The Rise of AI and the End of Hollywood as We Know It

Imagine a world … where blockbusters—those multi-hundred-million-dollar spectacles that anchor movie studio slates—simply disappeared. No reunion of Marvel’s heroes for another universe-saving battle in Avengers: Doomsday in May 2026. No Vin Diesel pushing car stunts to new extremes the following June in Fast X: Part 2. No Christopher Nolan epic adaptation of the Odyssey with Matt Damon and Zendaya that July. No Ryan Gosling learning the ways of the Force in Star Wars: Starfighter in 2027. Imagine if these upcoming tentpole productions were canceled not because audiences rejected them but because anyone with a laptop and the right software could make something visually comparable for a tiny fraction of the cost.

This is the future that Doug Shapiro, longtime media analyst and the author of The Mediator newsletter, sees emerging from advances in generative artificial intelligence—so-called GenAI—and shared in a Newsweek AI Impact interview series conversation with Marcus Weldon.

The entertainment industry has weathered numerous technological disruptions over the decades, from silent films to talkies, from broadcast to cable, from DVDs to streaming. But the coming GenAI revolution Shapiro describes strikes at a core assumption that has survived all previous changes: that creating high-quality entertainment requires substantial resources few possess, which has effectively created a persistent and effective moat around the Hollywood studio business.

Traditional studios are already struggling with streaming economics, leading to cost-cutting and wide layoffs, and GenAI will magnify these existing pressures. “Culturally, it’s been very hard,” Shapiro says. “It’s so foreign to the media business historically that it’s a really, really difficult transition.”

Hollywood has long recognized the threat of user-generated, short-form mobile video. For years, traditional studios have watched anxiously as viewer attention shifts from television and movie screens to the touchscreens of smartphones and tablets, where creator content dominates. But the attempts to leverage this trend for big-budget studio content pipelines have largely failed, suggesting that the kind of talent that Hollywood prizes doesn’t cross over to the creator economy, and vice versa.

In 2014, Disney acquired Maker Studios, a multichannel network of YouTube creators, for $675 million. The bet was that YouTube stars could be transformed into TV and movie stars—but that alchemy never materialized. Disney eventually wrote down the acquisition and folded Maker’s operations into its consumer products and interactive division. The Jeffrey Katzenberg–founded Quibi service that launched in 2020 with nearly $2 billion in funding represented the opposite approach: paying top-tier Hollywood studio talent to make short-form video content. It shut down just months later, citing a multitude of reasons, including market timing, the difficulty in driving acceptance of a new streaming app with no free option and the absence of a popular content library.

Despite these high-profile stumbles, social media video has continued its rise. Shapiro calculates that “social video represents about one-quarter of all time spent with video in the U.S.” and that the “total creator media economy revenue was a little shy of $250 billion last year.” By comparison, in 2024, the combined revenues of Disney, NBCUniversal, Warner Bros. Discovery, Netflix, Paramount, Sony Pictures and Lionsgate totaled less than $150 billion.

Up to now, traditional media companies could console themselves with the fact that popular TikTokers might attract massive audiences, but they couldn’t match the production values of a Hollywood blockbuster or a prestige TV series. But what happens when artificial intelligence dramatically reduces that barrier?

Invading Traditional Media Models

Shapiro has spent his career observing media business models from both inside and outside the industry. He spent 14 years as a media equity analyst on Wall Street and then 12 years at Time Warner, where he served as chief strategy officer of Turner Broadcasting System.

Now an independent consultant and senior advisor at Boston Consulting Group, Shapiro has brought his history of analyzing structural economic trends of the media business and charted out the potential impact of AI on entertainment. To understand Shapiro’s thesis about AI’s potential impact, it helps to first understand how he views the traditional media business model.

“Any value that exists is a function of the moats in the value chain,” Shapiro explains, referring to competitive barriers that protect profitable businesses. “And that is in Econ 101, page four or five of your textbook. If you have a business that has no barriers to entry, competitors will come in, compete away the profits. So, you need there to be moats, or barriers to entry, for there to be value in the value chain.

Shapiro identifies two historical critical moats in media: “There was a big moat around distribution because it was very capital-intensive, and there was a moat about content creation because it was expensive but risky [as it is] very hits-driven.”

The internet’s transformative impact was on the distribution moat. “What the internet did was unbundle information from the underlying infrastructure,” Shapiro explains. “So now you no longer needed to own movie theaters or record stores or hybrid fiber-coax or satellites or antennas or any of those things to be in the media distribution business.”

