What this means for you:

• Your subscription will continue seamlessly - Substack is handling the technical migration

• Paid subscribers: Your billing will transfer automatically through Stripe

• All free tier subscribers: Your email address will be imported to the new platform

No action is required on your part. I will share more details about this transition and exciting upcoming changes next week. Thank you for being part of The Medium community and for your understanding during this transition!]

[sAuthor's Note #2: I appeared in this week’s Next in Media podcast with Mike Shields, one of my favorite writers and reporters focused on the transformation of media business models. We discussed the next decade in sports streaming, and why sports fans are more confused, overwhelmed and under-served than ever. [Apple | Spotify | YouTube]

A surprise resolution emerged yesterday morning between Amazon and the Broccoli family—who have long controlled the rights to adapt Ian Fleming‘s 007 novels—ending a standoff that began with Amazon’s $6.5 billion acquisition of MGM Studios in 2022.

At its core, the dispute was “a clash between the 20th-century Hollywood of big screens and big swings and a new entertainment industry ruled by Silicon Valley firms that prize data, algorithms and streaming subscriptions.” 

The Broccoli family refused to entrust their James Bond intellectual property (IP) to an “algorithm-centric Amazon”. A previous contractual agreement with MGM Studios promised only theatrical releases of Bond movies, and Amazon had agreed to honor it. However, according to The Wall Street Journal, that did not stop executives from suggesting new ideas:

“Would Amazon produce a James Bond TV show for its Prime Video service? What about a Moneypenny spinoff? Or a TV spinoff centered on a female 007?”

Broccoli reportedly regarded the substance of those questions as a “death knell” for the Bond franchise. 

That all changed yesterday: The Broccolis will cede creative control to a new joint venture with Amazon MGM which will house the franchise’s IP rights. Amazon will now control who will play Bond, who will write the next script, and when the film goes into production. 

What spurred this change? Those answers will likely emerge through the media over time. 

A more important question is, what can this joint venture accomplish that neither the Broccolis or a traditional Hollywood studio could before? With generative artificial intelligence (AI) and internet distribution models reshaping the business of storytelling, the possibilities seem endless. It may be easiest to frame an analysis of the possibilities within the three areas of Amazon’s creative control.

Key Takeaways

The Bond franchise‘s future with Amazon MGM will be shaped less by the creative control of either Amazon or the Broccolis, and more by audiences’ desire to fund and/or create new stories across interactive, multi-format platforms

Total words: 1,400

Total time reading: 5 minutes

1. The Actor

Shortly after the announcement of the joint venture, Amazon founder Jeff Bezos posted a screenshot of a BBC News story on the deal and asked his 4.3 million followers on Instagram: “Who’d you pick as the next Bond?

In the hands of Amazon, there may be multiple answers to that question. Bond could simultaneously exist across movies, TV, games and/or generative AI. The TV proposals, above, imply a plan to fragment the James Bond universe much like Marvel and DC Comics TV and movie universes, where multiple actors play the same superhero exist simultaneously across multiverses (“Spider-Man: No Way Home”) or younger and older versions of the same character. It also mirrors how traditional TV studios extend popular shows through character-driven spin-offs.

Whereas in the James Bond franchise, there has always been one actor at a time playing Bond or a supporting character (e.g., Ms. Moneypenny). Also, Deadline reports there is “a slew of Bond villains and women in their own series or features” who are also available.

Theoretically, there are no limits to where Amazon can license the IP. According to The Wall Street Journal, Broccoli “had already turned down TV shows, videogames and at least one tie-in casino before Amazon entered the picture.” 

Should Amazon consider Generative AI or gaming, it will encounter layers of complexity that I discussed in “Meta's AI Vision vs. The Art of Storytelling”. One key concern is AI's ability to easily recreate one the voices and/or likenesses of actors, Actor Hank Azaria wrote in a recent New York Times opinion piece, “There’s so much of who I am that goes into creating a voice. How can the computer conjure all that?” 

These concerns drove Hollywood voice actors and motion capture artists in video games to strike against American video game companies last summer. The technology to reuse a likeness or modify a voice has existed for years, but actors argue that “AI ups the ante because it can scrape more information more efficiently and potentially turn it into a plausible clone of an actor, combine actors’ work or pass as a new, ersatz artist.” 

Two bills recently signed by California Governor Gavin Newsom protects both dead performers (granting rights to performers’ estates) and living performers (creating a use case of “reasonable specificity”). Even with a more flexible regime, Amazon faces obstacles to digitally resurrecting past James Bonds—like Sean Connery—or inventing new ones with AI.

2. The Script

The agreement allows Amazon to determine “who will write the next script” for the film. Because the joint venture gives Amazon creative control over the Bond IP, development across TV shows, games and movies seems inevitable. Generative AI offerings seem both possible—given the technology’s infancy—and speculative.

Scripts for a film versus a TV show or game are different end-products. A film has a single script developed by a single screenwriter, whereas TV shows require multiple scripts—typically eight to ten episodes per season in the streaming era—developed by a writing staff.

In gaming, there are no scripts. According to a Reddit post from a writer/designer of blockbuster (“AAA”) games, a team of writers start with “broad narrative conceits that (hopefully) have something to do with the game's planned genre, setting, and gameplay loop/mechanics.” He describes good writers in games as “those who are able to piece together a coherent story as the development process inevitably brutalizes their work.”

In generative AI, there also are no scripts, only the instructions a user enters into text-to-video prompts. My recent essays suggest two possible paths for the Bond IP.

Fable Simulation's Showrunner platform blends gaming and streaming models. The platform lets consumers create personalized TV animated episodes by entering text parameters for world, character, genre, and mood. The “script” is generated by AI based on a text-to-video prompt. An animated Bond property using this technology could be more lucrative than a traditional TV series on Amazon Prime Video.

Meta presents another compelling destination for James Bond IP in AI. CEO Mark Zuckerberg has recently argued that “individual creators or publishers tend to overestimate the value of their specific content in the grand scheme of [generative AI].” Only certain premium content from books, TV and movies will be “really important and valuable” and result in partnerships. His framing suggests Meta AI will enable creators and “hundreds of millions of small businesses” to use beloved characters from valuable IP like Bond in advertising and original, AI-generated content.

3. Film vs. Other Formats

If we believe yesterday’s press coverage, Amazon will begin production on its next Bond film after it makes two decisions: choosing the next James Bond actor and approving the script. A theatrical release makes sense: The last five Bond movies starring Daniel Craig grossed over $1 billion, with 2012’s “Skyfall” alone earning $300 million. 

However, the universe-building potential of the Bond IP suggests that may be the wrong takeaway. The internet as a medium favors multiple content models, as I argued in “Ben Affleck, Marshall McLuhan & The New Business of Living Room Content”. Traditional movies and TV shows are becoming “less bankable products” as content supply evolves within internet distribution models.

If Amazon indeed aims to build a universe where they can monetize IP in theatres and on Amazon, they are still competing against YouTube. They must solve for the “new business of living room content“, where ”YouTube is proving that consumers‘ ’choice and control’ over the medium of the internet means fans are the producers of the content they want to watch.” 

Netflix seems pressured by this dynamic to produce more content at lower costs, potentially forcing them “to break [the business]" and undermine their studio model. Amazon Prime Video faces similar challenges at a smaller scale: It captures 40% of the monthly TV viewing in the U.S., according to Nielsen’s most recent The Gauge—and with a smaller user base worldwide (200 million Prime subscribers to Netflix’s 300 million). 

Bond & The New Business of Living Room Content

It may be no accident that Bezos immediately turned to social media to crowdsource the next Bond. Amazon seems to have realized “the message of the internet medium” is the blurring of lines between fans as consumers and fans as producers. Consumer “investment” in content has evolved from emotional to financial engagement—YouTube’s Partner Program is driving this shift across 3 million-plus creators and hundreds of millions of televisions worldwide—“so supply must evolve, too.”

Amazon may envision the Bond franchise as *the* catalyst for a broader supply-side paradigm shift across movies, TV, games and/or generative AI. This would contrasts starkly with the Broccolis’ tradition of “big-screen storytelling and gut instinct.” 

The Amazon-Broccoli joint venture may enable Bond to transcend legacy media business models by embracing audiences desire to fund and/or create their own content with the hero, his allies, and his villains. Amazon MGM’s creative vision for Bond will be secondary to audiences’ imaginations and wallets.

Free ad-supported TV (FAST) is having a strong week. On Sunday, Fox-owned Tubi attracted 13.6MM of 127.7 million Super Bowl viewers (peak 137.7MM). That surpassed Amazon Prime’s average viewership for Thursday Night Football by only 400,000. Tubi’s Super Bowl ad—about a young man born with a fleshy cowboy hat as part of his head—delivered its own sales pitch of “choice and control”: “If It’s In You, It’s In Here.”

On Tuesday, YouTube CEO Neal Mohan revealed in his annual letter that “TV has surpassed mobile and is now the primary device for YouTube viewing in the U.S. (by watch time). According to Nielsen, YouTube has led streaming watch time in the U.S. for two years.

This validates my observation from last September:  “creator content increasingly competes with professionally produced content. Consumers no longer perceive Hollywood’s content ‘quality’ as an impetus for spending money on media or devoting time. ” 

Both developments present a difficult question for large media companies not named Fox (and including Netflix): How will they adapt to this new competitive balance?

Key Takeaways

The success of Tubi‘s Super Bowl strategy and YouTube’s living room dominance reveal “portfolio reconstruction” need not only focus on gaming and AI.

Total words: 1,000

Total time reading: 4 minutes

Three Models for Gaming & Television

In “Reconstructing The Media Conglomerate Portfolio For AI, Games & Streaming”, I identified three emerging business models that compete at the intersection of gaming and television:

1. Startups going down a rabbit hole with a startup where games, TV, streaming and decentralized media all intersect (e.g., Fable Simulation’s Showrunner)

2. Focus on earnings before interest, depreciation and amortization (EBITDA) and games but build the “wrong” product (Warner Bros. Discovery, “New Paramount”)

3. Focus on EBITDA and “crawl, walk, run” to games and the right product (Netflix)

These market dynamics each and all raise the question of why large media companies still need to exist. In particular, Tubi, YouTube and  Showrunner all present big media companies with a limited existential purpose: To receive checks from “makes” and “views” of its licensed IP. 

However, these companies face a fiduciary duty to shareholders to act in their best interests. That duty conflicts with a growing reliance on accounts receivable and/or funding business models in decline (and vulnerable to disruption from AI). 

The Trouble with Tubi

Fox’s bet on Tubi as a outlet for the Super Bowl offered cord-cutters free viewing in exchange for email registration.

That left it is less vulnerable—if not all-but-invulnerable—to churn than its competitors: Research firm Antenna reports 35% of Super Bowl LVIII weekend Sign-ups remained after eight months and 35% of Super LVI weekend Sign-ups remained on Peacock after ten months. Paramount+ also never disclosed the Nielsen figures for total number of viewers who tuned in.  

This presents uncomfortable questions for Disney, NBCUniversal and Paramount:

1. Should NBCUniversal and Paramount distribute the Super Bowl on free services (Xumo or Pluto, respectively)?

2. Has Tubi changed consumer expectations for tentpole sports events for subscription streaming? If so, how will that impact Disney’s upcoming launch of its ESPN flagship?

NBCU will test these questions with Super Bowl LX in 2026. They have 36 million subscribers to Peacock but have told investors that they would consider partnerships in streaming “despite their complexities.”

They are also pursuing “portfolio reconstruction” this year: A “SpinCo” will emerge later this year combining networks (USA Network, CNBC, MSNBC, Oxygen, E!, SYFY and Golf Channel) and digital assets (Fandango, Rotten Tomatoes, GolfNow and SportsEngine).

Tubi & Fox’s “Existential Challenge”

Big media companies ultimately face an “existential challenge” that invites both “portfolio reconstruction” and the need for “a generational shift in management teams and philosophies". 

Comcast’s SpinCo is a real world example of “portfolio reconstruction”: “If management cannot optimize shareholder value across unrelated assets, then they should either discard those assets or find managers who do understand how to connect them in ways that audiences want.” 

Tubi’s success reflects both. To date, Fox has actively pushed ad-supported Tubi (though Fox CEO Lachlan Murdoch says this will change by year-end). Tubi CEO Anjali Sud is an example of a generational shift in managers. She was previously CEO of Vimeo where consumer “choice and control” was central to the model. 

Since her hiring in 2023, Tubi has captured increasingly younger demographics without pursuing the complexity of pivoting into gaming or AI. It has done so by leaning hard into personalized TV viewing: More than 34% of Tubi viewers are ages 18-34 and more than half are Gen Z or Millennials and nearly half are multicultural. In 2024 it reported 55% growth in Multicultural demos including Latine, African American and LGBT audiences and 63% growth in female audiences, year over year.

Younger Audiences, Younger Entrepreneurs

Tubi’s success suggests that as long as broadcast remains a part of TV consumption—22.4% of monthly TV viewing in the U.S., just behind cable at 23.8%, according to Nielsen’s most recent The Gauge—the most drastic measure of ”portfolio reallocation” may be FAST models over subscription models.

Mohan points to alternative measures lying within Hollywood’s “healthy startup culture”: Creators are “leaning into new models of production, building studios to elevate their production quality, and exploring new creative avenues. They are creating a whole new playing field for entertainment and the businesses behind them.” 

With consumers facing an abundance of choices, Mohan envisions YouTube enabling as many permutations of business models for those choices as possible over the next twenty years. 

Against the backdrop of the three models at the intersection of streaming and gaming, the message to large media companies not named Fox is clear: There may be no single answer to what “portfolio reconstruction” should look like. Disruption is inevitable. Embracing consumer “choice and control” is the only path forward.

“More than 4 million advertisers are now using at least one of our generative AI ad creative tools, up from one million six months ago. There has been significant early adoption of our first video generation tool that we rolled out in October, Image Animation, with hundreds of thousands of advertisers already using it monthly.”

Last December, I predicted that a small business using Meta’s Movie Gen tools will create the most successful AI-generated ad campaign of 2025. Meta has since quadrupled the number of advertisers testing the waters and positioning themselves to create this campaign. However, success is not guaranteed.

As I have written in previous essays, Meta’s scale invites greater scrutiny —Zuckerberg shared that 3.3 billion people use at least one of the apps in its “Family of Apps” daily. “Hundreds of millions” of advertisers could reach 40% of the world’s population by leveraging Meta AI tools to create ads and target delivery.

That said, Zuckerberg’s more provocative revelation was that Meta believes “people don't all want to use the same AI, people want their AI to be personalized to their context, their interests, their personality, their culture, and how they think about the world.” The uncomfortable implication is that an AI-generated ad campaign will not rely uniformly upon the same creative. Storytelling will be in the hands of LLMs and not humans.

