Infinite Content: Introduction
You Can See it Coming from Odessa
This is the draft introductory chapter of my book, Infinite Content: AI, The Next Great Disruption of Media, and How to Navigate What’s Coming, due to be published by The MIT Press in 2026. This chapter is available for free. Subsequent draft chapters will be serialized for paid subscribers to The Mediator.
In spring 2018, I was one of several dozen Time Warner executives invited to AT&T headquarters in Dallas. AT&T and Time Warner agreed to merge almost two years before, and the deal was finally limping toward regulatory approval. AT&T management had limited say over Time Warner’s operational or strategic decisions before the close, but they wanted to have some broad discussions to prepare for the post-merger integration.
The Time Warner contingent coming from Burbank was delayed getting into Dallas by bad weather. In his opening remarks to the group, AT&T CEO Randall Stephenson quasi-apologized, with a disarming combination of folksiness and menace.
“I understand some of you had trouble getting in. Well, if you come to Dallas in May, there’s going to be thunderstorms. But, as they say, if you get wet, you’re a stupid sonofabitch, because you can see it comin’ all the way from Odessa!” (Like most of the Time Warner folks, I had no idea that Odessa, Texas is about 350 miles west of Dallas. But I got the point.)
I’m writing a book about media disruption because I can see it coming and I don’t want to be a stupid sonofabitch twice.
Not a Casual Observer
Over the last twenty years, I witnessed the disruption of the TV business firsthand. That doesn’t make me special. Anyone who has worked in media for a while has lived through disruption. What makes me special(ish) is that I haven’t been a casual observer. For three decades, it has been my job to understand the big shifts in media—recognize them as they’re happening, determine how they fit together, predict how they influence what happens next, and figure out what to do about it. Extract signal, synthesize, forecast, call to action. Rinse and repeat.
In 2005, I was working at Bank of America as an equity analyst, covering U.S. cable, satellite and entertainment stocks. The consumer internet was a relatively new phenomenon. Only about one-third of U.S. households had broadband access at home. Facebook was a year old and still called TheFacebook. The iPhone wouldn’t launch for a couple of years.
My clients were professional money managers. Some were starting to worry that internet-delivered television, or what was then called IPTV, would be a new competitive threat for the cable and satellite TV providers that I covered, like Comcast and DirecTV. (The phrases “over-the-top (OTT)” or today’s preferred term, “streaming video” weren’t in use yet.) A small company that rented DVDs by mail, with the aspirational name Netflix, was talking up its plans to eventually deliver video over the internet. The music and print industries were already deep in the throes of disruption. Maybe TV was next?
I wrote a report arguing IPTV wasn’t a threat to traditional TV. My rationale was that it wouldn’t be technically possible to offer a competitive product to cable at a comparable price for a very long time and the TV networks wouldn’t license their content to a new distributor and undermine their own businesses. It took almost a decade, but I would be proved wrong. I didn’t understand how quickly a general purpose technology, like the internet, can drive the cost of a critical input in the value chain toward zero. I also didn’t understand the disruption process, and how lower entry barriers can pave the way for an inferior product to turn a market upside down as well-intentioned, but hapless, incumbents stand by.
In 2007, I left Wall Street for Time Warner, to run investor relations. Time Warner would eventually turn out to be one of those well-intentioned, but hapless, incumbents. Over the next decade or so, I watched that disruption process play out in real time, kind of like watching those thunderheads roll in from Odessa. Seeing that disruption happen, despite our best efforts to stop it, was the most formative experience of my career.
As the head of investor relations, I was the primary conduit between Time Warner and Wall Street. It was my job to keep senior management and the board apprised about investor perception of the company and craft the messaging to investors. Time Warner was badly damaged by its merger with AOL in 2001, which is still often cited as the worst in corporate history. But by 2007, it had stabilized, had an $80 billion market capitalization, and was one of the most powerful media companies in the world, comprising AOL, HBO, Time, Inc., Time Warner Cable, Turner, and Warner Bros. Netflix’s market cap was $2 billion.
