Google-owned YouTube recently announced an alternative path to that outcome: A deepfake detection tool for creators which it opened up to politicians and journalists last month, and expanded to actors, athletes, creators and musicians last week. Anyone “at high risk of having their likeness abused” can sign up to identify and request removal of deepfakes on its platform, whether or not they have a YouTube account.

The model is the “same concept” as its Content ID system to identify copyrighted content uploaded to its servers. Effectively, a celebrity, creator or public figure is uploading their likeness into Google’s cloud infrastructure, where AI learns to identify—but not produce—those likenesses across all content on the platform. That is an enforcement mechanism that operates at both the cloud-level—where the likenesses are identified—and the platform-level—where the videos with violations are either flagged or removed. This mechanism is “air-gapped”: constrained within the cloud and isolated from other models and the broader internet.

That is a double-edged sword. Likenesses can be protected independent of studios, which are slow to partner with cloud companies. However, unlike content ID—where music rightsholders are paid a portion of ad revenue where use of their music was identified—celebrities are not being compensated when their likeness is flagged. They and/or their management team must request the removal manually.

But that request for removal “does not guarantee that YouTube will pull the video”, as its community guidelines allow for “a number of cases like parody and satire” to remain on the platform.

These are operational costs. The deeper problems are structural.

Past essays related to today’s analysis:

Read more

One core message of Netflix’s Q1 2026 letter to shareholders last week is that the “landscape” of the media marketplace is “more dynamic than ever”. In a section entitled “Our Focus”, management lists a series of factors impacting the business:

• “Streaming continues to take view share from linear.

• Technology is enabling a massive increase in video content from a wide variety of storytellers.

• There are more ways for consumers to connect to the stories they love, including through open content platforms, interactive and gaming services.

• ‘Premium’ content is increasingly defined by the viewer.

• The lines between entertainment on TV and mobile devices are blurring.”

This story has been told in different ways by Netflix management over the course of the last few years. The gist has always been Netflix competes with some mix of YouTube, gaming and social media platforms. The new detail is in the second bullet point about “a massive increase in video content from a wide variety of storytellers.” That phrasing broadly implies both more human creators, and more content from generative AI tools and platforms.

The result is more creators, more content and more platforms are fragmenting both the demand for consuming content on TV and the definitions of “premium content”. Netflix is positioning itself as a success story in subscription streaming—”slightly higher-than-planned subscription revenue”—and ad-supported streaming—projected to grow 2x year-over-year in 2026 to $3 billion.

These twin bets on subscriptions and advertising assume its content and inventory remain valuable to both consumers and advertisers. But the market dynamics it describes in that same section point toward a collision. Advertisers are increasingly turning to AI storytellers to create stories using their brands. It is their own version of premium content, and their natural monetization model is advertising. The implication is that the programmatic pipes for ad buying that Netflix is building are the root of both its future business model and its future competition.

Past essays related to today’s analysis:

Read more

Read more

Nanonoble is the sole operator of the version of Hailuo AI available to users in the United States. The lawsuit alleges that MiniMax, Nanonoble and SXJT disregarded U.S. copyright law by treating the studios’ copyrighted characters as if those were their own.

A motion to dismiss is, at its simplest, an argument for an outcome. As a rule, outcomes are determinative and arguments are not. In this case, this argument makes a number of counterpoints that mirror themes in my recent essays:

1. Cox Communications v. Sony Music Entertainment told IP holders that the legal system is the wrong venue for solving this problem at scale and that is the overall gist of the filing;

2. The 52 infringing videos were created by the studios themselves and not by any creators, which fails the Cox Commc’ns, Inc. v. Sony Music Ent. test of a service provider being secondarily liable

3. The studios may own popular IP like “Star Wars” but their complaint fails to explain why their copyright registrations also cover the characters within them—suggesting that there are “no sacred cows”;

4. U.S. copyright law does not reach conduct that occurred entirely abroad, suggesting Hailuo AI is one of those “thousands of tools” like Sora but which cannot be sued.

More importantly, the motion to dismiss also confirms the market coordination failure I highlighted last month with my Beer Marketplace paradigm: MiniMax and NanoNoble are arguing that they are neither able nor incentivized to police copyright the way the studios want both within and outside the U.S. The law is not the solution here.

Past essays related to today’s analysis:

Read more

The market reaction both in the media and in my recent conversations with executives is one of surprise— the writers folded easily on the issue of AI after demanding strict terms on its usage only three years ago. All eyes now turn to the AMPTP’s negotiations with the Screen Actor’s Guild (SAG), which will resume on April 27th. SAG has proposed a “Tilly Tax”, a fee named after the “first” AI actor in Hollywood—who I wrote about last October—which studios would have to pay to the union in exchange for using an AI actor.

No one expects SAG to fold as easily around the “Tilly Tax”. But there are lessons to be learned from Bollywood, where Eros Media World recut a 2013 hit movie “Raanjhanaa” over the lead actor’s public objection.

Past essays related to today’s analysis:

Read more

Google has previously estimated that there will be 1.2 billion new users who gain access to the internet between 2022 and the end of 2026. Services like SpaceX’s Starlink and Taara—an Alphabet X Moonshot company that uses beams of light to deliver high-speed data links—are expanding internet access to rural areas globally, and particularly in India and Africa.

As I wrote last August, this generation will be “the first to access the internet via AI platforms like ChatGPT and Claude, and not via start pages like Google, Yahoo or web browsers.” Reuters recently reported one impact of this new generation’s emergence in India: Bollywood studios “deploying AI at a scale unseen elsewhere” and audiences willing to pay for their AI-generated and AI-edited content. Bollywood producers are iterating and shipping AI-produced content, and audience demand is absorbing the rapidly evolving supply.

India is an emerging AI media marketplace that is unburdened by the regulatory and economic constraints of the U.S. marketplace. It is also rapidly being reshaped by growing internet access. It offers an important market test of Hollywood’s fear and AI platforms’ hopes for the future of entertainment.

The 1.2 Billion

India now has 958 million Active Internet Users according to the Internet and Mobile Association of India (IAMAI). That reflects 8% year-over-year growth. 548 million active internet users in rural areas account for over 57% of total users. Those are growing at nearly four times the pace of urban India. 38% of India’s population—nearly 579 million people—are not yet active internet users.

Reuters depicts this emerging generation of Indian consumers as having media diets evolving away from theatrical and shifting to streaming. This has resulted in production budgets being cut. Now they consume “full-fledged AI-generated films”, movies dubbed in numerous languages using AI and older titles recut with AI to revive ticket sales at lagging theaters.