This is the type of outcome actor Hank Azaria warned about in a recent New York Times opinion column.

Key Takeaways

 A campaign must tell a compelling story about why consumers should choose a brand. Hank Azaria's concerns about AI replicating his voice acting suggest the most successful AI-generated campaign of 2025 may not emerge on Meta. 

Azaria & Craftmanship

Azaria’s opinion piece wrestles with the physicality of voice acting—”our bodies and souls are involved to get the proper believability”—versus AI’s ability to easily recreate one of his voices. “There’s so much of who I am that goes into creating a voice. How can the computer conjure all that?” 

He adds, “Believability is earned through craftsmanship, with good storytelling and good performances, good cinematography and good directing and a good script and good music.” 

Actor, director and producer Ben Affleck made similar arguments on a CNBC panel last November: “The function of having two actors, or three or four actors, in a room, and the taste to discern and construct that, is something that currently entirely eludes AI's capability, and I think will for a meaningful period of time.” 

Last July, Hollywood voice actors and motion capture artists in video games launched a strike against American video game companies. The rationale—outlined in a Los Angeles Times interview in 2023—was also similar to Azaria’s point. The technology to reuse a likeness or modify a voice has existed for years, but actors argue that “AI ups the ante because it can scrape more information more efficiently and potentially turn it into a plausible clone of an actor, combine actors’ work or pass as a new, ersatz artist.” 

Last September, California governor Gavin Newsom officially signed two bills granting protections against AI being used both on dead performers (granting rights to performers' estates) and on living performers (creating a use case of “reasonable specificity”).

Azaria & Showrunner 

Voice work in video games occupies a middle ground between traditional TV shows and generative AI. In shows like 'The Simpsons,' voice work is delivered at specific moments, while in games, it responds to user interactions - much like Meta's vision for AI-generated content.

Notably missing from Azaria's opinion piece was any mention of his own experience with interactive content - his voice work in the 2007 'Simpsons' video game. This omission is particularly relevant given Fable Simulation's Showrunner platform which blends gaming and streaming models. The platform lets consumers create personalized TV episodes by entering text parameters for world, character, genre, and mood. 

Within Showrunner, Azaria's voice work would be AI-generated but the experience would mirror gaming's interactive nature. There is more ambiguity to Azaria’s argument than he lets on.

Meta v. Storytellers 

I have written before about how consumer demand is evolving away from traditional TV shows, movies and even console games. Consumers seem more sentimental for the intellectual property of media companies more than they are the formats. This shift suggests Azaria's fears about AI voices may miss a larger point: The consumer need for storytelling remains constant, as Saatchi and other AI entrepreneurs recognize. 

The best models put storytelling in the hands of consumers as their demand for "The Simpsons" as a show or game may decline. The question is, who will tell these stories—Humans or AI?

“Text-to-video” platforms put the storytelling in the hands of both. Meta has a more radical  ambition that storytelling can be widgetized and computers can replace humans as creative storytellers to drive advertising sales.

Two key differences separate these approaches. First, Meta commands extraordinary scale and the extraordinary resources to license from anyone ($62 billion in operating income and $45.4 billion in cash flow in 2024). Whereas Showrunner remains in early testing.

Second, Meta does not seem to believe in human storytelling in the same ways that Saatchi or other filmmakers in space still do. A filmmaker told me last December, AI lacks that “storytelling compression layer that we've got”. Meta believes otherwise.

Rethinking 2025

Meta's dominance in funding and distribution means Hank Azaria's AI voice is more likely to deliver sales pitches for retail brands than star in AI-generated episodes of The Simpsons.

Meta's vision of generative AI creating ads tailored to each user's context, interests, personality, culture, and worldview will exponentially multiply the permutations of ads served. Yet without human storytelling, these personalized ads may fail to connect.

This analysis leads me to reconsider my prediction. The most successful AI-generated campaign of 2025 may not emerge from Meta after all. Even with unprecedented reach and resources, a campaign still needs to tell a compelling story about why consumers should choose a brand. 

Meta AI alone will not deliver that.

1. First, Meta’s recent earnings call and the emergence of Deepseek challenge my prediction of “an explosion of new businesses of storytelling”; and, 

2. Second, artificial intelligence will reshape both military-industrial era strategy and its modern consulting-firm iterations.

New evidence since the holidays strengthens the first point, which merits its own essay.

My research into the second point remains preliminary, but I will outline my initial thesis below.

Key Takeaway

Why would management teams that failed to understand DTC technology and business models suddenly build effective AI strategies and operational structures? Even Netflix seems vulnerable now. 

Total words: 1,000

Total time reading: 4 minutes

Last summer, I received a head of strategy role offer from a media company's potential acquirer. I subsequently wrestled with two questions:

1. What defines a strategist's role in an era of rapid change? 

2. What constitutes corporate strategy when AI threatens to eliminate entire company divisions?

To date, my analysis of the media marketplace has focused on the direct-to-consumer (DTC) relationship as a new “first principle” of building a media business. I have frequently cited venture capitalist Marc Andreessen's test for management teams' understanding of DTC business models and their underlying software: Examine the percentage of top 100 executives and managers with computer science degrees.

Last September, I argued that there are two conflicting theories at play in the media industry. Legacy media executives, in pursuing DTC models, have overemphasized Clayton Christensen's disruption theory while neglecting Marshall McLuhan's principle that 'the medium is the message.'" 

Christensen argues that incumbents face a capital allocation choice: Generate long-term growth or adapt to change. McLuhan posits that adaptation demands understanding new mass media distribution technologies and the constraints they impose on both companies and consumers.

The middle ground between the two theories is that successful capital allocation and adaptation requires management to understand the underlying technology. Management teams that fail to comprehend these technologies will inevitably misallocate or suboptimally spend capital. I have long argued that this business logic explains why many media companies have struggled in interactive businesses like streaming and gaming.

In this light, the answer to my first question is simple: Heads of strategy at media companies must be as versed—if not more versed—in new mass media distribution technologies as they are in disruption theory.

However, the answer to my second question is less clear. Why would management teams that failed to understand DTC technology and business models suddenly build effective AI strategies and operational structures? Why would they be able to identify marketplace opportunities and allocate spending wisely?

If they could not adapt quickly (nor slowly) to DTC models, why would they suddenly develop the ability to invest at AI's rapid pace?

The AI Lords of Strategy

Management consulting businesses have emerged to help solve such questions over the past 60 years. However, their early wins focused more on data aggregation and frameworks to help management teams allocate capital more intelligently.

I am reading "The Lords of Strategy: The Secret Intellectual History of the New Corporate World" to understand the early days of management consulting firms because they emerged during a transformative period for industrial and post-industrial corporations (1960s and 1970s). Management teams struggled to comprehend the dizzying speed and scale of the significant market shifts: Government deregulated major industries, new technologies proliferated, capital markets shed past moral and legal constraints, and globalization reshaped and expanded supply chains and customer bases worldwide. 

We can theorize about AI's potential changes to marketplace supply and demand, but ultimately, we do not know precisely what will happen. The implication of Andreessen's test and streaming's precedent is that a select few executives with backgrounds in computer science and language learning models will have a more nuanced understanding of AI's trajectory. They are more likely to identify and build the “right” strategies.

The challenge in the age of AI is that LLMs trained on the capital allocation frameworks—like BCG’s Matrix —can now answer strategic questions, too. Even if imperfect, these responses reveal the core principles and fundamental outlines of established strategic approaches. One must wonder how management teams will balance the cost effectiveness of using AI platforms with the need for bespoke, more imaginative consulting that those AI platforms cannot deliver.

Needless to say, management teams favor cost-effective solutions with higher potential return on investment over bespoke human consulting. The question will be, at what cost to shareholders?

Watch Netflix

I have long argued that Netflix has “won” streaming partly because its management team has backgrounds in computer science. They understood from day one that “the medium is the message” and built technology that maps to the internet’s “invisible groundrules, pervasive structures and overall patterns”. Where legacy media perceived a new channel for content distribution emerging, Netflix management saw a technological means for delighting consumers with personalized entertainment recommendations.  

In last September’s “YouTube Is Forcing Netflix to ‘Break’ Its Business Model”, I argued that Netflix now seems vulnerable because YouTube has redefined “premium content” with the help of creators. It has transformed consumers' investment in content from something emotional into something financial.

Netflix now seems to be evolving more slowly than necessary. Management continues to repeat in its letter to shareholders that it accounts “for less than 10% of TV viewing in every country in which it operates." The data “suggests a long runway for growth as streaming continues to expand around the world.”

The performance of its stock—its price is up 75% year-over-year and 13% since last week’s earnings call—suggests that investors currently see this long runway and do not see those limits. 

The critical question for Netflix management is whether AI can help it compete with YouTube and other "formidable competitors across traditional entertainment and big tech":

• Can AI help it to capture more than 10% of TV viewing in every country?

• Can AI help it to pivot towards consumers’ growing role as producers of the content they watch?

• Can AI help it to adapt to consumer tastes being increasingly shaped by creators and short-form content? 

• Can AI help it to succeed at “a 'once in a generation’ inflection point for game development and player experiences using generative AI”, as Mike Verdu—Netflix’s VP, GenAI for Games—wrote last November?

I think the answer is no to the first three questions, and maybe to the last.

The implication is sobering: Even with the right management team, Netflix's future appears smaller and more uncertain than it should be. A recent pivot towards AI—evident in job listings from last October—seems unlikely to change this trajectory.

I now offer a generous discounted coupon code equivalent to the "I Can Expense It" tier on Substack. Please respond to this email if you are interested.

If you are feeling "on the fence" about subscribing, please read "Was I Right? My 2024 Media Predictions—The Hits & Misses" (free to all). I got most of predictions right or partially right in 2024.

Sign up here to have the clearest market signals for AI and media in your inbox 2x/week in 2025. 

According to a Wall Street Journal article last week, the answer is ambiguous: “Meta has told advertisers that Community Notes will apply to organic posts, not paid ads, but its official messages haven’t gone into detail about brands’ and influencers’ unpromoted posts.”

The problem is organic sponsored posts—where a content creator is paid to showcase a product—have historically made up about 25% to 50% of brands’ marketing campaigns on Meta’s platforms and TikTok. Creators rely on these sponsored posts to fill out their branded-content earnings, and brands use it to complement the posts they do pay to boost.

Meta CEO Mark Zuckerberg previously described the opportunity around generative AI as helping creators produce “more creative stuff" "and “UGC that people create for better, wilder use cases that they want”. There are “only so many hours in the day” and a creator’s community “probably has a nearly unlimited demand to interact with” that creator. AI tools will solve that pain point for creators.

The article suggests that “Meta’s new system might push more advertisers and creators to abandon organic content altogether to avoid the risk of critical Community Notes.”

In turn, Meta’s pivot to Community Notes seems to be creating a disincentive for creators to use Meta’s AI tools to create more content and drive more engagement. The question now is, how will Meta reconcile its ambitious vision for creators and generative AI with its pivot towards “free expression”?

Key Takeaway

Meta's “move fast and break things” culture that now prioritizes a Trump-era work culture shift over creator relationships may be disincentivizing the very creators and small businesses needed to realize Meta's AI ambitions. 

Total words: 1,000

Total time reading: 4 minutes

 “Ham-fisted”

Last week, The Information reported that Meta Platforms Chief Technology Officer Andrew Bosworth described Meta’s handling of parts of its recent changes as “pretty ham-fisted, borderline unintentional”. 

Meta now has nearly 3.3 billion daily users across its platforms and is generating more than $152 billion in annual advertising revenue. It does not seem worried about how its pivot will impact advertising spend.

However, if the objective is “more creative stuff" from creators "that's UGC that people create for better, wilder use cases that they want”, it is hard to see why this pivot helps incentivize creators towards using Meta’s generative AI tools towards that outcome. They need sponsored posts to help them grow.

Add in the uncertainty of TikTok’s future for its 170 million plus users in the U.S. looming in the background, the sense is that Meta’s relationship with creators is on shaky ground. 

Meta’s Spectrum of Creators and Small Businesses

When I spoke with Meta for last November’s “Armed With AI, Small Businesses Also Challenge Hollywood”, I asked about the spectrum of creators and small businesses on Meta. I learned that, in broad brushstrokes, the vast majority of creators are smaller, earlier in their journey. Some creators are small business only, some are creators only and many are creators who also have a small business (like the Kardashian family). However, the number of creators who also have a small business is smaller than the overall group of creators. 

From this picture, we can get a sense of why Meta’s pivot is contrary to the objectives of its generative AI vision. Effectively, the pivot leaves the vast majority of creators who are smaller and earlier in their journey most vulnerable. The AI vision sought to incentivize them to leverage AI to create more content and engage their audiences 24/7. The pivot to Community Notes appears to be a disincentive for organic content creation.

Small businesses and creators who also have a small business seem to be the least vulnerable because they rely on paid ads. But these are both smaller buckets of creators, and they still rely on organic content. This pivot is a disincentive for them to adopt Meta's generative AI tools, too.

Move Fast And Break Things

There were always two big assumptions in Meta’s AI vision: Creators would adopt the AI tools to create more content, and in doing so, they would drive more engagement. Meta's $46 billion investment in AI, AR and VR since 2021 reflects Zuckerberg's commitment to this vision.

Four months after Meta outlined this vision, both assumptions seem to be questionable, and only because of the pivot towards Community Notes. The “move fast and break things” culture that now prioritizes a work culture shift over creator relationships may be disincentivizing the very creators and small businesses needed to realize Meta's AI ambitions. 

The question is not whether to bet against Zuckerberg. Rather it is how Meta will resolve this fundamental tension between its Trump-era culture shift and its bet on AI to drive creator growth. There are no obvious answers right now, and that is cause for concern.

The health issue I mentioned in my previous mailings required hospital care and treatment last week.

I am now on a clearer path to recovery and have enough energy to write once per week. For the next month or so I will be publishing on Thursdays. This will give me time to write and edit without the pressure of deadlines.

While managing my health, I have also been laying the foundation for an important transition: In the coming months, The Medium will be moving off The Information's platform. This move serves two key objectives.

First, I want to offer you more value. Currently, I estimate that only 10% of my interviews with AI builders make it into published essays. Some valuable content never made it to subscribers—like my full conversation with Chris Gilberti, founder of AI data platform Avail, and an insightful tangent about creator business models from my February interview with Samir Chaudry of Colin & Samir.

The Information's platform limits me to essays only. In turn, essays limit me to the evidence and arguments relevant to my core arguments. A lot gets cut during the writing and editing processes.

I believe paid subscribers should have access to complete interviews through video, audio, and transcripts. PARQOR should be more than a library of essays—it should be an evergreen resource across multiple formats. It should also offer broader analyses across these interviews. A different platform will enable me to build that.

