dad, and that’s what we did at our dinner table. Well, thank you.

I’ll remind you — we’ll have an ending, since we didn’t have a beginning, but thank you for coming on to Stratechery.

AE: (lauging) I appreciate it.

This Daily Update Interview is also available as a podcast. To receive it in your podcast player, visit Stratechery.

The Daily Update is intended for a single recipient, but occasional forwarding is totally fine! If you would like to order multiple subscriptions for your team with a group discount (minimum 5), please contact me directly.

Thanks for being a supporter, and have a great day!