This disruption led to the dramatic industry reshaping we’ve all witnessed over the past 20 years—the rise of streaming platforms, the decline of physical media and the transformation of media business revenue models.

But through all this upheaval, the second moat around content creation has remained relatively intact. Making high-quality entertainment still requires significant capital investment, specialized expertise and established distribution channels, so creating a Hollywood movie or premium TV show remains expensive, typically reaching hundreds of millions of dollars, limiting production to deep-pocketed studios.

But this is where GenAI enters the picture.

AI Could Bridge the Second Moat

Shapiro believes we’re at the beginning of a new wave of disruption that could transform entertainment economics as fundamentally as video distribution was changed by the internet. “The internet caused the cost to move bits to go to zero,” he says, “and GenAI could cause the cost to make the bits to go to zero.”

When analyzing Hollywood production budgets, Shapiro identifies a vulnerability in the industry’s cost structure: “Above-the-line costs—which are less than 20 percent of the total—include the director, showrunner, producer, top writers and top actors. The other 80 percent is, well, everything else.” And that “everything else”—photography, sound, facilities, art department, costumes, effects, post-production—comprises elements increasingly within AI’s capabilities.

The evidence for such a dramatic cost reduction is already emerging in animation. Independent studios adopting GenAI workflows have transformed processes that previously took weeks into tasks that take only hours. Before Katzenberg founded Quibi, he was one of the founding partners of DreamWorks Animation, producers of computer-generated image hits like Shrek and Kung Fu Panda. He’s predicted AI will drop the costs by 90 percent. “In the good old days, when I made an animated movie, it took 500 artists five years to make a world-class animated movie,” he said in 2023. “Literally, I don’t think it will take 10 percent of that three years out from now.”

Less than a year after that prediction in October 24, a team of just nine people working with the assistance of AI for just over three months released Where the Robots Grow, a feature-length animated film, at a cost of roughly $700,000—over 99 percent less than the budget of a typical DreamWorks Animation film.

The democratization of animation tools has already begun reshaping the industry even without AI. At this year’s Academy Awards, the Best Animated Feature Oscar went to Flow, a Latvian film made for just $3.6 million using Blender, a free open-source animation software. The category is typically dominated by big-budget CGI production studios. Flow beat out Pixar’s Inside Out 2 (budget: $200 million) and DreamWorks’ The Wild Robot ($80 million), marking the first time an independent animated film has won in the category’s history. In his acceptance speech, director Gints Zilbalodis said he hoped the win would “open a lot of doors to independent animation filmmakers around the world.”

But the implications extend well beyond animation. What happens when high-quality live-action content can be created by small teams at a fraction of current costs? How would this reshape power dynamics within the entertainment industry?

Perhaps the most subtly revolutionary aspect of Shapiro’s analysis concerns what constitutes “quality” in entertainment. Traditionally, quality was defined by production values—big budgets, well-known talent and technical excellence. But consumer behavior reveals a different reality.

“When you slump down on the couch after a long day and scroll through [Instagram] Reels for 20 minutes rather than pick up the remote that’s an arm’s length away,” Shapiro writes, “you’re revealing that Reels is higher quality than anything on Netflix (or Disney+, Hulu, Max, Amazon Prime, etc.).” What consumers increasingly value, he argues, isn’t technical quality but engagement “quality,” which is defined by authenticity, relatability and personal relevance.

The connection between creators and audiences—what Shapiro calls the parasocial relationship—has become central to today’s media landscape. “That parasocial aspect replaces the [need that arises from the] relatively dispersed populations that we now are,” Shapiro says, and with recent trends like remote work, “we are losing direct physical-social, so parasocial will be an important part of our lives going forward.”

When GenAI tools make high production values accessible to everyone, these elements of connection and authenticity may become even more important differentiators than traditional production values. For media companies, this represents both a threat and an opportunity. But rather than abandoning established quality markers or throwing resources haphazardly at creator content, success will require identifying which attributes resonate most strongly with audiences.

Importantly, the gap between GenAI’s technical capabilities and genuine understanding of human experience may well preserve space for human creators, even in an AI-transformed landscape. The most successful content might combine AI’s production capabilities with the authentic human connection that drives today’s creator economy.

One possible benefit of this transformation could be a more diverse media landscape with a broader range of high-quality content available to consumers. However, Shapiro is skeptical that this will bring back the “middle” ground of entertainment that has been hollowed out by the dual pressures of increasingly gargantuan blockbusters and the user-generated wave of social media. He believes GenAI will move the entertainment industry toward further fragmentation.