Second, some technical issues have recently emerged between Stripe and The Information's third-party hosting platform. The easiest solution is to move off of The Information.

You might wonder: Why not integrate The Medium as one of The Information's newsletters? The Medium's focus has always been on in-depth analysis and frameworks, while The Information excels at delivering timely scoops to its subscribers. I will continue to write my Medium Shift column for The Information, where my analytical approach better complements their breaking news coverage.

I will share more details about this transition in the coming weeks. Please do not hesitate to reach out with any questions.

Thank you for your understanding and patience.

Andrew

I apologize, there will be no essay today because I am not yet recovered from the bug that I mentioned on Monday. I have confirmed it is neither COVID nor flu, but the symptoms are similar to both.

I hope to resume publishing by Monday, though the timing of my recovery is beyond my control. My energy comes and goes in waves, making it harder to focus consistently. Rather than push through, I believe that easiest solution may be for me to publish again when I feel strong enough to do so.

Thank you for your understanding.

Andrew

Disney and Fubo TV, a sports-focused virtual cable network (vMVPD), announced this morning they will combine Hulu + Live TV with Fubo’s operations in a joint venture. As part of the deal, Fubo will drop its antitrust lawsuit against Venu Sports, the sports-focused streaming package from Disney, Fox Corp. and Warner Bros. Discovery.

I wrote a few pieces on Fubo vs. Venu last summer:

• The Pain Point Is Recurring Revenues, The Solution Is...?

• David’s Choice / The Davids’ Choice

• Disney, Fox, Venu Sports, DirecTV & The Doom Loop of The Mogul

In the last piece, I warned of a "lurking danger" in the Venu Sports lawsuit: The management teams at Disney, Fox and Warner Bros. Discovery believe building "walled gardens" is more valuable than decentralizing their business models.

Today's deal weakens Disney's streaming walled garden by handing over Hulu + Live TV to a third party. For context, Hulu + Live TV represented 43% of all Hulu revenues in Q4 2024.

The move is also an example of something I described as "portfolio reconstruction": "if management cannot optimize shareholder value across unrelated assets, then they should either discard those assets or find managers who do understand how to connect them in ways that audiences want."

Here, Disney is handing the management of Hulu + Live TV over to Fubo TV managers who have experience in running a vMVPD business. Disney has not been able to substantially grow the Hulu + Live TV subscriber base since 2020 nor average revenue per user since 2021. 

It is unclear where Fubo TV management are better positioned to do so.

The Disconnect

The deal terms exposes a contradiction: vMVPDs are not strategically important to Disney but it is still embracing Venu Sports. Both are bundled packages of cable channels. The difference is Venu's narrower focus: It bundles virtually all of the U.S. national sports broadcast rights controlled by Disney, Fox and Warner Bros. Discovery in a single subscription, with limited entertainment and news content.

Venu publicly projected at least 5 million subscribers in five years, but evidence revealed in the Fubo TV trial suggested these numbers were conservative. As I wrote last September:

"[A]s imagined, Venu seems to be a desperate bid to preserve and grow recurring revenues at a low churn rate—assuming sports viewing has inelastic demand—by any means necessary. Effectively, a group of TV executives narrowly defined the solution to their expensive rights deals as network television distribution over the Internet. But, as their failures in streaming have proven, the Internet is a different medium that favors other media formats and pricing models."

This raises a reasonable question for Disney management: Does it or does it not still believe in the cable channel as a core value proposition of a streaming business?

The decision to spin out Hulu + Live TV suggests no. But, its deal-making on behalf of Venu Sports suggests it does. 

The Netflix Effect

Last November, I argued Netflix redefined sports distribution with its livestreamed Paul vs. Tyson boxing match. On Christmas Day, it redefined the standards for professional sports distribution with its broadcasts of two NFL games.

Netflix reported "a record-breaking day for Netflix and the NFL, with an unduplicated audience of nearly 65M US viewers, according to Nielsen." It added that viewers from 218 countries and territories tuned in to at least one of the games.

In other words, a substantial number of U.S. viewers did not associate a NFL broadcast with a cable channel but with an app. That seems to be both good news and bad news for Venu Sports.

The good news is that it suggests an app can indeed be a destination for audiences to stream a live broadcast at scale. The bad news is, it is unclear why Venu Sports—especially with its continued focus on cable channels—will be a competitive offering.

I now offer a generous discounted coupon code equivalent to the "I Can Expense It" tier on Substack. Please respond to this email if you are interested.

If you are feeling "on the fence" about subscribing, please read "Was I Right? My 2024 Media Predictions—The Hits & Misses" (free to all). I got most of predictions right or partially right in 2024.

Sign up here to have the clearest market signals for AI and media in your inbox 2x/week in 2025. 

In the meantime, if you have not taken the survey already, please do so over the holiday. I will leave it up and through the New Year. Your feedback is invaluable to me and to what I am building!]

Take the survey here.

Back in February’s “The Media Revolution Will Be Prompted”, I shared a prediction of Samir Chaudry of the YouTube creator duo Colin and Samir:

The best creators will use [artificial intelligence (AI) to make sure human audiences connect with the best storytelling, he said. Those at risk will be those who don’t know how to tell a story.

There are two stories within that prediction. The first is an elegant but Darwinist vision: Human storytelling will still matter but AI will be a necessary condition for connecting with audiences. Those who assume AI will compensate for their lack of storytelling skills will not survive. 

The second story is about connection: AI will not favor any one format (e.g., video, audio, games). Instead, it will only favor which stories and storytellers can connect with audiences.

The traditional media model of aggregating assets across TV, movies, books, theme parks, music and live events no longer connects audiences with the best storytelling. The decline of the cable business means it no longer optimally monetizes them. That model has been proven to be too expansive and uncoordinated to evolve with the emergence of retail-first, consumer-first models over the past two decades, it is irrelevant now.

In 2025, we will no longer be talking about the media businesses of the last century. Instead, we will be discussing an explosion of new businesses of storytelling.

Key Takeaways

1, There will be no AI-generated hits, blockbuster or niche 

2. AI emerges as a “storytelling” performer

3. A messy debate about the “Quality” vs. Possibility of storytelling emerges

4. “The new business of living room content” redefines the role of producer

5. The value of IP libraries for TV series and movies continues to decline 

6. Netflix continues to pivot and evolve away from its legacy media movie and TV DNA

7. A small business using Meta’s Movie Gen tools will create most successful AI-generated ad campaign of 2025

8. Only Spotify and YouTube will have outsized influence in the IP licensing marketplace for AI

And, some predictions for legacy media in 2025

Total words: 2,400

Total time reading: 10 minutes (~1 minute per prediction)

1. There will be no AI-generated hits, blockbuster or niche

Why we will see it

According to TCG founder Peter Chernin, technological disruption means AI-generated storytelling will “aggregate at two extremes: the big blockbuster hits and niche products”. The rest of the stories—what Chernin calls the “bland middle”—will be “gone, and gone forever.”

There is also a more AI technology-focused lens on technological disruption from Slow Ventures partner (and “intern” at The Information) Sam Lessin:

“AI is a new business weapon... who benefits? just like "cloud" before it - the moms and pops get powerful new tools, the $T big platforms profit from arms dealing.  and the "middle" gets crushed.”

Here is a slide from a deck with my business plan for 2025 (and beyond) that sums up these two takes (and if you’d like to walk through the deck with me, please respond to this email):

We are not yet seeing models for niche or blockbuster hits for AI-generated content. Last month, Tom Paton’s AiMation Studios released “Where the Robots Grow”, the first feature-length fully animated movie made entirely with artificial intelligence (AI) tools. It cost only $8,000 per minute, and was made with a team of nine people in 90 days.

Paton shared with me recently that trying to get the movie released “became like a dance with Hollywood where they were making offers but we couldn't get anyone to budge off of a [Q2] 2025 release.” He ended up releasing the movie for free on YouTube, instead.

That dynamic will not change in 2025, resulting in neither niche AI hits nor AI blockbuster movies.

Why I may be wrong

Payment platforms like Stripe already enable micropayments and AI platforms like OpenAI’s ChatGPT and Google’s Gemini and smaller players like Midjourney already charge monthly subscriptions. The conditions are in place for the economics of niche storytelling business models in AI to emerge quickly.

2. AI emerges as a “storytelling” performer

Why we will see it

I wrote in Monday’s “NotebookLM, AI Storytelling vs. AI Performances” that AI will be less a storyteller and more a producer of faster, cheaper “performances” of stories. This question is how and whether consumer demand for this service will scale in 2025. I think it will. 

On a related note, venture capital firm Andreessen Horowitz is predicting “A next generation Pixar” will emerge in the gaming sector in 2025 “using AI for interactive experiences.” Netflix seems to be making one version of this bet with its recent announcement that former Netflix VP of Games Mike Verdu is now VP, GenAI for Games.

He described the opportunity as “a 'once in a generation’ inflection point for game development and player experiences using generative AI.” He will be “focused on a creator-first vision for [artificial intelligence (AI)], one that puts creative talent at the center, with AI being a catalyst and an accelerant.”

Why it may not happen

In the case of NotebookLM, will audiences feel an affinity for two faceless, non-human AI co-hosts debating someone else’s work? That does not read compelling enough on its own.

In the case of Netflix, Fable Simulations founder Edward Saatchi predicted the earliest a Hollywood studio will “take the plunge” on interactive generative AI platforms will be in mid-2026. 

3. A messy debate about the “Quality” vs. Possibility of storytelling emerges

Why we will see it

When I interviewed creator Don Allen Stevenson III for last month’s "Armed With AI, Small Businesses Also Challenge Hollywood", he said something interesting about “creative possibility” in storytelling with AI: “some of the best stuff that I throw AI and like generative AI tools at are the stories that I wouldn't have gotten a budget for” when working at a Hollywood studio like Dreamworks. The ideas that studios “forced” visual effects designers to do were the ones that “the studios had confidence that it would make X amount of money.”

Tools like OpenAI’s Sora, Google’s Veo2 and RunwayML all present an uneasy tension between the more TV and movie-centric concepts of quality, and the promise of “creative possibility” from generative AI. Producer and actor Ben Affleck made the case against it at a CNBC conference last month:

“AI can write you excellent imitative verse that sounds Elizabethan. It cannot write you Shakespeare. The function of having two actors, or three or four actors, in a room, and the taste to discern and construct that, is something that currently entirely eludes AI's capability, and I think will for a meaningful period of time.”

Despite Affleck’s references to Shakespeare and his own movie “Good Will Hunting”, there is no universal standard of “quality” in Hollywood. According to Forbes, low-budget television animation sourced from overseas studios costs between $10,000 and $20,000 a minute and “isn’t anywhere near the quality of ‘Where the Robots Grow’”. 

Then there is a third variable: Newfound efficiencies and possibilities in production models from AI will matter less if consumers increasingly prefer their favorite IP in different formats than TV or movies.

So, anyone looking for a clear answer to how generative AI will impact Hollywood storytelling is going to find themselves trying to reconcile these three moving pieces.

Why I may be wrong

We will likely see some studios figure out how to integrate AI into their storytelling (and Disney recently integrated AI into the opening credits of its Disney+ series “Secret Invasion”). However, it will be inevitable that traditional studios will be disrupted by AI storytelling in 2025.  

4. “The new business of living room content” redefines the role of producer 

Why we will see it

Consumers are increasingly becoming the “producers” of their favorite stories. YouTube now dominates streaming on household televisions across the U.S. (according to Nielsen's The Gauge) and around the world. I wrote in “Netflix, YouTube & The New Business of Living Room Content”:

“The more YouTube competes with Netflix on television screens worldwide, the more YouTube is proving that consumers' ‘choice and control’ over the medium of the internet means fans are the producers of the content they want to watch. “

ReelShort, a short-form vertical video platform, is generating $40 million per month from a female-heavy audience—half of which is in the United States—with “soapy romance mini-dramas with one-minute episodes and up to 80 episodes in a “series.” Its app currently has 50 million monthly active users who make micro-payments by buying “coins” to watch content. 

Both reflect how consumers are increasingly playing direct roles in determining which content is produced and the “quality” of content they want to see. In 2025, that status will be further cemented in the marketplace with more consumer-as-producer models emerging.   

Why I may be wrong

I won't be.

5. The value of IP libraries for TV series and movies continues to decline 

Why we will see it

I previously argued in “Back To The Future Of Pre-Paramount Decrees Studio Distribution” that the next generation of consumers “value the user experience (UX) as much as the content, enough to prefer free services to paid services with historically popular IP.” 

In other words, the growing popularity of social video platforms and FASTs seems to have catalyzed both a decline in the value of IP libraries to consumers and in the business of monetizing them. Now, free services are growing a lot faster than paid ones.

If we follow Marshall McLuhan’s path of thinking, the internet is a more interactive medium than the more passive viewing mediums of television or theatrical. That means consumer standards of "quality" for TV and movie content in those mediums are also more passive, and their standards of "quality" within the medium of the internet are more interactive. 

It offers an abundance of choice. The value proposition for consumers is not just “possibility” but the ability to find interactive experiences with storytelling that map more narrowly to their needs.“Quality” storytelling increasingly reflects those narrow needs more than any previous standard established by TV and theatrical mediums.

Consequently, IP will continue to become less valuable for TV series and movies. It will become more valuable for more interactive offerings.

Why it may not happen

It will. This is an irreversible trend.

6. Netflix continues to pivot and evolve away from its legacy media movie and TV DNA

Why we will see it

In September, Co-CEO Ted Sarandos told the Royal Television Society’s (RTS) London Convention 2024 that Netflix must cannibalize its own business and “constantly challenge ourselves, to break [the business] and move our business forward on behalf of our consumers.” 

Netflix is leveraging mobile gaming to continue storylines between seasons of its reality shows. It is distributing live sports events globally (Jake Paul vs. Mike Tyson, NFL Christmas Day games). It also will be making a reality show called “Inside” with British YouTube superstars The Sidemen. It is evolving beyond TV and movies.

It is also moving beyond Hollywood. Back in March, research firm Ampere Analysis projected more than half of Netflix’s content spending this year — $7.9 billion of about $15.4 billion total — would go toward titles produced outside of North America. So, 2024 will likely mark the first time the majority of the streamer’s budget will be spent on either local original content for international markets or the licensing of internationally produced titles. 

A future of faster and cheaper TV shows, movies, some sports events and mobiles games seems counterintuitive for a platform that had Disney-like ambitions only three years ago. 

However, this seems to the result of its competition for consumer attention and advertiser dollars from Silicon Valley—YouTube and Amazon Prime Video especially—is forcing it to unlock new competitive advantages.