A year earlier, Netflix had begun offering its highest tier DVD-by-mail subscribers the ability to also stream some old movies online, at no extra charge. Few thought much of it, even within Netflix. According to The New York Times:
[CEO Reed] Hastings does not believe that electronic distribution, be it through downloads or streaming services, is ready for prime time. “The market is microscopic,” Mr. Hastings said. “DVD is going to be a very big market for a very long time.”
Within Time Warner, we didn’t pay much attention to the streaming service. We mostly talked about how Netflix’s subscription DVD rental business was undermining our DVD sales.
And then, within a few-week period at the end of 2008, we took notice. Netflix licensed rights to stream TV series from both Disney (a few Disney Channel shows, including its biggest hit at the time, Hannah Montana) and CBS (including hit scripted series, like CSI). More important, it struck a deal with Starz to license the entire Starz library (~2,500 movies), including the rights to stream Disney and Sony movies six–nine months after they were in theaters. Even though Starz was generating around $300 million annually by licensing its network to cable and satellite distributors, the deal called for Netflix to pay $25–30 million per year for the same content, delivered in a better experience (because it was all on-demand). Years later, subsequent Starz CEO Chris Albrecht would call the deal “terrible.” It wasn’t terrible just for Starz.
It was suddenly possible to foresee how Netflix might disrupt the traditional TV business. The big media companies might empower Netflix by licensing it successively better content with successively shorter windows. Consumers would shift their viewing over to Netflix and maybe even “cut the cable cord.”
Our peers and competitors were acting like licensing to Netflix was found money. At the beginning, it was. Wall Street was inconsistent about it: on the one hand, investors were concerned about the long-term implications; on the other, they applauded the near-term earnings boost. Within Time Warner, many of us believed it would inevitably come home to roost. Netflix became the foremost topic of debate and concern within the company and would remain so until the day I left, more than a decade later.
Fast forward to 2015, when I was running strategy at Turner, the largest division of Time Warner, and home to TNT, tbs, CNN, and the Cartoon Network, among other cable networks. By then, the threat posed by Netflix, Hulu, and Amazon Prime Video were no longer debatable. They were growing subscribers fast. Subscribers to our cable networks and viewing were indisputably in decline. The disruption was clearly underway.
I made a presentation to our CEO, Jeff Bewkes, walking him through the challenges facing our business and making the case, not for the first time, that we should launch a Time Warner-wide streaming service, branded HBO, and take on Netflix and the platforms (Google, Meta, etc.) head on. We needed to own the customer relationship and control the user experience and first party data. Or so I argued.
Bewkes intently listened, poking, prodding and debating, as he always did. But I didn’t convince him. The challenges I described resonated more than the opportunities. He didn’t think we could pull it off and a year later he arranged to sell Time Warner to AT&T. I disagreed then, but in hindsight, he was right. A year after that, in 2017, Rupert Murdoch agreed to sell most of the Fox assets to Disney, a decision that was even more shocking. Bewkes was a professional manager who was known for his economic rationality. Murdoch was a founder and considered by many as the archetype of the media empire builder. He bought, he didn’t sell. Until then.
Today, most of the major media companies are shells of their former selves and Bewkes and Murdoch look like the shrewdest men in media. AT&T threw in the towel only a couple of years after closing the Time Warner deal, merging the then-named Warner Media with Discovery, to create Warner Bros. Discovery (WBD). That company is now on the ropes, its stock down 65% since that merger and the constant subject of speculation that it will need to sell itself or some assets. Paramount recently agreed to merge with Skydance, a small studio, after controlling shareholder Shari Redstone realized the company couldn’t fight the trends on its own. Even Disney, long admired as the class act of Hollywood, is struggling. By contrast, Netflix now has over 300 million subscribers globally and, in the U.S., close to 70% of broadband homes subscribe. Its market capitalization exceeds $400 billion—more than Disney, Comcast, WarnerBros. Discovery, Fox and Paramount combined—and it is the most powerful company in Hollywood.