AI dubbing is particularly powerful in India, which has 22 official languages and hundreds of dialects. The technology substantially reduces the cost of serving submarkets of languages and dialects, which in turn creates new demand across those submarkets.

Aggressive Supply Side

Reuters also depicts a supply side aggressive on both costs and technology to meet this changing demand. A Bollywood director told Reuters: “In India, cinema isn’t about art. It’s purely business, so studios are going to use it to make mythologicals. Our audience is a sucker for it.”

The business logic favoring mythology has echoes in the AI-generated Revolutionary War series titled “On This Day ... 1776” from Hollywood director Darren Aronofsky and his AI-focused venture Primordial Soup—sponsored by Google, Salesforce and Time, and distributed on YouTube. On the 250th birthday of The United States of America, its founding mythology has never been more important.

They are proving there is a business in cheap AI-driven productions of American history guided by Hollywood production talent and without Hollywood acting talent (although George Washington often looks like Sean Penn). “On This Day ... 1776” is an independent venture structured to avoid paying expensive actors and production guilds, distributed on YouTube to bypass theatrical and streamer gatekeeping and sponsored to avoid relying on direct-to-consumer spending or streaming economics. It routes around Hollywood market constraints that are weaker or nonexistent in Bollywood.

Reuters noted that PVR Inox, India's largest cinema chain, is a partner in the experiment, not a casualty of it—AI recuts are being deployed to revive ticket sales at lagging theaters, inside the same window Hollywood is watching collapse (in 2025, location intelligence provider Kalibrate estimates average visit volumes declined by 15% year over year across major movie theater chains).

Technology firms are meeting the moment: Nvidia recently told an audience at India’s AI film fest that it is working to slash computing costs so that anyone can “create something substantial without putting a lot of money” into production.

This is structural alignment. Traditional storytellers are increasingly relying on AI and the technology firms are eager to deliver the infrastructure to help them. No one is waiting on anyone else to make the first move. Every constraint that paralyses the U.S. media and entertainment industry is missing in India. There is no SAG-AFTRA equivalent forcing consent for digital replicas, no copyright cases against AI platforms in a slow appeal process, no DMCA architecture to litigate around and no insurer refusing to underwrite AI risk.

The system does not need valuable IP. The myths are Indian, the audience is Indian, the distributors are Indian, and the only U.S. asset in the chain is the compute and model layer that the production companies rent by the hour.

A Hollywood Nightmare in Bollywood

India is the AI-native media marketplace Hollywood is fighting to prevent. It is evolving this fast because the 1.2 billion are not being converted from a prior media diet. The demand side is AI-native from first contact.

Their first internet experience is an AI platform, their first filmed entertainment is AI-generated mythology in their own dialect and their expectations are being set by what Bollywood ships this quarter. Supply and demand are forming together to create a new world of entertainment consumption—the only missing ingredient is Prosumer and Creator AI-generated content (see below). Google, Microsoft and Nvidia “partnering with local filmmakers” suggests that will emerge next.

The Hollywood nightmare is not that audiences will embrace AI-generated content. The Indian audience already answered that. The nightmare is that India is proving that an AI-first ecosystem does not need valuable U.S. IP, does not need U.S. consent law and does not need U.S. distribution. All it needs is U.S. tech infrastructure and the storytelling takes care of itself.

Bollywood production companies are already seeing the upside at the box office. Eros Media World recut a 2013 hit movie “Raanjhanaa” over the lead actor’s public objection and sold 35% of available Tamil-language tickets in the release month. Hollywood talent agents would have killed the recut before it reached production.

Past Essays Related to This Analysis:

Electricity followed this arc in the 19th and early 20th century. Bajarin and many other folks in my feeds and chat groups argue the build-out of today’s AI infrastructure now follows the same arc. Cynics counter that the inflated valuations of AI companies are a bubble like crypto, which also arrived as a buildout of “hardware and headlines” to replace traditional finance. 20 years into that marketplace, there are only $20 billion in onchain fees.1

The clearest signals from the advertising marketplace that “this time is different” with AI. Last week I argued that brands leveraging AI to build their own entertainment infrastructure will drive ad dollars out of the traditional advertising value chain altogether. I cited an argument that the barrier is lower for brands not to just become creators but to “develop original IP, put it in front of an audience, learn in real time, and iterate — the same way a tech company ships and improves a product — rather than betting everything on one expensive, slow-moving campaign.”

Past essays related to today’s analysis:

Read more

TBPN had captured a small but valuable audience—70,000 viewers per episode across three-hour weekday shows on various online platforms.. The Wall Street Journal reported it earned around $5 million in revenue from advertising last year and was on track to make more than $30 million in revenue in 2026.

OpenAI’s move is analogous to its short-lived business plan to integrate Sora into the ChatGPT app last month—putting TBPN video one-click away from a global audience of 920 million weekly active ChatGPT users will help the show to grow and OpenAI to drive the conversation around its brand.

TBPN founders were rewarded with an extraordinary sales price in “the low hundreds of millions”, according to The Financial Times. In exchange they gave up advertising sales, sponsorships from third parties and editorial independence.

Something I previously wrote about OpenAI’s move to integrate Sora into ChatGPT applies here, too: “OpenAI’s move confirms that storytelling creates moments but infrastructure captures value.”

OpenAI is defining those moments as both scoops and “idea-driven conversations”. It argues where those conversations happen and who drives them “matters”.

This point also echoes something I wrote on Wednesday.

Past essays related to today’s analysis:

Read more

Three developments from this past month reflect a weird moment for the $1 trillion advertising marketplace I wrote about in January.

First, the video market has become a zero-sum game with a shrinking ceiling. Streaming is pushing more consumers to cheaper, ad-supported tiers but only recapturing two-thirds of ad dollars being lost from linear, according to research firm MoffettNathanson. The rest of those ad dollars are leaking to YouTube and other digital platforms.

Second, Variety and The Wall Street Journal both reported that Nielsen’s popular “The Gauge” may be providing an inaccurate lens on the video marketplace. Nielsen had previously drawn estimates entirely from its own volunteer panels. In recent months, it began warning clients that “its broadcast and cable TV figures could see a boost in February after it began using a study from the industry’s Advertising Research Foundation to inform its estimates of the demographic groups in U.S. households and the technologies they use to watch TV.”

In response, Nielsen has yet to release The Gauge for February 2026. Streamers are pressuring Nielsen to delay the report because of this ARF methodology. The annual upfronts and newfronts events—where broadcasters and streaming platforms pitch upcoming inventory to marketers and ad buyers—start soon, and the industry does not have reliable data on where audiences actually consume content.