Why I may be wrong

According to its latest What We Watched report, people watched over 94 billion hours on Netflix in the first half of 2024. It is growing towards 300 million global subscribers and has churn rates below 2%, according to research firm Antenna. This may be a "molehill" of a “problem” in 2025 before it becomes a "mountain" of a problem in the longer run.

7. A small business using Meta’s Movie Gen tools will create most successful AI-generated ad campaign of 2025

Why we will see it

Back in September and early October, I argued Meta seems to be betting that "small businesses are becoming creators faster than creators are becoming small businesses."

The pace at which AI models are evolving—Google announced Veo 2 less than two weeks after it revealed Veo and OpenAI released its o1 version of its video generation model Sora only 10 months after its release—suggests that Meta’s Movie Gen will be publicly released in mid-2025 to “hundreds of millions” of small business advertisers. 

Meta’s advertising platform and “family of apps” are instant distribution outlets for its AI-generated content to over 3.27 billion daily active people. That is a significant competitive advantage. Odds are at least one small business figures out a viral advertising “hit” on the platform.

Why I may be wrong

We still do not know if AI-generated ads generate the type of brand awareness or conversion behavior when compared to other rich media format ads. It may take as long as a decade before we do. That is as long as it took for podcasts and the creator economy on platforms like Spotify, YouTube, Instagram and TikTok to prove itself out.

8. Only Spotify and YouTube will have outsized influence in the IP licensing marketplace for AI

I wrote in November that large media companies need more opportunities to monetize their large libraries of IP than within the closed marketplaces of their walled gardens. The logical solutions are either to aim to build free marketplaces where the IP can be more accurately valued or to let third-party technologies rethink their walled gardens for them. 

I think the latter will happen simply because platforms like Spotify, TikTok and YouTube (which I argued is “AI’s frenemy”) already leverage algorithmic technology to ensure third parties' IP rights are not violated by AI or creators. They are friendly to AI-generated content (as long as that content is labeled as such). There will be no market need for a new market entrant to technologically or legally reinvent the proverbial wheel here. 

Why I may be wrong

There may not be a solution. Intellectual property licensing is especially complicated in film and TV because of the myriad of rights holders (cast, crew, production, investors, and more). AI plays in the grey areas of IP licensing because it does not create “new” IP nor is it clear how it uses IP to train its models (this is the basis of ongoing lawsuits against Meta and OpenAI). 

Legacy Media Predictions

• Three years closer to 2029 means we’ll see NFL starting testing the waters for a new deal

• Disney and Epic hit the same operational and cultural roadblocks within The Walt Disney Company that past gaming and creator economy ventures have hit, and they dial back their ambitions. 

• A smaller streamer drops its owned-and-operated subscription streaming model and licenses its content library entirely to Amazon. 

• Neither Skydance nor Apple pursue the “cloud-usage model” I proposed in a September essay. However, Hollywood’s need for a free marketplace for content buying will push the market to find a solution, so we cannot rule it out?

Last Friday I decided to run an experiment with last Thursday's essay (“Was I Right?...”): I uploaded it to NotebookLM to create a podcast. Within minutes, a 19-minute “podcast” of my essay emerged with an anonymous, AI-generated male-female duo enthusiastically discussing key points from my essay (and you can listen to it on my old PARQOR Substack).

A reader texted me that it sounded just like one or more of the Penske Media-owned podcasts, and better than the ones he had listened to. Those podcasts have human hosts, production budgets and a loose script.

This was an AI-generated, “tightly”-produced “podcast based on an analytical essay, structured according to an ordered sequence of subheadings, and scripted as a debate around what I wrote. Most importantly, it was free.

This podcast production challenges both my essay format and any potential podcast I could create. The irony is not lost on me as someone who writes about all the new and exciting ways AI is challenging the movie and TV formats. I have written about the future of text-to-prompt video but this is essay-to-podcast audio. It is a different "animal".

The whole experience reminded me of something an AI filmmaker said in passing during our conversation the previous week: “a script is a 100-page analog prompt that we give to 300 people and say go and make this film.” NotebookLM took my essay as an "analog prompt" for producing a podcast, even if the essay was not intended to be a script.

That reads and feels different than anything I have discussed with entrepreneurs in the Generative AI space, to date.

Key Takeaway

NotebookLM suggests AI will be less an original storyteller and more a producer of faster, cheaper “performances”. But, a storytelling "performance" is still storytelling, and therein lies the rub.

Total words: 1,000

Total time reading: 4 minutes

The Analog Prompt

It reminds me of something Producer and Actor Ben Affleck said last month:

“AI can write you excellent imitative verse that sounds Elizabethan. It cannot write you Shakespeare. The function of having two actors, or three or four actors, in a room, and the taste to discern and construct that, is something that currently entirely eludes AI's capability, and I think will for a meaningful period of time.”

In other words, AI is bad at storytelling. The AI filmmaker told me this is because AI "lacks that storytelling compression layer that we've got.”

Samir Chaudry of YouTube creator duo Colin and Samir told me something similar in an interview earlier this year: “it is in our DNA to tell stories and to hear stories told” by human storytellers. 

NotebookLM seems to reveal a blind spot in these arguments. The podcast generated by NotebookLM may be a better “storyteller” about my analysis than the essay in which I delivered it, or a podcast I could record by myself.

This podcast seemed to have “worked” because my essay had a basic structure to it:

• Subheader for each prediction

• Emoji symbols reflecting my evaluation of how I did

• The discussion 

The “script” of the podcast follows that outline in order. It used my website URL to identify where I worked (and it missed the nuance that I do not work at The Information). There are no prompts in the script that tells the AI to make the voices enthusiastic or doubting in different areas.

Nevertheless, it was able to generate a compelling audio “story” out of my essay that sounded like any generic AI-generated male-female podcast hosting duo. In other words, it was AI-generated storytelling with human voices about a story I had told in essay format.

An Important Nuance

There is an important nuance here. Chaudry’s concern about flooding was that AI could generate its own compelling video stories from consumer text prompts. That is certainly still possible.

Here NotebookLM is generating compelling audio from an essay that has been outlined, written, edited and published. Both the input and the output have quality and utility to my readers and target audiences.

It is important to remember that on YouTube and Spotify, audio content competes with video content. According to Edison Research, YouTube has risen to the top as the most popular service used for podcast listening in the U.S.: 31% of weekly podcast listeners ages 13 and up choose YouTube as the service they use most to listen to podcasts, surpassing Spotify (27%) and Apple Podcasts (15%).

In the meantime, Spotify recently announced uninterrupted, ad-free video podcasts. It shared that nearly two-thirds of podcast listeners say they prefer podcasts with video. 

In these datapoints you can see why Chaudry’s read on storytelling is partially right and partially wrong. It is right in the sense that it reflects how consumers still need and value human storytelling. NotebookLM may be producing video within the next few years but it may not matter: The challenges that OpenAI’s Sora has with rendering gymnastics suggests AI is not built for certain types of video storytelling. 

But, it is wrong in the sense that consumer attention—the most valuable commodity in the Attention Economy—can still be captured by an AI-generated “performance” of an essay.

The Tension

There are no clear answers here.

The simplest takeaway from the above is that NotebookLM suggests AI will be less a storyteller and more a producer of faster, cheaper “performances”. It delivered the equivalent of a produced performance of my "story", and not an original story. Given that AI cannot produce compelling video of an essay analysis yet—if it ever will—the sense is that the “performances” AI will deliver may only be as good as the content humans feed it. 

NotebookLM suggests AI will be less an original storyteller and more a producer of faster, cheaper “performances”. But, a storytelling "performance" is still storytelling, and therein lies the rub.

The platform has uncovered a grey area in "storytelling" that builders and creators had not seen before. The sense is that the debate around whether AI can replace human storytellers needs to rethought immediately. Because "performances" are storytelling, too: AI may not be able to write Shakespeare, but it will be able to perform it.

Author's Note #1: A reminder that I am running a survey about the seven entrepreneurs or "builders" I interviewed this year.

Please take a couple of minutes to let me know what else would you want to learn about them in a new offering. And in which other formats?

It is invaluable to me to get your feedback. Thank you!

Take the survey here

Author's Note #2: Some quick housekeeping for this month:

1. Back in January, I made 10 predictions for 2024. With the holidays fast approaching, I am going to revisit them on Thursday

2. Next week, I will make my 10 predictions for 2025. This essay will be my last for 2024 barring any significant headlines or developments.

3. Given these two projects, there is no full essay today. Below, exclusive to subscribers are two short takes on Omnicom and Interpublic merger and recent headlines for Disney and Warner Bros. Discovery.

Total words: 700

Total time reading: 3 minutes

Observation #1: Omnicom + Interpublic

Omnicom Group announced this morning it will acquire Interpublic Group IPG. The deal would create the world’s largest advertising company.

The Wall Street Journal reports that Interpublic is “struggling to keep pace” with its rival Publicis, which “adapted faster to the technological shifts that reshaped how brands connect with consumers.”

Large holding companies more generally have been “struggling to develop new lines of revenue” and their best-known products—TV commercials and print ads—are seen as less effective in spurring consumer purchases and response. They are “working to stave off competition from tech companies such as Alphabet’s Google and Meta Platforms that are using AI to drive deeper into the business.”

According to the announcement, the purpose of the merger is "to accelerate innovation and harness the significant opportunities created by new technologies in this era of exponential change." 

This reads like an understatement if you have been reading my essays on Meta CEO Mark Zuckerberg’s vision of “hundreds of millions” of small businesses becoming creators. Their competition for consumer attention is exploding exponentially as creators become small businesses.

There is still consumer demand for large brands: Interpublic's client roster includes L’Oréal, Johnson & Johnson and Geico, and Omnicon's includes Disney, AT&T and PepsiCo. 

Broadly speaking, there are three challenges to this merge presented by creators as small businesses:

1. Creators compete for and capture attention that blue chip brands badly need;

2. Creators with small businesses have more resources to compete for the same inventory to reach those same consumers; and

3. Creators with small businesses have full-funnel models—meaning, they leverage advertising to drive awareness, purchase consideration, and purchase decisions—whereas blue chip brands rely more heavily on awareness.

This reads like an updated, more globalized version of the market dynamic the Interactive Advertising Bureau (IAB) has described as “The 200  vs. the 10 million”. I wrote about this dynamic in October's "Meta, Small Business Creators & “AI Television Universes”.

In the past, the television marketplace offered a fixed supply of linear inventory to a fixed demand of buyers. 200 “retail-cartel” advertisers supplied 88% of U.S. network television revenue (the term “retail-cartel” applies to the brick-and-mortar retailers who have historically bought from networks). 

But, Meta, TikTok and Google also have been proving over the previous six or seven years that over 10 million advertisers in the U.S. are willing and eager to spend on direct-to-consumer advertising. Those platforms have been democratizing the competitiveness of smaller brands in the advertising marketplace and pushing both advertiser supply and demand away from television and other traditional formats. 

In other words, this deal is a bullish signal for Meta.

Observation #2: Disney & Warner Bros. Discovery Go Tactical

Last Wednesday, Disney debuted a dedicated ESPN tile on Disney+ for people who subscribe to ESPN+. This morning, Warner Bros. Discovery announced it had finalized a wide-ranging cable distribution deal with Comcast, giving Comcast expanded rights in how it sells the Warner streaming service Max to its customers in the U.S. Also, Sky will become a distributor of Warner’s Max streaming service starting in early 2026.

For both streaming ventures, these are tactical moves with marginal value mostly targeted to lowering churn. Disney needs Disney+ consumers to watch ESPN and it is easier to consolidate those views on a single platform. The bet is that they are less likely to churn out if they do.

Warner Bros. Discovery needs to lower churn and Comcast has the rights to package the ad-supported versions of Max and Discovery+ in its streaming bundles. 

Incremental progress is progress. These businesses are starting to run more efficiently after five years of inefficiency.

However, it is worth remembering what I wrote last month on Comcast and Disney:

"[E]fficiency as an outcome will always be insufficient. Meta, Netflix and generative AI are redefining "efficiency" for these [media companies] faster than they can redefine it for themselves."

Both companies' big bets on streaming's future seem to be destined for disappointing outcomes. Disney and Warner Bros. Discovery should be pivoting more aggressively towards a licensing model for their libraries of intellectual property (IP). The pace of change being driven by artificial intelligence is extraordinary and is creating more opportunities for monetizing IP.

The rationale for the Omnicom-Interpublic merger is "to accelerate innovation and harness the significant opportunities created by new technologies in this era of exponential change." Neither of these deals accomplish this outcome for either Disney or Warner Bros. Discovery.

Author's Note #2: A reminder that I am running a survey about the seven entrepreneurs or "builders" I interviewed this year, all of whom sit at the intersection of Artificial Intelligence (AI) and media. 

What else would you want to learn about them in a new offering? And in which other formats?

Take the survey here

There have been two conflicting themes in my essays over the past five years: The exciting possibilities of technology and the “hard knocks” reality of constraints (market, legal, technological).

When I launched this subscription newsletter back in 2019, legacy media companies were boldly promising investors that their streaming services could reach hundreds of millions of subscribers worldwide. 

I was skeptical. 

I had learned as an executive within Viacom (now Paramount Global and soon to be “TBD”) that the older generations of TV and movie executives did not understand the emerging technology or retail business models of direct-to-consumer. The extraordinary profits from the cable business model—which effectively subsidized their media businesses—incentivized everyone to invest in their digital futures suboptimally.

Over the past five years, these executives have learned the “hard knocks” consequences of suboptimal investments in technology and direct-to-consumer business models. High churn rates suggest the demand for their content libraries is weaker than their business models require. Most are struggling to scale and/or to be profitable. 

Even Netflix faces a tougher march to 500 million subscribers. All are losing consumer attention and engagement to YouTube on TVs across the U.S..

Today, the seeds of a new five-year dance between possibility and constraints lie in emerging generative artificial intelligence (AI) tools. Over the past year, I have written about the possibilities of generative AI through the eyes of entrepreneurial “builders” like Jordi van den Bussche  aka "Kwebbelkop", Fable Simulation’s Edward Saatchi and Don Allen Stevenson III aka "Don Allen III". 

Generative AI technology is in a primitive phase where its use cases are being tested and proven out. So, there seem to be exponential possibilities and very few constraints. The question is how to think about what the constraints might be. That is what makes Marshall McLuhan’s work arguably more valuable than ever.

Key Takeaway

"Builders” like Jordi van den Bussche (Kwebbelkop), Fable Simulation’s Edward Saatchi and Don Allen Stevenson III are pushing past the traditional constraints of traditional media to identify new dynamics of storytelling within the medium of generative AI.

Total words: 1,400

Total time reading: 6 minutes

Three Problems With McLuhan

There are three fundamental problems with recommending Marshall McLuhan in 2024.