In the meantime, another disruptive force has been slowly brewing in video: so-called creator, user generated, or short form content, or my preferred term, social video. Creators reportedly upload about 300 million hours of video to YouTube annually, compared to my estimate of about 15,000 hours of new TV series and movies from Hollywood – 20,000 times as much. I also estimate that social video (which comprises viewing on YouTube, TikTok, Meta Reels and other social sites and apps) is now 25% of all video viewing in the U.S., and its share is growing. In the U.S., YouTube is the most streamed services to TVs, more than Netflix and more than almost all the other streaming services, combined.
Over the last two decades, like some merciless force of nature, disruption has left no corner of the media business untouched. Print has been devastated. Music has been to hell and back. Even gaming, the newest of these media, has not been unscathed.
It is a lot more fun to be a disruptor than a disruptee. Disruptors are on offense, their workforces unified in a sense of purpose and opportunity. Their investors are happy, public opinion is positive, founders are lionized, and early employees can create generational wealth. Disruptees are on defense. Investors flee, stock prices fall, senior executives are vilified, boards turn risk averse, and public opinion sours, often seeming to find perverse glee in the downfall of the once great. Employee morale turns nervous, anxious, depressed, and eventually desperate, as everyone who can leave, does. Being disrupted sucks.
But the disruption of media not only affects executives, employees, investors, creatives, and policymakers. It has even broader societal implications.
Media Matters
Media and entertainment may seem trivial. Perhaps the topic conjures images of self-important actors or directors preening at a movie premiere step-and-repeat. Or talking heads yelling at each other or trumping up some non-story in the hopes of drawing viewers. Or wasting hours scrolling TikTok, hooked to a dopamine drip. Or maybe a catchy but generic pop song that is the hit of the moment, but no one will remember in a few months.
I believe the opposite. Sixty years ago, Marshall McLuhan argued in Understanding Media: The Extensions of Man that media shapes the structure of society and the way people think. Yet even he didn’t foresee how profoundly media affects us today.
For most of us, media is our life, literally. One’s life is the sum of one’s experiences. It may be an unsettling thought, but most of those experiences are not with other people or nature, they are abstracted and refracted (and distracted?) through media. According to consulting firm Activate, in the U.S. the average person spends 75% of their waking hours with media of some kind. So, most of our life experience, from the time we first see a screen to the day our senses give out, comprises the media we consume.
Media affects our mental health, positively and negatively. It is a form of escapism that is cheaper and/or less dangerous than drugs, alcohol and psychotherapy, and has been empirically shown to reduce stress markers and blood pressure. But it has a flip side too. In the 1970s and ‘80s, many warned about the adverse consequences of TV. Today, debate rages about the effect of social media, particularly on young people. Many, such as social psychologist and author Jonathan Haidt, argue that social media usage is highly correlated with higher levels of depression, anxiety, and self-harm, especially among adolescents and young adults.
While scrolling Facebook or YouTube is called “social media,” all media is social. Increasingly, media is how we find our tribes. At a time when organized religion is in secular decline (yes, that is a pun), fandoms create a sense of belonging, a common vernacular, a shared value system, and, for some of us, even a moral philosophy.
Media informs much of how we see the world. It is the primary tool of conveying societal norms. It is also the lens through which we interpret most information outside of our direct experience, including what we believe to be true and righteous, false and detestable.
Media is a tool of mass coordination, for good or ill. In Sapiens, Yuval Harari argued that our success as a species is due to our ability to coordinate at scale and the mechanism of that coordination is shared stories. These stories are no longer conveyed around a campfire, but through a phone. Those seeking to gain power or keep it may use media to influence us, catalyze us, or manipulate us.
And yes, as McLuhan famously claimed, the medium is the message. The form of media we consume influences how the message is conveyed and how it is received. This includes changes that are trivial—such as the growing imperative in the “peak TV” era to hook viewers in the first episode of a series and the influence of Logic Pro on chord progressions in pop music—and profound.
Consider, for instance, that the advent of television brought with it the first telegenic U.S. president, John F. Kennedy. We can draw a line from Kennedy to Ronald Reagan, a former actor, to Donald Trump, a former celebrity game show host. The transition of news from print to TV caused another shift: from news-as-information to news-as-entertainment. The evolution of TV from broadcast to cable birthed the 24-hour news network and, with it, the realization that reinforcing what an audience already believes is more profitable than trying to educate it.