Last, Jason Zada—founder of entertainment studio Secret Level—argued in a recent essay that “the old playbook” of “interrupt, target, repeat” for TV and streaming advertisers is “getting harder and more expensive to run every year.” This model is also “always late”—”The gap between ‘this is culturally relevant right now’ and ‘our campaign is ready’ has always been the dirty secret of the industry.”

The message from all three different market signals: The traditional media value chain—produce content, attract audiences, sell advertising against those audiences, measure with Nielsen—is collapsing at each node.

Back to the $1 Trillion Ad Marketplace

In January I argued that the new AI advertising infrastructure being built—scene-level targeting, AI production tools, programmatic efficiency—”extracts value from storytelling without distributing value to storytellers.”

I added:

“The differentiation for monetizing storytelling is not quality—it is operating capabilities. For fat tail creators, that means infrastructure to monetize the moments algorithms surface. For successful mid-tail creators like TBPN, it means infrastructure to monetize without algorithmic ad targeting at all.”

Zada argues that in this market, the barrier is lower for brands not to just become creators but to “develop original IP, put it in front of an audience, learn in real time, and iterate — the same way a tech company ships and improves a product — rather than betting everything on one expensive, slow-moving campaign.”

The infrastructure of monetizing moments that algorithms surface favors this model: “The fear of the big swing gets a lot smaller when you can take fifty fast swings first, each one more informed than the last.” Agencies like Secret Level and AI storytellers working for brands are the ideal partners for this model.

Two Questions for Upfronts & Newfronts

This leaves two open questions for upfronts and newfronts.

First, what will happen to legacy media advertising spend?

Both Linear TV (-3.3%) and broadcast (-2.5%) Upfront spending fell last year, while streaming surged (13.9%). Whatever the revised Gauge says about streaming, YouTube is positioned to capture more. Measurement is broken. The upfronts are becoming less relevant—not only because key Nielsen data is unreliable but because there is an emerging alternative that does not need the sellers.

Second, if brands reduce spend with linear and streaming, will those dollars shift towards the creation of original IP?

If brands follow Zada’s model and build their own entertainment infrastructure, the dollars will not shift from linear to streaming or from streaming to YouTube. Instead, they will leave the traditional advertising value chain altogether. The upfronts are selling inventory inside a system that brands are beginning to build around.

A Hidden Catalyst

Zada’s proposal also connects to something I wrote in February:

“A Cambrian explosion of creators outside Hollywood may be coming. But the path to building and monetizing original IP requires physical infrastructure, legal literacy or an audience at scale. Most AI storytellers stand at the fork possessing none of these.

The successful path with brands has a lower bar—the skill set to orchestrate creative across multiple tools and a few days of schedule bandwidth.”

The implication is that Nielsen’s failure to measure audiences accurately is creating the conditions for brands to stop relying on third parties to create the moments to connect with those audiences. This will be a catalyst for generative AI creators like Zada to reap the gains.

Past Essays Related to This Analysis:

Justice Clarence Thomas wrote the opinion, ruling that suing ISPs is not the path forward for copyright enforcement on the Internet—a decision with obvious implications for future disputes between copyright holders and AI companies.

For every major IP holder in Hollywood except one, this ruling is a problem. For the Ellisons, it is confirmation of their bet on cloud infrastructure.

Past essays related to today’s analysis:

Read more

My go-to example of misalignment from the last six years has been the lack of direct-to-consumer, e-commerce and computer science skillsets in the C-suite. Each instance has undermined their expensive bets on streaming. Disney’s multiple failed bets on digital media over the past two decades are frequent examples. Another is Warner Bros. Discovery’s Q3 2023 earnings call, where management promised growth from gaming and streaming and Wall Street analysts responded by asking whether management’s TV skillsets could deliver on those promises.

Generative AI is rapidly creating new misalignments. I am using the global beer market as an analogy for identifying and explaining these various misalignments because the emerging AI content marketplace has the same basic structure:

• Oligopolies (AB InBev, Heineken) = Hyperscalers (OpenAI, Google) with large foundational models

• Craft breweries = Open-source / Niche AI model makers

• Distributors = Cloud platforms

• Homebrewers and enthusiasts = Prosumers and creators

The beer industry solved its coordination problems nearly 100 years ago through established regulations, industry associations and standard insurance products. The AI content market is emerging with few clear guidelines from intellectual property law, and that problem differs depending on which country you are in.

AI content deals and strategic decision-making are slow and uncertain without those guidelines. It is not just executives at companies like Disney who struggle to optimize for what the market moment requires—and this week’s sudden and spectacular failure of its bet on OpenAI was misalignment in its purest form.

Here are three versions of misalignment I have highlighted, each reframed to pinpoint where the misalignment exists.

Past Essays Related to This Analysis:

Read more

Disney’s $1 billion investment in OpenAI just unraveled in spectacular fashion, only three months after its signature and seven business days after former CEO Robert Iger stepped down. OpenAI announced it will discontinue Sora in the consumer app and API, and Disney announced it will seek other platforms as partners.

This outcome confirms my original concern from last November about Disney’s partnering with OpenAI instead of a company that owns a cloud solution. Disney management “believes it can have market leverage at the consumer-facing level. But, in the emerging marketplace, IP only has value to consumers when there is leverage at the cloud level.” The cloud is where AI models are trained on that IP, where derivative content is generated from it and where third-party distribution channels can pull the content.

Paramount Skydance and Oracle are quickly moving to build the cloud infrastructure that creates leverage to protect its IP in the AI era. Disney lacks similar available solutions. Instead, it has fragmented partnerships (OpenAI for Sora, AWS for cloud, Google for generative AI) that prevent the end-to-end control which Paramount Skydance achieves through the Ellisons’ control of both companies.

All coverage of this news has missed a crucial additional detail in the story: Disney sent Google a cease and desist letter on the same day as its OpenAI partnership announcement. Iger told CNBC “[W]e’ve been aggressive at protecting our IP, and we’ve gone after other companies that have not honored our IP, not respected our IP, not valued it. And this is another example of us doing just that.”

The message to the market was clear: Google was a bad actor and OpenAI was a good actor. Because OpenAI was a good actor, Disney shareholders would see upside in both operational cost-savings from the deal and value in Disney’s share of OpenAI’s growing market value. The message was re-emphasized with its recent cease and desist letter to Bytedance, the owner of the viral generative AI platform Seedance.