1. He was a philosopher;

2. Who was writing for an audience of academics (in “Understanding Media” he frequently cites the unreadable modern novel “Finnegan’s Wake” by James Joyce); and,

3. Unlike Clayton Christensen, he never utilized business case studies to support his arguments.

In short, his writing and argumentation can be dense and opaque. Anyone citing him beyond his adage “the medium is the message” may not necessarily understand what he wrote (and that was the gist of the joke behind McLuhan’s famous cameo in Woody Allen’s “Annie Hall”). 

A more accessible version of his writing can be found in “The Medium is the Massage: An Inventory of Effects”, a 160-page book published in 1966. It is composed in an experimental, collage style by graphic designer Quentin Fiore (and with a recently redesigned cover by the artist Shepard Fairey). 

That book offers a simpler, more digestible outline of his core argument: Mediums—or processes—matter more than the core content delivered within them. It argues that the key to understanding the possibility of generative AI business models is to identify the “invisible ‘groundrules [sic], pervasive structures and overall patterns” that work in the background of mediums.

Whereas, a focus on the various, emerging content formats with generative AI distracts us from identifying those constraints. McLuhan writes “the ‘content’ of a medium is like the juicy piece of meat carried by the burglar to distract the watchdog of the mind.” 

In 1967, television was “a totally new technology which demand[ed] different sensory responses” from audiences. In that framing, generative AI in 2024 seems to be television’s logical replacement in that sentence.

However, unlike television—constrained by radio waves (broadcast), fiber optic cable installation, (cable) and television manufacturer capabilities—the constraints and processing of generative AI are less familiar to us in large part because they are still emerging.

Possibility

This is why business models like Kwebbelkop AI, Showrunner or Stevenson’s consulting model seem more interesting than business models that leverage AI tools to make Hollywood production models faster and cheaper. They actively seek those constraints and then iterate accordingly.

Whereas van den Bussche, Saatchi and Stevenson seem focused on the possibilities for storytelling created by generative AI beyond movies and TV series. None are betting on using an arsenal of AI tools to produce the volume and the quality of content that previously was only expected of much larger studios.”movies or TV series as the final output to emerge from generative AI. 

Promise Advanced Imagination, an AI startup that emerged last week, is focused on the more familiar outputs of movies and TV series. It is focused on “using an arsenal of AI tools to produce the volume and the quality of content that previously was only expected of much larger studios.". There are possibilities in that model, but they are more technical, tactical and narrowly focused on building new workflows for studios “from the ground up”. 

Van den Bussche is hacking his way into building a suite of AI tools that generate every aspect of a YouTube video with his “Kwebbelkop’ persona: “audio, video, person, voice, concept, everything” as he told me in a recent conversation. He started down this path after suffering what he called “burnout” from posting videos almost every day for 10 years.

The interactive storytelling on Showrunner starts with the idea of the consumer—via text-to-video prompts—and then leverages generative AI to create an episode of a show. However, its founder Saatcxhi is not focused only on TV episodes as the final output. His company Fable Simulation recently shared research on SIM-1, “a framework for AI wargames that powered an AI simulation in Sim Francisco”. Fable recently released a simulation in Sim Francisco of the 5 days from when Sam Altman left OpenAI to when he returned. It also released a research paper on Showrunner Agents in Multi-Agent Simulations.

As for Stevenson, he told me "In the past, I would always kind of tailor my ideas off of how much time do I think I have to execute on that idea? If I started to get too creative with the idea, I would just pull it back and say, ‘You know what? Let's not go down that route because we won't have time if that's the wrong route.’”.

Now, with generative AI tools, he can explore “super complicated” concepts without punishment:

“And if it's the wrong direction an hour from now, we'll change directions. And that creative workflow, that creative pipeline is now one that's far more forgiving to the individual. You can explore without punishment. In the past, if you went to explore, there's a major risk. If you're wrong on your exploration, you've just wasted so many resources in trying to get to that idea that was wrong.”

He produces short Instagram videos, and as TV shows or movies, as the final outputs, and monetizes them with other business models: Instagram subscriptions, exclusive content for “luxury audiences”, and consulting for technology companies). He is using traditional media to bypass the traditional economics of content production.  

After TV, Movies, YouTube, Instagram…

All three entrepreneurs are pushing past the traditional constraints of traditional media. The question is how to identify the obstacles to all this possibility. Only van den Bussche and Stevenson are actively monetizing content produced with generative AI (van den Bussche via ads and sponsorships). 

Through the lens of McLuhan, the constraint that seems easiest to identify is consumer demand. McLuhan wrote that in 1967, what television’s audiences understood and television’s critics failed to understand is that “a totally new technology [...] demands different sensorial experiences. Van den Bussche’s fans rejected his previous attempt at an AI Kwebbelkop last Fall, and he suspended the earlier version of the project last December. He told me he will be announcing his next experiment in AI shortly.

Stevenson is more interesting because he has found audience demand but the majority are not paying to watch his content. Instead, they are paying to watch “luxury” exclusive content or for videos to inspire developers, as customers like Apple, Adobe, Snap and Nvidia do.

To paraphrase a point from McLuhan in “Understanding Media”, generative AI technology is “within the gates”, and “we are numb, deaf, blind, and mute” to its encounters with both the demand side and supply side of the media business. The sense is that by exploring the possibilities of storytelling, these entrepreneurs are discovering new dynamics of storytelling made possible by the medium of generative AI.

The danger with possibility is that it implies there are no constraints. It is in following “builders” more closely that we get early, valuable signals for what those new constraints around storytelling are and will be in the age of generative AI.

Take the survey here

My latest Medium Shift column for The Information went live last Tuesday. In “Armed With AI, Small Businesses Also Challenge Hollywood”, I wrote about Don Allen Stevenson III who is both a creator publishing content using AI tools like Meta's Movie Gen and also the owner of a small business.

Don Allen has built his own hybrid model for the AI era.

My favorite “a-ha” moment from the interview with Stevenson happened when he told me that 80% of his videos will be made with AI and require little or no additional labor. Social media audiences will watch, usually for free (he has paying Instagram subscribers). The followers, likes and views of the 80%—aka “vanity metrics”—will help attract the “luxury audience” who will pay to consume the other 20% of his content. 

Normally we think of the 80/20 rule—or, the Pareto Principle—as 80% of outcomes come from 20% of causes. Stevenson’s business model flips that logic on its head: 80% of his work helps him monetize 20% of his work better than he could do with advertising or sponsorships alone. 

This math is bread-and-butter business logic for marketers and retailers: Stevenson generates and captures attention with AI-generated content, first and frequently. He then drives those audiences down various buyer’s journeys—or conversion funnels—where he converts a smaller percentage of those audiences into paying customers with other content and services.

As Stevenson told me, “Three years ago I would have assumed that if you wanted to make stuff that looked like cinema-quality, feature-ready you would need a large studio. That’s not how I feel anymore.” In other words, visual effects (VFX) professionals no longer need to rely on the TV or theatrical mediums, only.

As consumer demand evolves away from traditional TV shows, movies and even console games, creators like Stephenson and platforms like Meta seem best-positioned to capture and shape that new demand.

Key Takeaway

There may be at least 100,000 Don Stephensons out there with direct access to audiences at scale who will not have a business need to monetize all or almost all impressions in order to break even. The implications seem terrifying for Hollywood and traditional media.

Total words: 1,100

Total time reading: 4 minutes

The Rise & Fall of VFX

Visual effects has long been "a global patchwork of studios, effects houses, boutique shops, and freelance contractors", as Defector’s Drew Magary wrote two years ago. VFX departments did not exist—particularly in television—until a decade ago. Three years ago, the post-production process became oversaturated with demand, creating a crisis that slowed TV series, movies and video games getting to market.

Stephenson highlighted how resource constraints—time and money being the main ones— drove the post-production market's struggles to meet that demand: 

[I]f I can generate something that a studio would take a month to generate in 15 seconds, then that has amplified my ability tremendously. I mean I couldn't hire that many people at DreamWorks to do that work for that one-off concept so you get time back.

In other words, a fear of waste factored into post-production, perhaps as much as if not more than creativity did. Stephenson explained how time and creative constraints go hand-in-hand: “

"In the past, I would always kind of tailor my ideas off of how much time do I think I have to execute on that idea? If I started to get too creative with the idea, I would just pull it back and say, ‘You know what? Let's not go down that route because we won't have time if that's the wrong route.’”

In 2022, Magary explored how those constraints impacted the demand-saturated post-production market. As one anonymous source told him:

“There are so many differing views and egos involved that it wasn’t uncommon to be on the hundredth version of a CGI animal or big action shot. Either might be on screen for only a split second. The amount of notes and revisions would inevitably push our delivery dates, which would always be an annoying conversation with the client. They usually wouldn’t understand the time that even rendering, let alone creative, takes.”

Magary highlights a web of relationships and responsibilities across Hollywood production studios, on-set creative talent, post-production shops and streaming platforms. It was a mess, and it still is. 

“I can innovate faster on my own”

Listening to Stephenson, the obvious suggestion is that the post-production market for VFX will be advanced by AI tools like Meta's Movie Gen. Both Hollywood studios and gaming studios will be less resource-constrained by the post-production process. In theory, that should mean productions will be more cost-efficient and time-efficient, resulting in more profitable theatrical runs and TV series. 

But, it will also mean the downfall of VFX shops. Last week, actor and producer Ben Affleck described them as being “in trouble” because “maybe it shouldn't take a thousand people to render something but [AI] is not going to replace human beings making film.”

Stephenson sounded more cynical after his experience at Dreamworks during the pandemic. He shared the story of how one of the first teams let go by Dreamworks during that period was the future technologies team. He shared his reaction then: “I was so blown away that, when things got rough at the studio, they threw away their future technologies team.” It was then that he realized “that's the wrong place for me to be at. I can innovate faster on my own.”

The implication is that the same people who were quick to throw out the next-generation technologies are also still in charge of the studios. The more things change, the more things stay the same. Perhaps AI will transform post-production studio processes to become faster and cheaper, but it is unlikely to happen at the speed at which creators like Stephenson are moving.

The Crossroads

If we take Stephenson's model, scale it up across “hundreds of millions” of creators and small businesses, and then assume the top 0.1% to 1% will succeed, an important point emerges: There are at least 100,000 Don Stephensons out there with direct access to audiences at scale who will not have a business need to monetize all or almost all impressions in order to break even.

Moreover, Stephenson’s model is not utilizing this newfound, unburdened creativity to create movies like “Where the Robots Grow”, the first feature-length fully animated movie made entirely with AI tools. Instead, he is exploring faster and cheaper creative outputs made possible by AI. 

He does not need to produce movies or TV series because, on Instagram, audiences prefer short-form content to long-form movies or TV series (which are not available). There is the lurking sense that his creativity—and the creativity of hundreds of thousands of other creators—will ultimately produce a different kind of output from short-form content, movies or TV series. 

It is unclear yet what that will be. The uncertainty seems terrifying for Hollywood and traditional media. What will be the long-term of all this innovation in the hands of hundreds of thousands of creators who will not need the post-production marketplace for income? 

At the same time, this uncertainty is exciting and promising for savvy creators like Stephenson, too. The only certainty seems to be that Meta is increasingly positioning itself to be the heart of all this change.

Affleck predicted that consumers will want “to make 5 minute, 30 second Tik Tok videos where they look like “The Avengers”… Just like you used to be able to buy your Iron Man costume at the store, you're going to buy your Iron Man pack and you and your buddies are going to look like Iron Man and Hawkeye on Twitch.” Done right, that model will create “an additional revenue stream that can replace [the] DVD which took 15 to 20% out of the economy of film making.”

This example touches upon an argument in last month’s “Is There Consumer Demand For Hollywood Content Produced With Generative AI?: Newfound efficiencies in production models from AI will matter less if consumers increasingly prefer their favorite IP in different formats than TV or movies.

It is also an example of the four “back to the people” business models where consumers might license “beloved worlds and characters” to create original content.  In other words, Affleck is aware that consumers are increasingly becoming producers. He also conceded that “YouTube is kicking peoples’ ass” and therefore talent “have to work harder and be better.” 

Affleck's vision ultimately mirrors the “new business of living room content” I wrote about in September: “The more YouTube competes with Netflix on television screens worldwide, the more YouTube is proving that consumers' 'choice and control' over the medium of the internet means fans are the producers of the content they want to watch.” 

Affleck’s interview avoided answering what this new competitive dynamic will mean for producers.

Key Takeaway

The business of being a “producer” is being more broadly redefined. Even if Ben Affleck's Artist's Equity can successfully reduce upfront costs, it may not be enough to compete against an increasingly democratized pool of producers.

Total words: 1,000

Total time reading: 4 minutes

What Should Supply Look Like?

Lurking within Affleck’s predictions and insights are the early seeds of realization that we are living through the next chapter of Marshall McLuhan’s “the medium is the message”.

The business of distributing interactive content to billions of devices (mostly smartphones) over the internet is not the same as the linear distribution business model. Interactive content is not the same as static light signals sent via miles and miles and miles of fiber optic cable to a finite number of homes and businesses (over 100 million homes in the early 2010s and now 76 million). Nor is it the same as convincing hundreds of people sitting in a theater to simultaneously experience the same movie on a giant screen for a few hours.

McLuhan’s point was that mediums matter more than the content. The “invisible ‘groundrules [sic], pervasive structures and overall patterns” working in the background of the internet are doing more to involve humans in both the consumption and production of content than the more passive mediums of TV and theatrical distribution ever did. But, we miss that because we focus on the content instead of how the medium shapes human behavior.

Consumers’ relationships with the content and its creators will no longer be passive in an increasingly interactive medium, especially as they assume the role of producers subsidizing creator content at a smaller scale. For these reasons, demand is evolving, so supply must evolve, too. 

TV shows and movies are less bankable products and the internet as a medium favors multiple content models. Affleck concedes this. But, he is less clear on what supply should look like as consumer demand evolves away from TV shows and movies.

The Signals

The vision presented by Affleck and Cardinale at the Delivering Alpha conference has the qualities of an incomplete jigsaw puzzle. There are certain signals that they have identified—the emerging cost efficiencies from AI, the “cultural footprint” of some but not all talent and the growing value of IP beyond TV and movies—that point to changing how the supply of content is likely to evolve away from movies and TV shows. 

For the remaining TV shows and movies, his bet with Artist’s Equity—with the backing of Redbird Capital, whose founder, managing partner, and chief investment officer Gerry Cardinale appeared on-stage with him—is to make sure creative talent “act like owners” in the creative product and are not just participating for the “cash grab”. 

In short, the talent becomes the producers. That means realigning project incentives away from studio executive compensation packages so that bonuses are awarded in “escalating tranches of money”. “Real success” in a streaming project with the likes of Netflix will mean talent will earn “a hell of a lot more.”