More recently, the recognition by politicians that attention is more valuable than likability has incented ever-more extreme positions. Today, noise trumps reason. The decline of mass media has brought with it the decline of a shared truth, as we splinter into like-minded groups, our echo chambers reinforced by the algorithms that feed us what they statistically infer we want to see and hear. The state of media isn’t the only cause of our current political polarization, but it has surely helped.
So, media matters, a lot. When it changes, it changes us.
The Next Disruption
The primary enabler of media disruption over the last 15-20 years was, of course, “the internet.”
The internet, combined with digitization before it, had many profound and far-reaching effects across media. Most important, it unbundled information from costly infrastructure and, with the help of a host of associated technologies, caused the cost to move bits around to plummet and lower barriers for new entrants. The current state of media can be directly traced back to these effects.
In this book, I refer to this as the “last great disruption of media.” There have been other disruptions before. The Gutenberg printing press dramatically reduced the cost of print distribution. Newspapers, radio/TV, and desktop publishing all probably felt disruptive to incumbents at the time.
Now we are on the cusp of another great disruption, but this one is different. Most prior disruptions in media primarily affected the cost of distribution. Today, the cost of a different critical input in the value chain is racing toward zero: the cost of content creation. The cost to make bits is set to plummet too, enabled by generative AI (GenAI). It will result in what I call “infinite content,” the title of this book.
You could take issue with both of those words. Hopefully, you’ll indulge me the use of the word “infinite.” I don’t argue that there will be literally infinite content, but rather that it will feel infinite to any one consumer, because content will be, and already is, created far faster than anyone can consume it. And then there’s “content,” a polarizing word that many believe denigrates art, relegating it to no more than widgets. Point taken. But I think content is the right blanket term for the stuff we watch, read, play, and listen to. Some of it is art and a lot of it isn’t.
GenAI is evolving extraordinarily fast. There are a lot of moving parts and critical unknowns. This book won’t tell you precisely how this next disruption is going to play out, who will emerge as a clear winner and clear loser and how consumer behavior will change again. But we can make sure we’re asking the right questions, map the broad outlines of the answers, and sketch out the implications. For anyone with a stake in the media business—creatives, creators, those who work at big media companies, agencies, advertisers, investors, policymakers, etc.—or even those who care about the societal and cultural effects of media—it is critical to understand the potential implications of this transition.
Layered on top of the past disruption, I am convinced it will make the media business unrecognizable within a decade. It will upend businesses, investments, and careers, but it will create great opportunities too.
For me, it is both fascinating and personal. The Winston Churchill quote that “Those that fail to learn from history are doomed to repeat it” is overworn, but for a good reason. It’s true. I don’t want to miss the importance of this change, as I did twenty years ago. If you’re reading this, you probably want to be prepared too. As the thunderheads approach, consider this book to be your umbrella.
The Right Questions
I just promised to ask the right questions. Easy to write, hard to do. To ask the right questions—and form tentative answers—here are a few meta-themes that you’ll see throughout:
Pattern recognition. The book is, ultimately, an exercise in pattern recognition. It explores the origins and implications of the prior disruption of media as a prelude to the next one. The first half or so is history lesson and the second is gazing into a murky crystal ball.
First principles. While pattern recognition is a useful jumping off point, like all heuristics, it’s a blunt tool. I also try to understand what has happened and what may happen at a first-principles level. Fair warning: It gets a little wonky in parts. I found balancing the right amount of exposition and brevity to be tough. I’m sure I missed the mark, one direction or the other, in parts. But I used myself as a proxy. How much detail do I think I need to know to explain what’s happening in an intuitive and accessible way? Not that I always achieve it, but that was my goal.
Multi-disciplinary. The future of media will be largely dictated by what happens at the intersection of three disciplines: technology (what’s possible); economics (what’s viable); and consumer behavior (what’s desirable). Neglecting any will probably lead to wrong conclusions. I especially try to steer clear of naïve technological determinism, namely thinking that just because something is technically possible, it will happen.
Empirical. Where possible, I support my conclusions with data, as you will see in the roughly 100 figures.