Google and Bytedance are now Disney’s only remaining “dance partners” in the hyperscaler world. Working with either or both will require walking back legal action and an agreement for cloud training. There is also Midjourney, which Disney sued alongside Universal last June.

Its other choices include smaller open source and niche platforms—I have previously highlighted Fable Simulation’s Showrunner and AiMation as examples. Runway seems to be off the table, recently raising $315 million in Series E funding to pivot away from Hollywood and “pre-train the next generation of world models and bring them to new products and industries.”

Whatever it does, its choices are more limited than before because it has opted for litigation over a cloud training partnership strategy.

Misalignment

The diagram above maps the AI content market onto the structure of the global beer industry. To remind you:

• Oligopolies (AB InBev, Heineken) = Hyperscalers (OpenAI, Google) with large foundational models

• Craft breweries = Open-source / Niche AI model makers

• Distributors = Cloud platforms

• Homebrewers and enthusiasts = Prosumers and creators

I previously wrote that the beer market analogy is valuable because it shows how oligopolies adapt when enthusiasts get access to quality equipment.

That same dynamic is true in AI but at 1,000x the scale—there are 10,000 craft breweries worldwide but over 20 million Midjourney users. That means both creators and smaller platforms have an outsized impact in the AI marketplace relative to homebrewers and enthusiasts in the beer marketplace.

The key difference between the two marketplaces is that post-Prohibition-Era state laws prevent beer companies from owning distribution. That means a brewery can make the best product in the market and still fail if it does not have distribution. But in the AI marketplace, distribution can come in the form of either a streaming platform and a cloud solution where models are trained, content is generated and derivative works emerge that can be distributed wherever the cloud solution permits.

So Google can own Gemini, Google Cloud and YouTube, and the Ellisons can control Paramount, Oracle Cloud and Paramount+. Owning the distribution creates leverage over every platform—niche, open source or hyperscaler—that needs cloud infrastructure to train models and generate content using owned and original IP.

I argued last November why Disney should focus on partnering with companies in the Distribution part of the marketplace instead of the producers:

“Disney should focus on finding its Oracle equivalent—a cloud partner with the aligned incentives that come from shared ownership (like the Ellisons have with Paramount and Oracle), providing end-to-end integration from IP licensing through infrastructure to AI tools. Because leverage in the cloud translates into leverage at both the levels of generative AI tools—both from hyperscalers and open source and niche tools—and creators and prosumers using generative AI apps.”

That partner is harder to find now because Disney spent six months signaling that litigation was its preferred strategy. Disney had also relied on OpenAI to help it tell a story of AI driving operational cost-efficiencies. That story is lost.

Disney now finds itself fundamentally misaligned with both Wall Street and the broader AI marketplace. It is an unfortunate but poetic ending to Iger’s tenure—beginning with the video iPod, he found success in pivoting Disney’s brand to digital media by using IP as leverage over new technologies.

This strategy became “long in the tooth” after Disney’s Maker Studios acquisition failed in 2016. The OpenAI failure bookends the arc of that strategy and leaves an unfortunate mess for new CEO Josh D’Amaro.

Past Essays Related to This Analysis:

If it could only be that simple. He is understating the consequences.

I previously argued the Ellison family is pursuing WBD because it is the only deal in the marketplace that connects two large libraries of content to cloud infrastructure—”more IP delivers more business opportunities within the cloud.” The long-term business objective of the merger is to control and monetize all possible use cases with owned IP in a world where the “new medium” of generative AI will become the norm.

If the deal does not close, WBD faces a variety of problems around market timing. AI videos from Bytedance’s Seedance and Google’s Veo 3 are growing in popularity across Bytedance-owned TikTok, Google-owned YouTube and other digital media platforms. “Pirated” Hollywood IP and likenesses are showing up in more of these videos.

Past essays related to today’s analysis:

Read more

Today, Josh D’Amaro officially takes the reins as Disney’s CEO. There are many takes on Bob Iger’s legacy. Here is one I have not seen.

Iger’s expensive bet on streaming and expensive $71B acquisition of 21st Century Fox left a giant database of Disney + 21st Century Fox content at the heart of Walt Disney’s famed flywheel. That decision is playing out in the reorganization under Disney Entertainment announced on Monday. Streaming, games and digital entertainment now directly report into Dana Walden, Disney’s new President and Chief Creative Officer.

Substantively, this move means that two different databases now drive Disney’s business—one sits in Entertainment under Walden, and the other sits in Experiences under D’Amaro’s successor, Thomas Mazloum.1

Iger’s unintended legacy is that the Entertainment database now sits at the heart of the flywheel. But the appointment of Walden—a longtime TV executive with strong ties to Hollywood’s creative community—is intended to signal that human creativity still sits at the heart of Disney’s flywheel. The long-held belief is that “creativity will drive the profits”, as former CEO Michael Eisner posted on X upon D’Amaro’s appointment.

An Inherent Contradiction

This conflict beneath the reorg reflects an inherent contradiction within their pivot to streaming. As I wrote last month, “streaming is, at its core, a database business.” In streaming, content delivered by Netflix serves technical problems like retention curves, churn reduction and ARPU optimization. It serves the distribution software and the core metrics that drive the economics of that software.

Former CEO Bob Chapek understood this dynamic. Before his ouster, he told investors that the customer relationship—not content—is the engine for direct-to-consumer businesses. Theme parks were Disney’s “original direct-to-consumer business” and streaming was “our newest direct-to-consumer business.” Disney’s future lay in a data-driven model operating across the two databases. The closest analogue was Amazon Prime, putting customer satisfaction first and content second. But that operates upon one giant customer database.

This reorg explicitly refuses to acknowledge those synergies. At the very least, it implies those synergies are secondary to political control over each database. Synergies are a negotiation, not a central corporate priority.

From a financial perspective, the Experiences database is more important to the flywheel—it drove 57% of operating income ($10 billion of $17.6 billion) in FY 2025. But, given the corporate adage “creativity will drive the profits”, the Entertainment database is more important.

So which database is at the heart of the flywheel? If it is actually both, no one has spelled out how that will work.

In a black box world, there is one database to rule them all. In the notoriously ugly internal politics of Disney, the CEO who envisioned the single database is out via a putsch. The new generation of leaders is being territorial around their databases.

It is not clear what lies at the heart of the flywheel now. All we know is that it is not creativity alone driving it anymore.