“I think that the middle [in] popularity—I would bet against that. I would bet that what you’ll see is this increasing atomization into microcultures,” he says. I think what you’ll see is more and more time will be spent in personalization and in these very small cultures.”

Indeed, this fragmentation might accelerate as GenAI makes it easier to create content tailored to specific niches. The success of Flow demonstrates how a small team with limited resources can create work that resonates worldwide despite not having the production muscle of major studios. But for every Flow that breaks through to mainstream audiences, countless other productions may find sustainable audiences within specific communities without ever needing to achieve blockbuster status, due to the GenAI-enabled asymptotic reduction in production costs.

While Shapiro paints a picture of significant technologically driven disruption, he also emphasizes that human creativity remains essential. When discussing the limitations of today’s generative AI systems in our conversation, he acknowledges they lack a crucial creative capability. “Most generative systems simply work forward from the prompt and then generate,” he says, “but it can’t do what most human creators do, which is start with the end point of the story you want to tell and then work back from there.”

“The answer,” he says, “is humans [will] probably have overarching design control, creative control, narrative control, but GenAI can be used in many places throughout that workflow.”

Who Will Control the New Flow?

Another critical aspect of this transformation is the role of platform companies—the tech giants that increasingly control access to audiences. As content creation becomes democratized, the distribution channel may become the new scarce resource. “I think a good general question to ask is when one input becomes more abundant, what becomes more scarce? And clearly distribution is becoming more scarce,” Shapiro observes. “Owning the end user, owning the platform, being the curator is probably more valuable than ever.”

This suggests that even as content creation becomes more accessible, control over audience attention may become even more concentrated—potentially reinforcing the power of the largest technology platforms. The platform companies already dwarf traditional media in both market capitalization and advertising revenue, and they are also the creators of the key GenAI tools and platforms, making them potentially key gatekeepers of—and catalysts for—this future.

How quickly will these changes unfold? Shapiro predicts that over the next two to three years, we’ll see significant but not total transformation. He expects the adoption will be driven by external pressure. “You’re going to have some GenAI-enabled movie made by five or six people and for a fraction of the cost of blockbusters, he says. “It will be hugely successful, and the studios will be like, Oh my God, we now have to pursue it.”

This mirrors the historical pattern of technological adoption in Hollywood, where innovations often come from outside the established system. “I think an interesting parallel in the movie business was Pixar,” Shapiro notes. “Pixar came from the outside, produced Toy Story. It was the first CGI movie, and it heralded the end of hand-drawn animation in Hollywood. Everyone’s like, Oh my God, we now have to have a CGI strategy!”

At the same time, Shapiro predicts that “on the user-generated end of the spectrum,” we’ll see quality continue to improve, with more scripted content appearing in spaces previously dominated by unscripted videos. “Your TikTok feed could be dominated by scripted content and serialized shows,” he says.

These dual trends—major studios forced to adopt AI tools and independent creators leveraging them to push into the traditional sanctums of professional production—could create a period of intense creative disruption, similar to the early days of streaming video but potentially more far-reaching in their impact on who creates entertainment and how it’s made.

The vision Shapiro sketches in our conversation isn’t complete disruption or replacement of traditional media but rather a fundamental transformation in creative and economic power. And he argues that GenAI won’t eliminate human creativity, but it will democratize who can express that creativity at scale. “I don’t think the traditional media business, as we know it, goes away,” Shapiro says. But, he believes, “it’s going to be a very tough transition.”

In essence, the world he envisions is one where the barriers between professional and amateur content blur further, where entertainment becomes more personalized and fragmented across microcultures and where traditional studios must adapt to a landscape with fundamentally different economics.

In this future, content would be produced along a continuous spectrum, running from highly personalized experiences on one end to microculture communities to occasional mass cultural moments at the other end. As Shapiro notes, “Mass media is only a hundred years old. It’s not necessarily the normal state of affairs that we should be having these mass cultural events.” The Dunbar number—that is the number of other people researchers say our human brains are designed to interact with—is only 150, not the nearly 4 million people who bought tickets to Sinners last weekend. But, as Yuval Harari has noted in his classic text Sapiens, our species will still need the common narratives that unite us, so there will always be a vestigial desire for massive communal experiences at the same instance in time—like trekking to megaplexes for the opening weekend of Avatar 3: Fire & Ash—as well as for common experiences that are asynchronous and more parasocial.