Frameworks. I also lean heavily on frameworks, by which I mean structured ways of thinking through problems. Since this book is about disruption, Clayton Christensen’s theory of disruptive innovation looms large throughout.
Here are some of the questions we’ll raise:
GenAI has the hallmarks of a disruptive innovation and will almost undoubtedly prove so for traditional media. But how disruptive? Will it wreak total disruption, like digital photography did to Kodak, or only partial disruption, as Airbnb has to the lodging industry?
How does GenAI shift value up and down the value chain? Who stands to win and lose?
If AI drives the cost of content creation toward zero, what will still be scarce and, for that reason, probably even more valuable?
To what degree will consumers embrace AI-generated or assisted media? How much AI is too much? And for which use cases?
Will GenAI exacerbate the fragmentation of our attention? If so, what are the implications of fewer shared cultural experiences? What does it mean for our collective culture, the social fabric, and our shared sense of truth?
If infinite content creates an infinitely splintered world, how do brand advertisers reach people?
As GenAI democratizes the means of creation, will the distinction between creation and consumption increasingly blur?
What does AI-native media look and sound like?
What new business models will AI-native media allow?
What to Expect
To be clear, I haven’t finished the book yet! But here’s the plan:
Chapters 1 and 2 provide the theoretical foundation of the book. If you’re a disruptive innovation expert and a media maven, feel free to jump ahead!
Chapter 1 explains disruption, why it happens, and why it’s important, paving the way for the discussion of media disruption throughout the book. A lot of this is based on the work of Christensen. But in this chapter, I start earlier and go further. I trace Christensen’s ideas back to Joseph Schumpeter and address questions Christensen didn’t, like why the speed and extent of disruption can vary widely; how industries and ecosystems can be disrupted more than once; and the characteristics of the most disruptive innovations—general purpose technologies (GPT). Both the internet and GenAI are GPTs.
Chapter 2 is an MVP—a minimal viable primer—on the media industry: how it makes money, how it is organized and who does what. I explain the unique economic properties of information goods and how they shape media business models. This grounding will help us understand how the media industry has changed and will likely change in the future.
Chapters 3-8 discuss the current state of media and how we got here.
Chapter 3 provides an overview of the “last great disruption” of media over the last 20 years, namely why and how the internet, building on the foundation of digitization, changed the architecture of media. Most important, it set the cost of media distribution on a path to be functionally free. I explain how this and other changes led to the tectonic trends that define the current state of the media and entertainment industry: stagnation; fragmentation; deflation; disintermediation; globalization; and concentration. We’ll walk through each over the next several chapters.
Chapter 4 surveys the cost of free distribution: stagnation, fragmentation, and deflation. Falling barriers to distribute media have resulted in an enormous supply of content. But with so many hours in a day, time spent with media has stagnated. That’s a structural constraint on industry growth. Fragmentation is as old as media itself, but it’s more extreme today, for two reasons: the tsunami of choice; and a changing consumer definition of quality in media. The deflationary effects of new media aren’t new either. But, more than 20 years into the disruption process, they are surprisingly persistent.
Chapter 5 explores disintermediation and the rise of creators. Historically, most of the biggest companies in media and entertainment—movie studios, TV networks, major music labels, game publishers, and national newspapers—have been middlemen between creatives and consumers. They have created value by doing things that are hard or practically impossible for creatives to do themselves (finance production, assemble talent, market, distribute, and monetize). But technology is systematically democratizing each of these steps in the media product development process. This is shifting bargaining power to creatives or allowing them to circumvent the traditional middlemen altogether, forge direct connections with consumers, and become creators. In this chapter, I size the creator economy and explain the many reasons why it will inevitably keep taking share of media time and money.
Chapter 6 examines another systemic change brought about by the internet: globalization. Media distribution is no longer bound by physical or geographical constraints, but by regulation, contractual rights, and cultural boundaries. Those cultural boundaries are getting more porous. As a result, the U.S. is slowly ceding its role as the dominant cultural exporter globally, as hits can emerge from anywhere—video games from China, music from Latin America, TV shows from South Korea. For the same reason, the U.S. cultural trade surplus is narrowing, as it begins to import culture.