Iger’s “Monolith”

In season 7 of AMC’s hit show “Mad Men”—Episode 4: “The Monolith”— Sterling, Cooper & Partners’ (SC&P) acquire a room-sized IBM System/360 mainframe computer. It is placed in a room that SC&P’s creatives had used as a lobby. The presence of the computer and its loud humming drives one of the creatives—Michael Ginsberg—to have a mental breakdown and the rest of the creative team to resent it.

Iger’s legacy is a “monolith” of a database at the heart of where human creativity once resided. It does not “hum” because its databases of content are stored across Amazon Web Services (AWS) servers. No creatives were impacted by the installation of Iger’s “monolith”.

But it does pose an existential challenge to human creativity at Disney. The deeper challenge presented by Iger’s “monolith”—one reflected in the reorg—is that it embodies a fundamentally different corporate relationship with the consumer. The most successful flywheel models in digital media put customer satisfaction first and content second. Disney’s Experiences division already operates this way.

This reorganization leaves Disney with organizational, procedural and technological ambiguities around what drives the flywheel after creativity has been supplanted by a database.

D’Amaro’s appointment suggests that Experiences—once a beneficiary of creativity at the center of the flywheel—will now drive the flywheel itself. This reorg leaves each division’s role in the famed flywheel ambiguous. Perhaps it is D’Amaro’s objective to resolve those conflicts. But his former Parks boss Chapek sought to do the same, and he was ousted.

That leaves D’Amaro, like Chapek, starting his new job in tricky political territory with internal stakeholders and external shareholders.

“Antibodies in the System”

I argued in February that “Chapek was right about the business and wrong about the politics.” He misread Hollywood and he misread the willingness of Disney’s corporate culture to embrace the culture change catalyzed by Iger’s bold pivot into streaming. He learned there are “antibodies in the system”—a phrase I read in the book “Unit X” about the Pentagon’s attempts to accelerate the U.S. military’s adoption of Silicon Valley innovations.

In the book, “antibodies” describes Pentagon bureaucrats and congressional staffers who were slow to cooperate, prioritizing their own privileges and entrenched processes, and working to block or kill the progress of the Pentagon’s “Unit X” program.

Iger wrote in his autobiography “The Ride of A Lifetime” that he believed Disney would not have been able to pivot to streaming and survive the acceleration of cord-cutting trends without a realignment of managerial incentives—effectively, fighting the antibodies. That pivot to Disney+ required those managers to focus on the “new thing” instead of the businesses at which they had been successful, to date. It required “the purposeful erosion and disruption of their businesses” and the incentives offered needed to match the assumption of that risk.

Despite all that work, “antibodies” in Disney’s system later took down Chapek.

D’Amaro needs to simplify the role of the database in Disney’s flywheel. The new reality is that creativity has become a commodity in Disney’s distribution model. Marginal value from creativity must be extracted elsewhere across the flywheel. It is operationally and technologically more complex and inefficient to pursue that model across multiple databases.

In the Attention Wars, the consumer has fundamentally more power over Disney than they ever had when Walt Disney was alive. They now navigate exponentially more digital and real-world alternatives for their money, time and attention.

Bet on D’Amaro

My bet is that D’Amaro will figure it out. He has to. But the past is precedent and it will not be pretty. My guess is that at some point over the next 12 months, D’Amaro will rip the band-aid off Iger’s legacy and start building anew.

That means:

1. Unify the two databases under a single customer-data architecture,

2. Subordinate content decisions to retention metrics; and,

3. Streamline the fiefdoms of Entertainment and Experiences.

Last November, I wrote that Disney management “believes it can have market leverage at the consumer-facing level. But, in the emerging marketplace, IP only has value to consumers when there is leverage at the cloud level.”

The market’s message is that Disney collectively needs to honor and listen to the “monolith” now at the center of its Entertainment division. That means a strategic investment from a cloud company to wall off Disney’s IP at the infrastructure level and charge AI models for training access.

Licensing alone will not drive the flywheel. Infrastructure ownership will.

Past essays related to today’s analysis:

Yesterday, The Information reported that ByteDance has delayed the global launch of Seedance 2.0 due to “a string of copyright disputes with major Hollywood studios and streaming platforms”. ByteDance had previously assured the market of IP holders that it is “taking steps to strengthen current safeguards as we work to prevent the unauthorised use of intellectual property and likeness by users.”

Those assurances do not yet appear to be enough. Its legal team is “working to identify and resolve any potential legal problems”, and its engineers are “putting in guardrails to prevent the model from spitting out content that could cause further intellectual property breaches.”

Both this story and OpenAI’s decision to integrate Sora into ChatGPT—a move which would launch Disney likenesses to an exponentially larger audience—imply that the supply of legacy media IP is strategically important to the international growth of hyperscaler platforms. If Seedance’s launch is precedent, “fanfiction” will become the dominant use case—so legacy media must pre-empt a future of anyone with access to a hyperscaler platform to use their IP freely and without parameters.

Past essays related to today’s analysis:

Read more

This essay is free for all subscribers]

This morning The Information reported that “OpenAI plans to soon launch its Sora AI-generated video capabilities in ChatGPT.”

ChatGPT has over 920 million weekly active users, whereas Sora has 20.7M lifetime downloads and flat growth—only 2M additional downloads in roughly two months— since its launch last October, according to recent data from research firm AppMagic. That said, it reportedly will continue to operate the Sora standalone app, though its long-term plans for the app are “not clear”.

Now one-click away from a global audience of ChatGPT users, Sora will inevitably drive greater usage than it has seen, to date. Google’s data for usage of its video generation platform Veo—230 million videos have been generated with Veo 3, up 600% since June 2025—suggests the business rationale: OpenAI needs to expand distribution for Sora both because the app needs it and ChatGPT needs to compete with Gemini.

Will it scale? According to the AppMagic Mobile Market Landscape 2026 report I analyzed last week, downloads of chatbots like ChatGPT are around 4x as large as AI art generators and AI music generators. Both this data and Sora’s download history (below) suggest consumer demand for video generation may underwhelm relative to ChatGPT’s scale.

This confirms the IP bifurcation thesis I outlined in November. IP behind hyperscaler paywalls (e.g., ChatGPT, Google Gemini) retains value because licensing and generation can be controlled. IP distributed through cable syndication, free streaming platforms, or social media clips loses value because it becomes training data for open-source models.

Sora as an independent app offered a narrow pathway into OpenAI’s ecosystem. Its integration into ChatGPT pulls it deeper within that ecosystem and, like Veo 3 in Gemini, makes it core to the platform’s value proposition.

Good News for Disney?