However, the big picture seems otherwise incomplete. The problem with YouTube ”kicking peoples’ ass” is that it implies YouTube is simultaneously redefining success for a Hollywood “production". It is doing so by democratizing the pool of producers to the subscriber bases of a creator and enabling them to fund creators directly via subscriptions, merchandise and tips (all enabled by YouTube’s Partner Program).

I made a version of this point in September's “YouTube Is Forcing Netflix to "Break" Its Business Model”: “The more YouTube pressures Netflix into producing more content at a lower cost, the more Netflix will have “to break [the business]” and undermine the studio side of its model.”

Even if Artist’s Equity figures out how to produce lower-cost content that will succeed on the back end—and it is a reasonable bet to assume they will—market forces are working to push those costs even lower while expanding the possible pool of creators and producers. In the long run, that competitive dynamic is not viable for Hollywood talent.

Rethinking The "Producer"

For this reason, Affleck and Cardinale may be too focused on the competitive dynamics of the talent side of the equation. They have fascinating competitive insights into these, particularly the “cultural footprint” of talent like LeBron James or Matt Damon offering a safer return on investment.

The uncomfortable question is whether they should be rethinking how the business of being a “producer” is being more broadly redefined. They have exponentially more competition in the production space from fans than they seem to realize, and those fans do not always need TV shows or movies. 

Comcast announced on Tuesday that its proposed spin-out will indeed happen. It will separate off entertainment and news channels including MSNBC, CNBC, USA, Oxygen, E!, Syfy and Golf Channel—all of which generated $7 billion in revenue in 2023—into a separate public company.

Last week I argued both solutions reflected media conglomerates pursuing either consolidation (Disney) or contraction (Comcast).

They seem inefficient and insufficient as Meta, Netflix, YouTube and generative artificial intelligence (AI) companies gun for the inefficiencies in their conglomerate business models. 

Key Takeaway

Neither Disney nor Comcast seems willing to pursue efficiency more aggressively. In doing so, they do not seem to be focused on the emerging disruptive forces in the marketplace. 

Total words: 1,300

Total time reading: 5 minutes

Disney

Disney finally has a story of fiscal efficiency in streaming. Two years ago it reported $4 billion in annual losses. In its fiscal Q4 2024, its direct-to-consumer businesses generated $253 million in operating income and $143 million for the year. It projects for 2025 that the division will reach $1 billion in operating income in its fiscal 2025 and a 10% operating margin in 2026.

Disney has accomplished that outcome with a mix of both layoffs and cost-cutting. It reduced its content budget from $27 billion in FY 2023 to $25 billion in FY 2024, and by another 4% in FY 2025.

Its recent Disney+ series “Agatha All Along” had a reported budget of around $40 million. Previous series like Loki (2021) and Moon Knight (2022) had budgets of $23 million and $24 million per episode, respectively.

However, there is also the sense that there is clever accounting that the Disney conglomerates can only pull off. After carving out a new Sports division for its ESPN properties—including streaming service ESPN+—in Q1 2024, it reported its first profit in streaming in Q3 2024. Its Entertainment streaming businesses lost $19 million but ESPN+ earned $66 million. 

In Q4, it reported that its Entertainment streaming businesses had generated $253 million in profit—14x more than the previous quarter—while its ESPN+ had only grown by 3%.

A conversation I had with a streaming executive on Tuesday noted that Disney management could still cut costs further and become more operationally and technologically efficient if it simplified its business to a single streaming platform. However, the problem is that Hulu and BAMTech remain two separate platforms, and—as The Information reported in 2019—the  “longstanding tensions” that both engineering teams brought with them into their merger remain unresolved. 

That is related to another conglomerate-specific problem: Disney and Comcast are locked in a dispute as to whether Disney owes Comcast more than $8.6 billion in exchange for the remaining 33% ownership of Hulu. The deal was one element of Disney’s acquisition of Fox’s entertainment assets in 2019. 

A final third-party appraisal of Hulu’s valuation was expected but not delivered in its annual report. That should not prevent further financial, operational and technological efficiencies because Hulu’s equity fair value is being assessed as of September 30, 2023. This obstacle is a symptom of Disney’s struggles to find these efficiencies easily as a large, $206 billion conglomerate that prefers its bet on consolidation over contraction.

Comcast

Comcast’s spin-out offers a mixed bag on efficiency. Cord-cutting accelerated in Q3 2024 to 9.3% among the top six U.S. operators—a loss of 1.3 million households—who publicly report their customer metrics. Virtual MVPD YouTube TV is now projected to be the largest Pay-TV distributor in 2026, and it currently reaches 8 million households, nearly 60% of total households for Comcast’s video business.

The spin-out has an accompanying reorganization:

• Chief Content Officer Donna Langley will become chairman of NBCUniversal Entertainment and Studios, gaining streamlined authority to greenlight productions and more financial visibility and control over content spending. 

• Matt Strauss, a Comcast veteran who heads the direct-to-consumer business, will be named chairman of NBCUniversal Media Group, with control over areas such as sports, ad sales and distribution.

With this move, the organization becomes both more and less efficient. It is more efficient because Comcast may better focus on and allocate resources to its growth businesses, and particularly Peacock. It is less efficient because Comcast’s media business has less scale at a time when advertisers are shifting their digital spend to platforms with more scale (Amazon Prime Video, Netflix, YouTube). 

Wall Street has reacted to this news with mixed reactions. There is a relief that the cable businesses no longer will be a weight on Comcast’s stock price. There is also the belief that the spin-out could get larger as “a potential partner and acquirer of other complementary media businesses.” But, the upside to both moves is not yet obvious or unlikely to ever manifest.

In terms of consolidation and contraction, the spin-out delivers both outcomes. The contraction of Comcast opens the door to consolidation opportunities for the cable networks in its spin-out. Those opportunities were less likely with the cable networks in-house. The contraction leaves Comcast increasingly relying on nearly $50B in revenues and 40% Adjusted EBITDA margins from its video and broadband businesses.

Efficiency

Both outcomes share the optics of efficiency but not the substance. Meanwhile, technology companies and new technologies like generative AI are redefining efficiency for them.

Netflix just redefined sports distribution with last week’s Paul vs. Tyson fight. Now, the NFL—a partner to NBC Sports and ESPN—is excited that “Netflix can bring you a gigantic audience to a major sporting event’, according to Netflix chief content officer Bela Bajaria. Global scale in content distribution is now possible and millions will tune in despite technical glitches

I also argued in “Media Conglomerates' Tough Choices With Meta's Generative AI” that Comcast and Disney increasingly need new sources of cash flow as cable network revenues decline. I asked at the end, “If small businesses can afford to re-imagine the value of old [intellectual property] better than a studio can, why not let them—and Meta—deliver and participate in the upside?”

Another way to frame this given the headlines above is to ask whether Disney and Comcast are too conservative given the speed of technological change. They should be pivoting faster towards a licensing model. 

Unfortunately, the answer is a hypothetical. Yes, a licensing business with recurring revenues can be profitable—Warner Music generated 27% profit margins on $1.2 billion in revenue for its Music Publishing business in its fiscal 2024—and therefore precedent for licensing models from Comcast and Disney. However, licensing TV series and movies is far more complicated given the long list of creative and productive talent involved.

The takeaway is that in this marketplace, efficiency as an outcome will always be insufficient. Meta, Netflix and generative AI are redefining "efficiency" for these conglomerates faster than they can redefine it for themselves. 

Two questions linger in both Disney's and Comcast's success stories:

1. Can they solve for the new standards of technological efficiency being dictated by their competition?

2. If they cannot, what circumstances will force their hand?

Neither company seems willing to pursue efficiency more aggressively. In doing so, they remain vulnerable to the emerging disruptive forces targeting those same inefficiencies. 

Author's Note On "Conglomerate"

It is worth reconsidering the term “conglomerate”.

I first used the term two years ago in "Can Disney TikTok-ify or Amazon Prime-ify Itself?" In 2023 I used it in 19 essays, and this year I have used it in 41 essays. I now plan to cut that usage back substantially.

The rationale for using the term has been simple: "Conglomerate" is a single word that summarizes a complex, centralized and evolvingcorporate decision-making structure across a diverse collection of assets. It also succinctly sums up a long, complex and ongoing history of mergers and acquisitions and corporate restructurings. 

However, the term is a mouthful of clunky consonants. The “ng” and “om” require a more nasal pronunciation. Unsurprisingly, there are few, if any, heated debates about the "conglomerate model" in mainstream media or social media.

Nor is the conglomerate the end-all-be-all target of disruption for the next generation of media entrepreneurs. For example, in the connected TV space—where Amazon, YouTube and Netflix now compete for linear upfront dollars—the market focuses on capturing advertising spend traditionally allocated to inventory of cable and broadcast TV channels. None of those companies are also gunning for Comcast’s or Disney’s theme parks businesses.

For this reason, "efficiency" may be the better term and concept going forward. Conglomerates capture and exploit efficiency in all its various forms, specifically financial, operational, technological and legal. Disruptive market forces always gun for inefficiencies within conglomerates but not to topple the conglomerates themselves. 

Going forward, I will be reorienting my arguments and analyses around efficiency because—even if the business rationale for the media conglomerate is disappearing with cord-cutting—there are more market inefficiencies created by conglomerates than there are actual media conglomerates. Those inefficiencies are where the market excitement lies.

Yes, the user experience of watching the Paul vs. Tyson fight was a minimum viable product (MVP). The cultural event of the year often showed “a spinning red wheel with a random percentage in the middle of it”.  The broadcast did not out-ESPN ESPN, out-Eurosport Eurosport or out-Hotstar Hotstar with the quality of its viewing experience. 

But, its commentary, celebrity appearances, and graphics seemed to match the experience of watching a live sports event on any of those services. 

In the end, an international viewing audience of 60 million households tuned into the same streaming app to watch the fight live around the world, peaking at 65 million concurrent streams. Fighter Jake Paul said after the fight that 120 million viewers watched the fight globally.

If we estimate off a metric Netflix Co-CEO Ted Sarandos shared with The New York Times in May—”we’re currently programming for about 650 million people around the world”—then that number goes up to 140 and 160 million viewers, at least [NOTE: They had just announced 270 million members in their Q1 2024 earnings call, implying around 2.4 viewers per household.]

Perhaps most importantly, those ~150 million streaming viewers seemed not to need the quality of a television broadcast to tune in and stick around. According to the website Down Detector, nearly 85,000 viewers logged problems with outages or streaming leading up to the fight.

That is less than 0.01% of the total estimated viewing audience and 0.1% of all households, suggesting almost 100% of viewers tolerated the imperfect, MVP viewing experience on Netflix that no other platform built yet.

Key Takeaway

The NFL's 2029 opt-out from its recent distribution deal seems more likely after the Paul vs. Tyson fight. The best path ahead for Disney and ESPN lies in a profitable purgatory of their own making.

Total words: 1,300

Total time reading: 5 minutes

Kayfabe

There is an obvious counterargument to the above: Paul vs. Tyson had more in common with the “kayfabe” nature of a WWE match. “Kayfabe” is the illusion that the storylines performed in the ring are real.

There were unusual rules for the fight approved by the Texas Department of Licensing and Regulation:

• Both fighters wore 14-ounce gloves instead of the standard 10-ounce gloves typically used by heavyweight fighters;

• Each round lasted two minutes rather than the usual three minutes; 

• The fight consisted of eight rounds instead of the standard 10 or 12; and,

• Because the fight was sanctioned, it needed reach four rounds to become official.

Also lurking in the background were the rumored paydays for both fighters:

• Tyson was estimated to make around $20 million for the fight; and,

• Paul was estimated to make $40 million, as he helped to put this event together as the co-founder of Most Valuable Promotions.

All these constraints and incentives suggest both fighters gamed the fight to last eight rounds. Consumers were watching a circus act—Mike Tyson is 58 years old—dressed up to look like a professional boxing match. 

Netflix Goes From Crawl to Walk

It is reasonable to assume Netflix needed the fight to last eight rounds, too. This was the first test of its live-streaming technology at a global scale larger than previous comedy specials with Joe Rogan (up to 2.5 million views est.)  and Chris Rock (up to 15.47 million views est.).

The previous record for concurrent streams was set by Peacock in January with 16.3 million tuning into the AFC Wild Card game. Netflix quadrupled that less  than 10 months later.

Co-CEO Greg Peters has long talked up its “crawl, walk, run” approach. We can reasonably assume the previous comedy specials, its Netflix Cup broadcast, its tennis event The Netflix Slam, and a “Love Is Blind” Season 4 reunion special were the "crawl" phase. 

This fight appeared to mark the “walk” phase, where the platform is not ready to run but is willing to suffer the bumps and bruises of a learning curve. A four-hour broadcast offered an unusually long window for the team to learn and problem-solve. According to Bloomberg, its CTO Elizabeth Stone wrote in an internal memo:

“This unprecedented scale created many technical challenges, which the launch team tackled brilliantly by prioritizing stability of the stream for the majority of viewers. I'm sure many of you have seen the chatter in the press and on social media about the quality issues. We don't want to dismiss the poor experience of some members, and know we have room for improvement, but still consider this event a huge success.”

I read this as, “Ignore the criticisms and social media snark. Subscribers stuck around, our launch team learned a ton and Netflix changed the paradigm for sports streaming distribution.”

Netflix vs. Disney

Netflix’s win is unfortunate timing for Disney, which had its earnings call only the night before. On the FY Q4 2024 earnings call, Disney CEO Robert Iger talked up the launch of its ESPN flagship:

“I think one of the things that [...] hasn't been appreciated yet is that when you apply technology to the presentation of sports, almost anything is possible. So imagine an AI-driven personalized sports center as a feature, for instance. Imagine essentially being able to tailor your sports experience, not just watching SportsCenter, but in a thoroughly personalized way. I think essentially, it will be designed to serve the consumer in the most compelling way ESPN has ever served the consumer.”

There is an obvious contrast between Iger selling investors on technological possibilities and the reality of Netflix distributing a live event globally, albeit imperfectly. Netflix has delivered investors and consumers a tangible paradigm for what a sports event streamed at scale will look like. 

That contrast widens when considering Netflix’s three-season deal with the NFL for Christmas Day football games that launches next month. Whatever Netflix may be able to achieve with an international broadcast of an NFL game will have a significant benchmark to beat: A global TV audience of 62.5 million viewers watched the Kansas City Chiefs defeat the San Francisco​ 49ers in Super Bowl LVIII. That was a 10% increase over the previous Super Bowl broadcast. It was also high quality.

In this light, the Paul vs. Tyson fight seems significant: Global audiences at scale will suffer through an MVP experience to watch an event on the same streaming app. That is the antithesis of the quality upon which Disney’s brand and business are built and sold. Risk-taking seems to be an advantage for Netflix.