Chapter 7 digs into the last of the tectonic themes, concentration. I try to reconcile an apparent contradiction. In Chapter 4, I described how the internet fragmented attention and in Chapter 5, how technology is weakening traditional intermediaries. But there are still huge hits and there have emerged massive new intermediaries—technology platforms. How can the internet both fragment and concentrate power and attention at the same time? The reason is that today most media is consumed on networks. Networks are subject to powerful positive feedback loops that make the strong stronger. On the supply side, these manifest as network effects that put the platforms (Apple, Amazon, Google and Meta) in an (almost) unassailable competitive position. On the demand side, they result in increasingly extreme distributions of popularity, which look like a power law: a skinny “head” of a few massive hits, a dwindling middle and a near-infinite “tail.”
Chapter 8 ties together the first half of the book by exploring case studies in disruption and how and why the fates of the print, music, TV, and gaming industries have been similar in some regards and very different in others. We’ll try to pull out some lessons here that we’ll apply in the second half of the book.
Chapters 9-14 turn to what happens next and how to navigate it.
Chapter 9 frames out the next disruption, enabled by GenAI. We start with the “known knowns.” I explain what GenAI is, why it is so revolutionary, why it is ideally suited to media, and, as a result, why it will cause the cost to make bits to plummet. It is impractical to provide a current assessment of GenAI tools in a book, because they are moving far too fast. Instead, I discuss the underlying technology of GenAI from first principles and what this tells us about its current capabilities, future potential, and inherent advantages and limitations.
Chapter 10 gets into the “known unknowns” about how GenAI will evolve. These unknowns will dictate how disruptive GenAI ultimately proves to be. They include: how realistic it will get; whether it will be capable of the fine-grained creative control that professionals will demand; to what degree consumers will accept it, and for which use cases; what GenAI-native forms of media will emerge and how compelling they’ll be; and whether and how the thorny legal issues surrounding GenAI will resolve.
Chapter 11 tries to place some shape around this uncertainty through scenario analysis. This helps to frame the most important questions and the range of potential outcomes. In this chapter, we create a scenario planning matrix by varying two critical variables, abstracted from our list of unknowns in chapter 10: technological development and consumer acceptance. This yields four scenarios: low, low (“Novelty and Niche”); high, low (“The Wary Consumer”); low, high (“Stuck in the Valley”); and high, high (“Media Mayhem”). Then we explore the most likely outcomes. One clear conclusion: at a time when the traditional markers of quality are becoming less important to consumers, GenAI doesn’t need to enable comparable production values to be very disruptive to traditional media.
Chapter 12 raises more questions than answers. It discusses the complex cultural and societal implications of an infinite content world. What happens to social cohesion, collective culture, and quality of life in a world of infinite content? A lot of us long for the good old days of mass cultural events, but mass media is only 100 years old. Maybe mass cultural events aren’t our natural state and media experiences will self-organize around different size affinity groups? To what degree will borderless digital affinity networks supplant old social structures, like local communities, religions, and nations? What will it mean to our culture if there are fewer shared references? What will be the cognitive costs of infinite choice and how will we manage them?
Chapter 13 tries to sort out winners and losers. Value always flows to scarcity, so the key question is this: as content becomes more abundant, what new scarcities will emerge, what existing scarcities will become more valuable, and what businesses (and business models) will be newly viable? In this chapter I try to answer that question and discuss what that will mean for traditional studios and publishers, creatives, creators, distributors, platforms, and consumers.
Chapter 14 lays out what to do about all this. Change, as unsettling as it is, creates the greatest opportunities for those who seek to understand and leverage it. That is, after all, why I’m writing this book. Much is unknown and unknowable about our infinite content future, but here I lay out what I believe are the 10 key principles for anyone trying to navigate this transformation.
Subsequent draft chapters of Infinite Content will be serialized for paid subscribers to The Mediator.
Never tell anybody your plans, unless telling helps you achieve them.
Speaking of "the medium IS the message...I notice you used the verb "to trump" twice in this chapter. DONT let your editor remove these. The use is totally apropos.