On its surface, this appears to be a good development for Disney’s $1 billion deal with OpenAI which licensed 200-plus characters to OpenAI under a three-year agreement. The licensing terms apply to the Sora platform specifically. Embedding Sora in ChatGPT exponentially expands the existing audience and potential audience. Disney appears positioned to have more user-generated content to choose from for Disney+. OpenAI will have more content it will be able to monetize as it expands its advertising model.

That said, Sora’s integration also raises the question whether the constrained, Disney-approved use cases for its IP will generate enough engagement on ChatGPT to justify the partnership. Hilary Mason—Co-founder and CEO of generative AI roleplaying platform Hidden Door—predicts “Disney characters on Sora will be either so locked down as to be boring or will feature in humiliating, offensive, and pornographic content within hours, or maybe both.” The odds of those types of content emerging seemed smaller when Sora was smaller. Now, with 920 million users getting access to Sora, they seem much higher.

The other challenge is whether Sora within ChatGPT will compete with popular generative AI platforms Seedance and Veo 3. Beyond Disney’s IP, Sora has not demonstrated a compelling advantage over Seedance or Veo 3 for the creators—professional and amateur—who are actually making generative AI content. As I argued before, they may prefer celebrity likenesses to Disney IP.

That leaves both Disney and OpenAI facing some new questions about their partnership. If demand for the approved use cases is thin, Disney’s $1 billion will have bought positioning on a platform where its IP will be underutilized. If demand is strong, Disney will prove that licensing behind a hyperscaler paywall works—but only within that wall. Outside, content like “Rasta Wars” and the Stormtrooper vlogs suggest that creators are happy with Veo 3 and open source models.

Given these videos and the explosion of Seedance-produced videos, OpenAI and Disney will have to renegotiate or kill the Sora elements of their partnership. Exiting CEO Robert Iger made the deal. New CEO Josh D’Amaro inherits the consequences. How he handles them will be one of the closely watched decisions of his early tenure.

Market Coordination Failure

At 20 million users and low usage, Sora was a manageable threat to studios. At 920 million, that is no longer possible. Disney’s only options are to enforce selectively and hope the volume stays low, or to accept that its IP will circulate uncontrolled inside the largest consumer platform in the world. Neither involves outcomes that Disney will embrace. In the second instance, Disney could issue takedown notices while simultaneously benefiting from unauthorized use as earned media. However, if any version of Mason’s predictions proves to be accurate, D’Amaro will opt to kill the Sora elements of the partnership.

Disney is not alone in this bind. Every studio, insurer and investor described in my prior essay now faces Sora’s larger scale—and the same absence of shared frameworks to act on it.

Whatever the outcome, OpenAI’s move confirms that storytelling creates moments but infrastructure captures value. The scale of storytelling just changed within OpenAI. The question now is how the rest of the market will adapt if ChatGPT users do not flock to Sora. They may not. The chatbot still wins.

Past essays related to today’s analysis:

Streaming marked a fundamental shift in the value of movies and TV shows. A monthly fee made content available at a fraction of the cost of cable or DVDs. Amazon Prime equated watching Hollywood movies to free deliveries of paper towels. The competitive advantage of owning a giant library eroded. User data—an understanding of what the consumer needed to watch in the moment—became more valuable than the content itself.

The promised solution—an owned-and-operated streaming service—is barely profitable at 5% net margins for legacy media companies. A service requires increasingly sophisticated technology to compete that their executives lack the skillsets and experience to identify and integrate.

Generative AI threatens to finish what streaming started. Streaming killed pay windows and compressed the value of expensive productions to one title among thousands in a monthly subscription.

AI is compressing the value of IP itself. A comedy writer Heather Ann Campbell explained it plainly on Bluesky: “All of these people who have access to the latest AI visualisation engines like Seedance they’re being given total control to create anything they can imagine - and they’re turning out fanfiction.”

A Seedance clip of Brad Pitt fighting Tom Cruise resulted in Hollywood studios and the Motion Picture Association (MPA) sending cease-and-desist letters to ByteDance. Disney partnered with OpenAI and sent Google cease-and-desist letters. The lesson is the same in both cases: The technology stack of an IP owner must now be sophisticated enough to deliver a profitable service consumers need and to regulate how AI training models access their IP libraries. Open-source AI tools are proliferating—most beyond U.S. and EU enforcement reach—and are replicating Hollywood IP with zero consequences.

At the same time, a new generation of AI creators is already building original IP. But owning and licensing that work requires human authorship that any use of generative AI tools may not satisfy, with the current exceptions of motion capture and 3-D rendering. Licensing agreements restrict behavior on specific platforms. They do nothing about the billions of users in jurisdictions with weaker IP regulations where enforcement is impractical.

The best AI creators are earning well by producing branded and agency content—think early radio and TV sponsorships—but for their own content, there are few settled answers for basic questions: Do I own this? Can I license it? What happens if someone claims the training data was stolen?

The risks are real—training data provenance, likeness usage, conflicting tool licenses—but lawyers are limited to analogies from traditional media that are partially applicable, at best. On the other side, institutional buyers such as streamers and studios operate under intense reputational and litigation risk. Internal Business Affairs teams currently adopt blanket prohibitions that kill promising AI-native deals without systematically assessing the actual exposure. Insurers cannot price AI-heavy productions, so they exclude them entirely.

No coverage means no closed deals. Investors cannot model the long-term defensibility of AI-native IP, so capital cannot deploy where paperwork does not exist. The Supreme Court recently declined to hear the only generative AI copyright case on its docket. Other cases are currently working their way through the appeals system.

The problem extends beyond Hollywood. Over 90% of U.S. ad agencies are using or experimenting with generative AI, but legal and IP concerns remain among the main barriers to scaled deployment. Eighty percent of brands have concerns about agency use of generative AI and are revisiting contracts, governance and guardrails. More than half of surveyed audiences say they are uncomfortable with AI-generated brand content. The coordination failure runs the entire value chain from the person who makes the content to the person who watches it.

Nobody—excluding pirates—is acting in bad faith. Each actor is making the correct decision for their own position. A creator who cannot prove ownership is right to take the brand deal. A studio that cannot evaluate risk is right to reject the submission. An insurer that cannot model exposure is right to exclude the policy. An investor who cannot model defensibility is right to pass on the deal. A court that sees no urgency is right to wait.

Economists call this a market coordination failure. It is not a disagreement. It is a system where every participant needs someone else to move first and no one has the incentive to be that someone. A creator cannot invest in building a franchise until they know the IP is defensible. But the IP is not defensible until courts or regulators establish standards. Courts will not establish standards until better cases arrive. Better cases will not arrive until creators invest in building franchises. The loop has no starting point.