However, it also is the antithesis of why advertisers value Disney over other companies, including Netflix. That implies Netflix will need to get to the "Run" stage sooner than later.

ESPN & John Malone

Meanwhile, ESPN is stuck in a purgatory of both a declining but profitable linear business and a growing and increasingly profitable streaming business.

It reads like the right strategy in light of last week’s comments from Liberty Media chairman John Malone at the Paley Center for Media. He argued Disney and NBCUniversal made a mistake in deciding that “it was better to jump in and go the direct-to-consumer route.” He believes they would have been better off trying to “evolve the traditional wholesale-retail distribution approach in order to take advantage of this network neutrality free ride and bypassing their traditional distributors.”

Disney’s recent distribution deals with Charter and DirecTV are examples of Malone’s “better” approach. Consumers may access Disney’s cable channels and opt-in at no cost to complimentary ad-supported tiers of Disney streaming services. The ESPN flagship app will also be bundled by Charter and DirecTV when it launches next year. 

The question is, effectively, why should Disney try to compete with Netflix in streaming while pursuing a strategy limited by both linear and streaming distribution?

In 2029 the NFL may terminate any or all of its $100 billion worth of deals with Amazon, ABC/ESPN, CBS, FOX and NBC via an opt-out clause. If Netflix can successfully launch and grow a new model for sports streaming distribution at a global scale, neither of Disney's options seem viable in the long run. It must either ride out cable's demise (Malone) or it must rapidly evolve to a global distributor (Netflix). 

Disney's best and most profitable business rationale lies the former, and Netflix's lies in the latter. 

On Wednesday, Liberty Media announced he would return as interim CEO  and current CEO Greg Maffei would be stepping down into an advisory role at year’s end.

The press release outlined the (wonky) corporate reasons for Malone's return are “transactions to simplify corporate structure.” Those include a split of Liberty Media into two entities, Liberty Media and Liberty Live. 

Both headlines hint that a "next wave of consolidation and contraction" for media conglomerates—which I wrote about last August—has started. According to a report from CNBC, the Liberty Media conglomerate is now “in its final act”.

However, Malone’s vision for Liberty and arguments for consolidation offer mixed messages about any forthcoming wave emerging. It also seems to discount the technological shifts underfoot.

Key Takeaway

Malone's vision for the future of media seems to discount the speed of technological disruption and the positioning of companies like Meta to shape demand for generative AI. That makes it flawed, regardless of whether the future of media conglomerates is consolidation or contraction.

Total words: 1,000

Total time reading: 4 minutes

Contraction

Over the past 24 hours, we have had two market signals about media conglomerates contracting.

First, Maffei said in a press release that Liberty Live’s spin-out “should reduce the discount to net asset value of our Liberty Live stock and enhance trading liquidity at both entities.” Meaning, its Live assets should be valued more by investors outside of the Liberty Media conglomerate than within it. The Liberty Media conglomerate has hit a wall. 

That rationale mirrors a proposal from Comcast’s Q3 2024 earnings call from President Mike Cavanagh. He shared that Comcast is exploring whether a new well-capitalized “spin-off” company—owned by its shareholders and composed of NBCUniversal’s “strong portfolio of cable networks”—would be in a better position “to take advantage of opportunities in the changing media landscape.” 

Second, on this morning’s earnings call Disney management was asked about whether it would divest its cable networks because CEO Robert Iger floated the idea last summer.

Disney CFO Hugh Johnston responded that the company has ruled out divestitures to create shareholder value, and shared that he considers two variables when evaluating a divestiture:

1. Prices Disney can get for the assets, and

2. Any resulting operational friction from the divestment process.

A divestiture of cable assets did not meet those criteria. Disney needs to remain a media conglomerate because the divestiture costs are higher than the value that could be unlocked from spinning those assets out.

The takeaway from both market signals was that contraction may not be in the cards for many conglomerates except for those with businesses that the market currently values. Media conglomerates are likely persist in one form or another. There may not be an anticipated wave of contraction.

Consolidation

In Monday’s essay I wrote about how a “big” media company that emerges from megamerger-type consolidatin will not need to look like Disney's enormous library of intellectual property (“IP”) or“flywheel” of business divisions built upon that IP. Instead, it will need the scale of its owner to compete like a Sony Corporation or Comcast with a studio integrated into a broader ecosystem.

Iger told investors that Disney was not pursuing consolidation. It has “already consolidated” via its merger with Fox’s entertainment business and does not need more assets “right now from a distribution or a content perspective to thrive in basically a disrupted media world.” Instead, the story was operational efficiency: Iger and Johnston promised double digit growth in adjusted earnings per share and in cash provided by operations. 

Warner Bros. Discovery has been a candidate for consolidation since April, the WBD merger’s complicated tax structure (a Reverse Morris Trust) was legally permitted to unwind. But, a recent S4 regulatory filing with the SEC about Skydance’s acquisition of Paramount reveals that Warner Bros. Discovery engaged in merger talks with Paramount Global. Those talks fell through after a few months in February 2024.

Malone told the Paley Center that Warner Bros. Discovery was burdened by debt but now is “starting to experience what the original theory was, which was to take underexploited Warner Bros. programming and library content, create a brand and then distribute it globally. ” He believes the bet is “starting to work… with very rapid global growth.”

The sense from both companies is they believe they have enough to fight and succeed on their own. At the same time, though, both Warner Bros. Discovery and Disney are doing so through operational efficiencies while battling to build compelling growth stories for investors (Both stocks are up 6% since their respective earnings calls). 

Notably Malone did not discuss the type of consolidation that Sony Entertainment Pictures CEO Tony Vinciquerra recently envisioned: A studio needs the scale of a Sony Corporation or Comcast to maximize the value of its IP. It can be ruled out for Disney, which has a $200 billion market capitalization. But, it cannot be ruled out for Warner Bros. Discovery, which has a market cap at $24 billion or 12% of Disney's.

The Disconnect

The key variable that differentiates consolidation and consolidation in Malone's vision is ownership of IP. If the conglomerate owns IP, it should remain as is. If it does not own IP, it should rethink its corporate purpose and contract.

Monday’s essay asked and answered, “How can a merger or acquisition create more opportunities for IP to be monetized than within the closed marketplaces of walled gardens?”

Emerging technologies will increasingly enable creators to monetize this IP. That seems inevitable. However, Malone’s vision seems to discount conglomerates' need for cash flows from beyond traditional distribution and their ability to evolve under legacy media management (Warner Bros. Discovery struggles with gaming). It may just be that Malone has confirmation bias towards traditional distribution, even if he sits at the heart of efficient markets of information as a shareholder in multiple media businesses. 

That said, Malone's vision seems to discount the speed of technological disruption and the positioning of companies like Meta to shape demand for generative AI. That makes it flawed, regardless of whether the future of media conglomerates is consolidation or contraction.

Donald Trump’s incoming Administration will have less intrusive antitrust regulation than Lina Khan's Federal Trade Commission under President Joe Biden. Now, media and Wall Street coverage has instantly shifted to speculation of the first megamergers since Disney acquired much of Fox’s entertainment empire in 2019 for $71 billion.

If one believes Sony Pictures Entertainment CEO Tony Vinciquerra, we will see “mergers and bankruptcies and sales and all kinds of fun things happening over the next couple of years.” He also predicted at an investor conference in September that “big companies will be the ones that survive unless they make some massive mistake or miscalculation.”

The other part of this quote has been summarized by Hollywood trades, and its missing parts are worth adding: 

“So, scale will be important. And people say we're under scale at Sony Pictures. But we have PlayStation music, all the electronics background of Sony. I don't think -- we're not under scale if you look at the company in total”.

In other words, a “big” media company that emerges from megamergers will not need to look like Disney's enormous library of intellectual property (“IP”) or“flywheel” of business divisions built upon that IP. Instead, it will need the scale of its owner to compete like a Sony Corporation or Comcast with a studio integrated into a broader ecosystem.

That presents a difficult question for a media marketplace in secular decline:

Will any merger solve for the suboptimal monetization of IP libraries in the streaming era? 

Key Takeaway

A merger or acquisition likely will not create more opportunities for IP to be monetized. Instead, the logical solutions are either to aim for free marketplaces for IP or to let third-party technologies rethink walled gardens.    

Total words: 1,700

Total time reading: 7 minutes

Suboptimal Monetization Of IP Libraries

Consumers increasingly seem more sentimental for the IP of media companies than they are for the TV series and movie formats. Therefore, there are three business problems a mega-merger will need to solve for IP libraries:

1. Linear is declining; or, as Vinciquerra described it, cable networks have become an “albatross” around the necks of media companies that still own them;

2. Theatrical revenues are declining; and,

3. Streaming is not a growth business that compensates for either.

A megamerger that puts a studio in a Sony-like ecosystem solves for the second, only. Comcast recently proposed a version of that outcome where it will spin out its cable networks into a separate company. That would leave the Universal Studios division within Comcast. Disney CEO Robert Iger floated the same idea in Sun Valley in July 2023 to CNBC’s David Faber, but never got any traction.

A spin-off of cable channels may create a better outcome for cable channels to be more efficient and generate more cash in decline. But, it complicates the content supply for streaming service libraries.

For example, if Comcast spins out both Bravo and USA networks, Peacock will still need to exclusively license content from those two companies. However, Bravo and USA networks will deliver more shareholder value by letting free markets dictate the value of their IP. Peacock is worse off in this outcome.

To date, selling IP to owned-and-operated streamers—which I described in step 2 of “The Doom Loop of The Mogul”—has resulted in “no real marketplace in Hollywood right now”. Given that Hollywood increasingly needs free marketplace dynamics to maximize the value of its IP, neither a spin-off nor a licensing deal will result in that outcome.

That raises the question: How can a merger or acquisition create more opportunities for IP to be monetized than within the closed marketplaces of walled gardens?

The logical solutions are either to aim for free marketplaces or to let third-party technologies rethink walled gardens.    

Alternative Solution #1: Cloud-Usage Model

One solution for a free marketplace is a cloud-usage model that I proposed in September and speculated that Oracle will pursue with David Ellison’s acquisition of Paramount (Oracle Chairman Larry Ellison personally funded the acquisition). In this model, Apple, Amazon or Paramount would tear down their walled gardens, toss out their subscription models and become marketplaces like Amazon Web Services (AWS).

However, I noted that licensing is still an issue:

“No licensing agreements are structured to reflect cloud usage. The contracts with Hollywood unions are structured to reflect benchmarks in the streaming business model. This will neither be an instantaneous nor seamless pivot.”

As for expanding walled gardens, Comcast President Mike Cavanagh suggested to the market that the outcome is appealing but complicated: “[W]e would consider partnerships in streaming despite their complexities.” That is an implicit concession that it has learned the same lesson as Disney: Expanding Peacock’s library will not change consumer demand enough to mitigate losses of $436 million in Adjusted EBITDA in one quarter.

I also argued last week that: 

Cavanagh’s concession of “complexities” reflects an understanding that Peacock may be part of a flywheel without needing to be a “walled garden” anymore. He is referring specifically to the complexities around ad delivery and consumer data. Both are optimal in a single, centralized solution. So, the fate of a valuable IP library is intertwined with advertising (more on this below).

It is clear that executives at Comcast and other media companies understand there a merger or acquisition can accomplish only so much for a valuable library of IP.

Alternative Solution #2: Spotify for IP 

That said, a merger or acquisition delivers a concrete outcome and return for shareholders, whereas technology-focused models promise exciting but more speculative outcomes.

The core argument of my essays on four broad buckets of “back to the people” business models—where studios are likely to license IP to professional amateur content creators— is that technology can deliver a solution for unlocking hidden value.

The most concrete proposal was made by Gavin Purcell—a producer and host of the podcast “AI for Humans”—who imagined a Spotify-type platform emerging to enable these “back to the people” models. IP rights holders would receive a share of royalties per use, much as artists receive a payment per stream on Spotify. IP rights holders likely “will be paid less than they have in the past”. But, like music labels, they will get paid. 

Showrunner founder Edward Saatchi imagines his generative AI, interactive storytelling platform “passing on a portion of its subscription fees based on how many shows a creator makes and how many people watch.” He also envisions that “people will pay to use elements of TV shows or movies to create and tell their own stories. 

Showrunner is currently in its Alpha stage and does not envision a studio licensing its most valuable IP for another two years. Inkitt—“a startup that lets writers self-publish books and uses artificial intelligence to edit, distribute and sell the most promising works on its Galatea app”—is live with a similar model in the fan fiction space. Inkitt readers can select their favorite Galatea story for inspiration, and customize elements like character names and the story’s setting.  

Inkitt pays authors a royalty when a reader generates fan fiction, and it is the same royalty that they earn from the sale of an original title. It appears to be one market precedent for a Spotify for IP-type solution.

Alternative Solutions

A “Spotify for licensing” may not be a competitive investment opportunity given that Spotify, Netflix, Meta and YouTube already have licensing solutions in place. One could expand that list to platforms like Epic Games’ Fortnite or Roblox, where users pay licensed third-party character “skins” to play as their favorite characters in virtual worlds.

A key problem faced by sophisticated third-party platforms licensing valuable IP face lies in game theory  (or, the study of mathematical models of strategic interactions with rational consumers). Technology is fragmenting the books, movies and TV shows of famous IP into value propositions where the “beloved characters and worlds” are front and center. 

Warner Bros. Discovery may know which users like DC Comics’ intellectual property from its Max streaming platform or “Multiversus”, its free-to-play crossover fighting game. But, Netflix, Amazon, Meta or YouTube have built multidimensional lenses on what those DC fans also like across shopping and other affinities. All are built upon the assumption that IP is one variable among many to understand the consumer. Their key to success is targeting the right advertisement to the right consumer.

In this light, the only competitive solution to ensure libraries of valuable IP do not become beholden to technology companies is for a white knight to come in with a competitive solution. Two immediate solutions come to mind.

Alternative Solution #3(a): Data Warehouses

One feasible white knight solution would be a data warehouse service offering itself to media companies as the default platform for managing their IP licensing and advertising. As I wrote in May, data “warehouse-native” marketing technology (which allows marketers to create, run, and manage online marketing campaigns and conduct onsite marketing) is “creeping” into marketer data stacks. 

First-party apps that now run within the data warehouses can assume the roles of third-party ad servers, demand-side platforms (DSPs), and attribution/measurement partners. Snowflake and Oracle are leading providers. Adding licensed IP libraries to these solutions would make them more robust on the open market, especially as walled gardens.

Alternative Solution #3(b): Curation

Another solution borrows from curation from the ad tech space, or the process of advertisers using “audience, contextual and supply chain signals from publishers to curate collections of auctions across various sellers.” A handful of first-party data signals from both the publisher and the advertiser are combined to help to streamline open auctions in programmatic bidding toward the inventory that is most valuable to both parties. 

