On last week’s earnings call, Meta CEO Mark Zuckerberg updated investors on Meta’s progress with its generative artificial tools (AI), now part of a larger suite of tools for advertisers called Advantage+:
“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.