This is why analyzing any single element in isolation produces the wrong answer. A studio executive who studies only copyright law will miss that insurers are excluding AI productions entirely. An investor who studies only deal structures will miss that the underlying IP may not survive a platform terms-of-service change. An insurer who studies only loss exposure will miss that creators are generating the signal data needed to price the risk. Every constituency holds a piece of the picture. No constituency can see the whole board.

The sum of those rational decisions is paralysis. Billions of dollars in potential value sit trapped behind uncertainty rather than actual legal impossibility. That is not a content problem or a technology problem. It is a coordination failure.

What Am I Going to Do About It

Past essays related to today’s analysis:

Read more

Both are free for all subscribers.]

Last month AppMagic released its Mobile Market Landscape Report.

They shed new light on the three themes from the last Monthly Series with AppMagic data:

1. The dropoff in downloads of OpenAI’s Sora App

2. Netflix’s pivot in its mobile games strategy

3. The Rise of “Drama Slop” Apps

NOTE: A reminder that AppMagic only tracks in-app mobile revenue in the App Store and Google Play.

Generative AI Apps (Sora + Chatbot apps)

Key takeaway: User retention is falling, which means the AI platforms monetizing this attention are capturing value from a shrinking core of power users, not a broadening base.

AppMagic reported that the overall generative AI marketplace grew nearly 3x from 2024 to 2025 and is now a $3 billion marketplace. This marketplace includes Sora alongside OpenAI’s ChatGPT, Google’s Gemini and X’s Grok.

Revenue for these apps grew faster than downloads in 2025—273.6% to 178%. ChatGPT is estimated to have generated $2.33B on 990M downloads (+193% YoY) and Google Gemini hit 474M downloads (+381%).

However, user retention has decreased. AppMagic observed the largest drop in the first days of user life (–6.2%), while mid- and long-term retention decreased by 3.9%–4.5%.

This is all occurring against the backdrop of apps surpassing games in mobile revenue for the first time in September 2025. Generative AI platforms increasingly monetize attention through recurring revenue while the games selling discrete entertainment experiences fall behind.

This outcome reflects a market dynamic I identified back in January: When algorithms increasingly define which moments matter and platforms build scene-level precision to monetize them, how do individual storytellers capture value?

My answer: “In the Algorithmic Era, storytelling creates these moments, but infrastructure captures the value.” ChatGPT is infrastructure. Sora at 20.7M lifetime downloads and flat growth—only 2M additional downloads in roughly two months—is selling intermittent, interactive creative experiences in a market that rewards recurring subscriptions.

Netflix’s Pivot Away From Mobile Games

Key takeaway: AppMagic plus other data shows stalled growth in the mobile market. Netflix may have seen many opportunities four years ago, but its pivot suggests it sees fewer now.

The gaming market is reaching saturation, suggesting Netflix’s retreat from mobile games was impeccably timed.

Revenue grew just 0.2% YoY (down from 3% in 2024). RPG revenue dropped 16.6%. Strategy grew 16.1%, driven by 4X titles like “Last War: Survival” and “Whiteout Survival” where $99 offers generate roughly 30% of revenue. LATAM—which was the growth story a year ago—is now declining in downloads across Colombia, Peru, Ecuador and Argentina.

Matthew Ball’s Annual State of Video Gaming deck shed additional light on why the mobile gaming market is flat. It highlights broader declines in user engagement and revenues of games across demographics (slide one). It also points to declining hours played (in the U.S.) and livestreams of games globally.

Given these trendlines were emerging in 2022, why was Netflix so bullish on mobile games over the past four years?

The logic was sound. The internet is interactive. Netflix’s platform captures value from interactivity with media. If Netflix could capture additional interactivity cost-effectively, it would reduce churn and drive growth in certain markets, too.

But the global market stalled. Revenue growth collapsed from 3% to 0.2%.

Particular game formats are more popular than the ones Netflix has bet on. Netflix’s “Stranger Things: 1984” is a casual adventure game. AppMagic’s report shows downloads of the adventure format for hypercasual games were -30.7% year-over-year. We also see from AppMagic’s download data that downloads of “Stranger Things: 1984” were -29.7% from 2022.

It recently bet on the popular midcore format—meaning, playable in a relatively short period of time while retaining some complexity—with action-adventure title “Red Dead Redemption”. That should be a better performing bet, according to AppMagic data. Downloads of Adventure titles are up 7% between 2024 (163.3M) and 2025 (176M). Also, downloads of Action titles are effectively flat year-over-year at 2.2B (-1.4%).

However, since its release, downloads of the app suggest steadily waning interest. Even with smarter market bets, the data points to obvious disconnects in Netflix’s mobile business model—it bundles games into a subscription rather than selling them directly. At over 330 million members, it is hard to imagine how 10 million downloads–and slowing—are moving the needles on engagement and reduced churn.

This data both justifies its recent pivot and raises more questions.

The Rise of “Drama Slop” Apps

Key takeaway: Investors wonder how Netflix will defend itself against Drama Slop on iOS and Android after this report.

Short Drama aka “Drama Slop” as a category hit $1.8B in revenue (+93.5% YoY) on 2.1B downloads (+279.5%). The market grew nearly 4x in downloads in a single year. ReelShort and DramaBox were the top two grossing Short Drama apps of 2025, more than 2x the size of their third place competitor GoodShort.

The report also highlights how the segment is showing a “clear stability” phase after a 1,000%+ revenue spike in 2024. Monthly revenue now holds steady between $145M and $160M even in weak months. And monetization runs on 7-day subscriptions at roughly $18–$20, with over 70% of revenue coming from that single offer type.

The cost of a Standard monthly Netflix subscription is spent weekly on these apps. This data point strongly supports my argument from last July that the Drama Slop model has “increasingly more value to consumers than Netflix’s walled garden”.

ReelShort and DramaBox together pulled roughly $700M in 2025 from a format that did not exist three years ago. The challenge for Netflix is not only that this new format is competitive with the 27-year-old incumbent in such a short period of time, but that its weekly and monthly spend suggests consumers value the new format more. There is a meatier story to the rise of “Drama Slop” apps in the rankings than just their outperformance of Netflix on AppMagic’s download charts.

Investors have growing questions about the strategic direction of Netflix’s business, especially after its failed bid for Warner Bros. They will wonder how Netflix will defend itself against Drama Slop on iOS and Android after this report.