In this model, a curation partner would capture first-party data signals specific to its IP from both the publisher and the advertiser. (e.g., on which platforms consumers most consume Superman IP). It could then “curate” the optimal signals specific to “beloved characters and worlds” and help publishers optimally monetize IP with creators, gaming publishers, generative AI platforms and (perhaps) meme coins out in the “wild” of the open internet.

Such a solution at scale across multiple media formats beyond video could create leverage over video-focused platforms like Amazon, Apple, Meta, YouTube or TikTok that media companies have failed to build. The catch is that data warehouse providers Oracle (Paramount) and Amazon (MGM Studios, Prime Video) have head starts in this space.

Presumably, this outcome would work based on data partnerships in the short term. If this solution is viable in the long term, a wave of mergers toward the aggregation of IP libraries could make sense to improve studios’ bargaining leverage over the data warehouses.

For now, the need for mergers in this instance and in the other scenarios, above, is more speculative than necessary.

The post answers my question from last month: Why did Netflix believe it needed to deliver blockbuster, major-level (“AAA”) games from blockbuster budgets to win? 

Verdu’s answer implies that Netflix management realized earlier this year— he stepped down as VP of Games in June—that they did not need to anymore. As Verdu wrote, Generative AI is “a 'once in a generation’ inflection point for game development and player experiences using generative AI.” 

In other words, the technology has matured to the point where Netflix is best off taking a blockbuster budget—AAA games greenlit in 2023 for potential releases in 2024 or 2025 typically received development budgets of $200 million or higher—and reallocating it across multiple teams using generative AI tools.

The business rationale makes sense. There are two logical questions:

1. Is this what consumers want? And,

2. If so, how is Netflix's competitive future being shaped by generative AI? 

Key Takeaway

Netflix VP Mike Verdu’s post does not read like generative AI will “break” either of its gaming or streaming business models. But, the initiative seems to be one big step closer to that outcome.

Total words: 1,300

Total time reading: 5 minutes

What Do Consumers Want?—Netflix Version

Netflix has launched over 100 games, to date. It told investors in August that it has another 80 in development. The implication is that with over 280 million members, 

Co-CEO Ted Sarandos told investors on its Q2 earnings call that its objective with mobile games is to give “the superfan a place to be in between seasons [of a show], and to be able to use the game platform to introduce new characters and new storylines or new plot twist events."

Faster and cheaper gaming production with generative AI will translate to a higher volume of games at a lower cost. In turn, fans of Netflix’s original IP will have more options for engaging with their favorite shows and characters. Higher engagement translates into lower churn between seasons and more “customer delight”.

Verdu’s post also implies that “an almost unimaginable collection of new capabilities” will enable gaming capabilities that consumers have not seen before. The post notes that “AI will enable big game teams to move much faster”, meaning the bet on AAA games is not off the table for Netflix. AAA games can take as long as seven years to develop. Generative AI appears to be able to accelerate that process.

Last, there are headwinds in the broader gaming marketplace. Studios are being shut down following a drop in post-pandemic consumer demand for games. Higher interest rates have made it harder for studios to borrow. Advanced technologies are expensive and consumer expectations for better experiences have driven up development budgets. 10,000 jobs were lost in 2023 and an additional 13,000 layoffs have taken place in 2024, to date. 

Netflix’s vision of faster and cheaper game development with generative AI seems to be one solution to this crisis. Like Microsoft (Xbox) and Sony (PlayStation), it has an enormous installed base of nearly 300 million members it can serve immediately upon the release of a new game. It can capture gamers that gaming studios are increasingly struggling to capture.

However, it remains an open question as to whether its members want or need exponentially more than 200 games, and especially games produced with generative AI. “GTA: San Andreas” is Netflix's most downloaded game of all time with nearly 10 million downloads since its December 2023 launch, which is only 3.5% of its membership.

Netflix vs. YouTube & “Flooding”

Games and videos are an apples to oranges comparison. But, in the attention economy they compete. YouTube and Netflix are now locked in a battle for attention and the efficiencies in both video and games production from generative AI content risks “flooding” both platforms with AI content.

The term comes from an interview of YouTube CEO Neal Mohan by YouTube creators Colin & Samir last October. About 25:00 into the interview, Colin Rosenblum shared with Mohan a fear about the threats of AI-generated content to creators:

“I’m concerned about the influx of semi-autonomous content and it perhaps floods YouTube... and there is perhaps an appetite for it and it starts to change what it means for viewers to enjoy the platform and what it means for creators to succeed on the platform, if there truly is so much [content]”.

Chaudry told me in an interview earlier this year for my Medium Shift column. that he and Colin worry that “inauthentic, AI-generated storytelling” will win at the expense of human-created content. The connections between human storytellers and human audiences will be lost.

Netflix’s growing competitive threat from cheaper “premium” content on YouTube capturing audience share, especially on connected TVs, reflects how vulnerable it is to "flooding", too.  I wrote in September’s “YouTube Is Forcing Netflix to ‘Break’ Its Business Model”: “The more YouTube pressures Netflix into producing more content at a lower cost, the more Netflix will have “to break [the business]” and undermine the studio side of its model.”

Netflix is unlikely to flood its platform with games at the expense of TV series and movies, even if it substantially increases its production using Generative AI. It would have to increase its game production 60x to match the 6,500+ TV series and movies it offers in various countries around the world.

But, I have also argued in that essay: 

“Netflix appears to be getting squeezed into letting ‘fans [become] the producers of the content they want to watch.’ YouTube has redefined ‘premium content’ with the help of creators. In doing so, it has redefined consumers' investment in content and creators from something emotional into something financial.”

Netflix is responding to this challenge by building more games. It is within the realm of possibility that it could leverage generative AI to build fan-first storytelling platforms like Showrunner, too. In either case, it will be producing more fan-driven interactive content in a model with economics closer to YouTube than Netflix’s distribution model for TV series and movies.

What If…

Netflix has offered its subscribers choose-your-own-adventure, interactive video content. It first launched interactive titles in 2017 with Puss in Book: Trapped in an Epic Tale, and was nominated for and won multiple awards with Black Mirror: Bandersnatch.

Earlier this week, Netflix confirmed to The Verge that it is removing all but four of the 24 titles currently on the platform. Both this announcement and Verdu’s LinkedIn post reminds me of an observation made by Chaudry told me earlier this year: 

“text-entry search boxes on platforms like Netflix and YouTube have been conditioning users for text-to-video AI-generated content. There are only a few steps between searching for a title using text and creating a video with a text prompt.”

Meaning, Netflix has built and defined the “text to prompt” user experience and is now embracing generative AI for gaming. “Text to video” inhabits the logical territory in between. The implication is that Verdu’s remit may be broader than his LinkedIn post suggests. 

I wrote last week in the conclusion to last week’s “Netflix's Risky Pivot To Moments” that “the question in the long-term—and that may be as soon as 2025—is what Netflix’s model will become in the age of AI content generation.” This may be the working answer, for now.

Either way, the announcement emerges as Netflix navigates growing competitive pressures from Silicon Valley to rethink both its business model and broader value proposition for consumers. In September, Co-CEO Ted Sarandos foreshadowed these challenges in a speech to the Royal Television Society’s (RTS) London Convention 2024. He told the audience that Netflix must cannibalize its own business and “constantly challenge ourselves, to break [the business] and move our business forward on behalf of our consumers.” 

Verdu’s post does not read like generative AI is “breaking” Netflix’s gaming or streaming business models. But, the initiative seems to be one big step closer to that outcome.

He added, first, “[W]e would consider partnerships in streaming despite their complexities.” In other words, Peacock can no longer exist on its own. It lost $436 million in Adjusted EBITDA and gained only 3 million net new subscribers despite an “exceptional quarter” with the Paris Olympics, an exclusive NFL game from Brazil, the return of the Big Ten, and “several entertainment hits during the quarter including Love Island, Bel Air and Fight Night.” 

Second, Comcast is exploring whether a new well-capitalized “spin-off” company—owned by its shareholders and composed of its “strong portfolio of cable networks”—would be in a better position “to take advantage of opportunities in the changing media landscape.” 

If the business logic sounds familiar, Disney CEO Robert Iger floated the same idea in Sun Valley in July 2023 to CNBC’s David Faber: “The creativity and content [Disney’s cable networks] create is core to Disney, but the distribution model, the business model that forms the underpinning of that business, and that has delivered great profits over the years, is definitely broken.”

The interesting difference between the two proposals is a single word: “complexities”. Cavanagh used it to describe Peacock’s challenges within and without the NBCUniversal “flywheel”. However, Iger only hinted at it in his framing of Disney’s inextricably intertwined content and cable distribution businesses.

Both raise two important questions. Given these "complexities":

1. Can media conglomerates disrupt their flywheels? And,

2. What will happen if they can’t?

Key Takeaway

Comcast seems to be telling Wall Street that "flywheel" models within a "walled garden" have become self-defeating in the streaming era. This presents some difficult questions for Disney.

Total words: 1,400

Total time reading: 5 minutes

Can Media Conglomerates Disrupt Their Flywheels?

A “flywheel” is a connected ecosystem of businesses where improvements in one business segment can help to boost the performance of other business segments. The flywheel is a growth model popularized by Walt Disney’s famous 1957 multispoke diagram. Disney and NBCUniversal bought cable networks to boost cash flow for the flywheel and enable individual businesses to take financial risks they otherwise would not have been able to take advantage of.

The “flywheel” model suffers as cash flow from cable declines. Also, audience demand for the content fragments: It is hard to aggregate audiences at scale both in cable and streaming when competing against technology platforms with enormous global scale like Amazon, Netflix and YouTube. It is even harder to aggregate those audiences when asking for a monthly or annual subscription fee.

Notably, Iger seems to believe the cable business has been “broken” by streaming and that streaming is Disney’s best solution for aggregating those audiences. Cavanagh also seems to believe that streaming is the future as a  “growth business” but refuses to call the cable model “broken”. 

He is implying that NBCUniversal will be able to spin out its cable channels because the decline of its linear businesses will be slower. That may be because MSNBC and USA Network are in the top 11 most-watched networks, and Bravo is 30th. Also, top titles from NBCUniversal’s libraries of content from Bravo and USA Network have performed well when licensed to Hulu and Netflix, respectively. 

A potential spin-out has a better business rationale for Comcast than it does for Disney. Excluding ESPN channels and ABC, Disney's channels all rank below 60th and their Nielsen audiences averaged around or below 100,000 households last week. Disney tends not to license its titles but is making exceptions in the streaming era.

This means Comcast may be better positioned to generate cash flow for its "flywheel "while Disney struggles to find new sources to replace its lost cable profits.

And What Will Happen If They Can’t?

It is important to note that Disney and Comcast are the only two media conglomerates with a theme parks business. In this light, NBCUniversal has one advantage over Disney in the scenario of a spin-out: Neither the reality content of Bravo nor the dramas of NBCUniversal are the stuff of theme park rides. In comparison, everything Disney produces is intended to become either be a ride, an attraction or simply merchandise for sale.

Whatever may remain of NBCUniversal after a spinout of its cable businesses will look like a smaller Disney. In its fiscal Q2 2024, it generated 1.98 billion in revenues and $632 million in Adjusted EBITDA. Disney is almost 4x the Universal theme parks business: It generated $8.4 billion in revenues and $2.2 billion in operating income. 

In some ways, NBCUniversal’s flywheel may be better positioned for a future without cable channels. On its earnings call, Comcast announced the grand opening of Universal Epic Universe in May 2025 and sold it as “the most ambitious and technologically sophisticated theme park ever created.” The park will offer “worlds” from IP it does not own, like “The Wizarding World of Harry Potter – Ministry of Magic” and “Super Nintendo World”. The only IP that it owns for that new park are “How to Train Your Dragon – Isle of Berk” and “Dark Universe”. 

The implication is that Comcast’s theme parks are less vulnerable to whatever NBCUniversal chooses to do with its cable assets because the latter do not produce IP for the “flywheel”. Meaning, Comcast and Disney face the same challenges with their linear businesses but Comcast has more flexibility in evolving its flywheel without some of its cable networks.

This reads counterintuitive given the financial heft of Disney’s theme parks business: $26 billion in revenues through FY Q3 2024 and $7.6 billion in operating income (64% of total operating income). But, Disney seems constrained by the strengths of its “flywheel” and “walled garden” model whereas Comcast is unconstrained by the weaknesses of its “walled garden” model.

As For Streaming…

The difficult question for both companies is whether streaming belongs within the “flywheel” model. Both companies are arguing streaming is a growth business despite growing evidence to the contrary. But, they also both have a unique mix of libraries of intellectual property and expensive sports distribution deals that do not seem to be able to compete at scale. 

In the cable model, sports distribution has been a lucrative source of additional cash flow for the “flywheel” model. Every household paid for those channels, the channels were the most expensive in the bundle but a minority watched them. Both Peacock’s data from the Paris Olympics and ESPN+ consumption in Nielsen’s The Gauge (0.16% of all viewing) suggest that streaming is unlikely to replace that lucrative model.

Comcast’s willingness to partner around Peacock seems to acknowledge this reality. Cavanagh’s concession of “complexities” reflects an understanding that Peacock may be part of a flywheel without needing to be a “walled garden” anymore. He is specifically referring to the complexities around ad delivery and consumer data, both of which are optimal in a single, centralized solution. 

More broadly, he is saying streaming may not power a “flywheel” within a “walled garden” but “flywheel” models may still benefit from streaming. That presents some difficult questions for Disney, which has persisted with its “walled garden” approach to streaming and is doubling down with the launch of a “flagship” ESPN streaming product next year. 

It needs the latter to succeed. That said, its attempt to participate in the Venu Sports joint venture with Warner Bros. Discovery and Fox was a similar acknowledgement. 

If we assume Peacock is a “stalking horse” for ESPN, it already has proved that sports no longer has the same value to the “flywheel” model in the streaming era. In believing both sports and streaming can replace cable in its “flywheel” model, its need for scale will be greater and more existentially risky than NBCUniversal’s. 

The Flywheel Is Dead, Long Live The Flywheel

If Comcast is able to spin out its cable channels after Disney failed to, that will be a strong signal that Disney’s bet on streaming is far more wrong-headed than it has been letting investors know. That may read counterintuitive because Disney has a streaming business that is generating $20 billion in revenue per year and is “profitable”. 

But, Disney’s “flywheel” within its "walled garden" has long been its secret sauce. Iger’s vision was that streaming could be a logical replacement for cable and at a global scale.

Comcast seems to be telling Wall Street with both data and its proposed spin-out that "flywheel" models within a "walled garden" have become self-defeating in the streaming era.

They require a different calculus. Agility matters most.