Good afternoon!
The Medium delivers in-depth analyses of the media marketplace’s transformation as creators, tech companies and 10 million emerging advertisers revolutionize the business models for “premium content”.
On Tuesday, Disney CEO Robert Iger conceded to the 2024 Morgan Stanley Technology, Media & Telecom Conference, “We need to be at Netflix’s level in terms of technology capability.”
The diagnosis is right—as I argued three years ago—and the timing is poor. It has been five years since Iger publicly rejected the need for Netflix-type recommendation algorithms to The Wall Street Journal: “I think if people are clicking on Mickey Mouse, they mostly want Mickey Mouse.”
Mickey Mouse— specifically, Walt Disney’s 1928 “Steamboat Willie” version—is now in the public domain. Writing the prompt “drawing of Mickey” in the Stable Diffusion image generator, then following it with any description (e.g., “holding a sword”), will generate endless variations of Disney’s first animated Mickey Mouse (but which copyright law also forbids them to label “Mickey Mouse”).
In this use case, who needs Disney+ anymore to click on Mickey Mouse if they want Mickey Mouse? Isn’t it better to create and engage with a version of Mickey Mouse that delights them personally?
There is tension here between a subscription platform with billions invested behind it and a new, free platform with disruptive technology delivering a version of its most valuable IP. At the root of it lies a dynamic I described in last May’s "It's about the free money... and it's about the free money”:
The TV and movie industries — as we understand them today — were subsidized by “free money” from (1) the DVD and VHS rental market, (2) affiliate revenues from the linear TV business model and (3) zero interest rate policies (ZIRP) from central banks (which drove streaming).
That source of capital is steadily disappearing as cord-cutting accelerates and movie distribution evolves towards streaming.
Nature abhors a vacuum, and it is unclear what will replace “free money”.
So, if the past is precedent—and I believe it is—understanding which platform will ultimately win requires understanding what will be the next source of “free money”. One answer lies in venture capital, which funded Generative AI and AI-related startups with nearly $50 billion in 2023, according to Crunchbase. The other is the YouTube ecosystem, where YouTube is offering free AI tools to all creators.
Key Takeaway
The smartest bet in the media business has been to follow the "free money". The disruption of streaming subscription models by generative AI tools and YouTube's creator program has a lot in common with past disruptions with "free money" models.
Total words: 1,300
Total time reading: 5 minutes
Redefining "Creative Ability"
Last month, YouTube CEO Neal Mohan wrote in his annual letter: “Creators should be recognized as next-generation studios.” The major studios were disrupted by streaming and five years later, creators are disrupting streaming “with top-notch storytelling that can’t be dismissed as simply ‘user-generated content.’”
Mohan also wrote that YouTube is “leveling the playing field and developing AI tools that empower everyone.” Last September—with the help of creators including Colin and Samir—it introduced a suite of generative AI tools for creators to help them brainstorm ideas in YouTube Studio, find soundtracks for videos more easily in Creator Music and dub their videos into other languages with a tool called Aloud. A key part of its strategy is its Shorts format, which he shared is averaging over 70 billion daily views, and the number of channels uploading Shorts has grown 50% year over year.
As I wrote in last week’s Medium Shift column “The Media Revolution Will Be Prompted”, YouTube is emerging as the testing ground for the competition between videos fully generated by humans and those fully generated by AI. More than 3 million channels are in the YouTube Partner Program (YPP), so the math of content made partially or mostly with AI-generated tools seems infinite.
“A fair exchange” of value
Over the last three years, the YouTube Partner Program has paid out over $70 billion to creators, artists, and media companies. The program has paid out 55% of advertising revenues to creators and 45% to creators producing shorts, to date. It has enabled audiences to pay for merchandise (both creator-owned and third-party-owned products), for Super Chat (having messages highlighted for the creator), for Super Thanks (effectively, tips for content), and for channel memberships (exclusive perks and content for monthly paying members). Mohan announced YouTube is investing more in these initiatives and shared that the number of creators using memberships increased more than 50% last year.
Through the logic of “but for YouTube, these revenue channels for creators would not exist”, YouTube is effectively subsidizing these 3 million channels and all future ones with "free money". Audiences who compensate creators believe they are getting a “fair exchange of value” (more on this in this section) in exchange for the creative ability of YouTube creators.
Mohan is marketing YouTube's AI tools—constrained by YouTube’s existing policies and with new layers of transparency and protections—will both expand the pool of YouTube creators and enhance all creators’ creative abilities.
Diller & Disruption
This sales pitch, and the broader business logic of disruption emerging in this “weird” market moment, mirror a dynamic IAC Chairman Barry Diller described to Kara Swisher in an interview five years ago:
Amazon’s business model has nothing to do with anything anybody who’s been in the entertainment business has lived with their whole lives, which is, we have one job. We entertain the folks. If they like what we do, they buy a ticket or subscribe to this or get that, and then they get that in a fair exchange for their creative ability being essentially exchanged by somebody who says, ‘I’ll buy that thing.’ Amazon’s model has nothing to do with it.”
Diller’s core point was that the six major Hollywood movie studios were being disintermediated by streaming apps with alternative business models—whether direct-to-consumer (Netflix) or as a perk (Amazon). Streaming technology was a disruptive force, and the most successful streaming models completely subverted the value of both the jobs of entertainment executives and their end product. Netflix and Amazon had redefined the fair exchange of value and consequently discounted the marginal value of the product of an individual’s creative ability to nearly zero.
Five years later, Mohan's sales pitch is that generative AI tools will enhance both the marginal value of creative products on the platform and the fair exchange of value between the consumer and the creator.
Appetite
There is a market fear—some of which I discussed in “The Media Revolution Will Be Prompted”—that generative AI will reduce the individual’s contribution to zero. Without the individual involved, there will be no exchange of value.
That was the fear shared with Mohan last October by creators Colin Rosenberg and Samir Chaudry, aka Colin and Samir. YouTube audiences may develop an “appetite” for inauthentic, AI-generated storytelling at the expense of human-created content. It is implicitly a fear about how algorithms on ad-supported platforms like YouTube and Spotify tend to serve content that is both brand-safe and generates the most impressions for the platform. The more the algorithms favor purely AI-generated content, the fewer creators will receive compensation for their work.
Chaudry told me in an interview that he is less worried about audiences developing this “appetite” than he was a few months ago because “it is in our DNA to tell stories and to hear stories told” by human storytellers. He was confident that YouTube’s algorithm would continue to recognize and prioritize human-created content over “synthetically generated content”.
There is also an interesting question of “appetite” in the example of Mickey Mouse in the public domain, above. Both Chaudry and Jordi van den Bussche, aka Kwebbelkop, envisioned alternative platforms emerging that will enable users to create a video with a text prompt. These effectively would be future versions of OpenAI’s Sora, where “Mickey with a sword” could be animated by simply adding the verb “waving”.
Would that be "storytelling"? Would it be "inauthentic"?
We are a ways away from the “appetite” for a prompt-to-text video platform reaching 149 million subscribers, globally, like Disney+. The use case still needs to emerge and prove its worth for a monthly recurring subscription fee.
But, if/when it emerges, the most interesting question will where Disney's intellectual property should be: In the hands of creators or behind a Disney controlled paywall? Iger is betting on the latter, but YouTube is proving out that there are many more ways to monetize Disney IP by betting on creators with AI tools.
The smartest bet in the media business has been to follow the "free money", and that may be the better solution for Iger and other leaders worrying about the sophistication of their technology relative to Netflix's.