Final Takeaways

AppMagic's Mobile Market Landscape Report added sheds new light on three new and dynamic business cases of disruption in digital media:

• GenAI platforms capture recurring revenue despite falling retention

• The mobile gaming market and Netflix’s bets within it have stalled

• Drama Slop consumers spend in one week what they could spend on a Standard monthly Netflix subscription.

Past essays related to today’s analysis:

March 2, 2026 was a big day for the future of intellectual property. It may be the history major in me, but it marks one of those days of unusual coincidence.

First, the Supreme Court declined to hear Thaler v. Perlmutter—the only case on generative AI and copyright in its docket. The Court had already rejected Thaler previously. He wants to register a copyright for an artwork listing DABUS—an AI system he designed—as its author.

Thaler was testing whether human authorship was a necessary condition for a copyright to be issued. The U.S. Copyright Office wrote in 2023 that when authors use AI in the creative process, “what matters is the extent to which the human had creative control over the work’s expression and ‘actually formed’ the traditional elements of authorship.”

The ruling is a mixed result for those betting that a “Cambrian explosion of creativity, weirdness, new media, and new forms of expression and art” is “just around the corner.” Human authorship remains a necessary condition to register a copyright. But outside of motion capture and 3D imaging, other categories where copyright can be registered after using AI platforms are still being tested. The Supreme Court sees no need to rush into this market and establish the clarity that creators will need to monetize their work.

Second, Paramount management spoke to investors about its acquisition of Warner Brothers Discovery. “Consolidating Streaming Technology Stacks” was listed as the first objective for corporate synergies in its investor presentation.

Past essays related to today’s analysis:

Read more

The Ellisons won. Netflix lost more than people seem to realize. Warner Bros. Discovery management will leave with share prices at least 4x greater than where they were 10 months ago.

Acquire IP

I predicted that 2026 will be the “the year of MOAR content”—reflecting the popular internet slang for “humorously or emphatically demanding additional material”—to compete with YouTube.

Netflix’s sales pitch was that acquiring Warner Bros. would “give audiences more of what they love and help define the next century of storytelling.” They could monetize that IP across streaming and gaming. I argued:

The acquisition of Warner Bros. would reduce [Netflix’s] licensing costs and give Netflix the IP resources to produce more content that its data shows audiences love on the services. It would also allow Netflix creatives across the world to imagine more localized versions of DC characters, something analogous to Marvel’s Indian version of Spider-Man (Pavitr Prabhakar) who appeared in Sony’s “Spider-Man: Across The Multiverse” (2023).

Paramount Skydance’s pitch was to become “the most technologically capable media company”. A “truly competitive” version of this company is “a combination of great content working with tech product hand in hand“.

The less obvious rationale is that the Ellisons control both Oracle and Paramount, enabling end-to-end IP control between the studio and its cloud provider. That is a valuable technological defense mechanism as generative AI platforms like Seedance threaten to drive IP value toward zero entirely.

Don’t Lose IP

Netflix needed to bid because WBD’s library—Batman, Harry Potter, HBO’s prestige catalog—represents some of the last remaining premium IP at scale.

In Netflix’s most recent What We Watched report for the first half of 2025, Warner Bros. and HBO IP are strong performers. “Sex and the City: Season 1” was in the top 7.5% of shows with 19.2 million views. Its second season, “Band of Brothers” (25.6 million) and “The Pacific” (20.1 million) were all in the top 14% of shows watched. The rest of the licensed HBO shows fell somewhere in the middle of the tail, around 40% of views.

Netflix has no obvious library alternatives that could perform as well as WBD IP. A bid for WBD was acting in the best interests of shareholders, even in defeat.

Monetizing IP Beyond TV/Film

Warner Bros. Games—home to major studios like TT Games, Rocksteady, and NetherRealm—merited only two mentions in the acquisition pitch deck to investors.

Its most successful title is “Hogwarts Legacy”, released in 2023 and developed by its Avalanche Games studio under its Portkey Games label. To date, sales are 34 million and revenue has crossed the $1 billion mark.

Paramount Skydance’s gaming operation is thin by comparison. Skydance Interactive focuses almost entirely on VR. Major Paramount titles like “Teenage Mutant Ninja Turtles: Mutants Unleashed” and recent “Paw Patrol” games are published through Outright Games, and not an in-house studio.

Management has said little about gaming ambitions beyond the broad promise of combining technology and content.

Licensing IP as Training Data

Controlling the cloud is an advantage in the AI era, where protecting IP from piracy will be an existential business. Open source and niche AI platforms enable piracy from Prosumers & Creators (see image above) in ways that are faster, cheaper and better than past forms of piracy.

If the objective is controlling and monetizing all possible use cases with owned IP, proprietary cloud infrastructure and Paramount content backed by Ellison patient capital will enable consumer-facing AI innovation, not just operational efficiency.

Paramount Skydance’s integrated solution with Oracle will allow it to work more easily with the likes of Seedance—to which it recently sent a cease-and-desist letter—and Google’s Veo, or smaller platforms like Showrunner or AiMation.

A New Era for Netflix

I argued that Netflix’s initial agreement with Warner Bros. Discovery may have been “strategic theater that both parties expect to fail.”This outcome would be “brilliant fiduciary duty engineering by both parties”: “WBD locks in the narrative that it pursued maximum value and Netflix locks in the narrative that it fought hard to solve content scarcity.”

Now, having lost the fight to solve for content scarcity, it seems more vulnerable than it ever has. Its next best defensive story about its vulnerable content library is its bets on YouTube creators and Spotify podcasts. Those bring premium podcast content from proven creators. But they do not bring “the architecture that makes video podcasts work on YouTube—the comments, the community, the many-to-many interaction.”

That means they no longer have “a combination of great content working with tech product hand in hand” at a time when the Ellisons are seeking to rethink and rebuild that model using its cloud infrastructure.

That all said, they have bought themselves time—both Warner Bros. Discovery and Paramount Skydance will be locked up in a regulatory review process over the next 10 to 12 months—and a $2.8 billion breakup fee to be paid by Warner Bros. Discovery.

They still need to answer the question of where their business is headed with less content and no games strategy to drive engagement. Video podcasts seem like a tactical fix. What is the strategy?

My guess as to its next big move: A strategic investment from—or partnership with—a cloud infrastructure company to help it defend its IP. The volume of cease-and-desist orders it will need to issue will only increase exponentially. Lawyers are expensive. Cloud solutions are cheaper. Seedance has proven that fans love Netflix IP. Why not use the cloud to enable and monetize more of that engagement—on Netflix’s terms?

Past essays related to today’s analysis: