A meditation on reasoning versus creating, and why the future of artificial intelligence will be settled by the terms — not the tools.
Katzenberg reads the coming age of generative AI through a century of Hollywood history — and lands on a single, deceptively simple conclusion.
A founder's demo in Silicon Valley made Katzenberg feel exactly what he felt in 1986 watching Pixar's Luxo Jr. — awe at a machine that seemed to breathe. But the real question, texted to him by an artist friend 350 miles to the south, was simpler: "Is this the end of us?"
His answer runs through John Philip Sousa's 1906 fight over the phonograph, the arrival of sound in cinema, and Walt Disney's quiet lesson that tools were never the point. Drawing on his own decades on both sides of the divide, he argues that AI today operates on the reasoning side of the line — not yet the creating side — and that the future belongs to those who build with the storytellers, not on top of them.
His central claims are that tools versus taste is a false choice, that the terms — consent, credit, and compensation — settle every revolution, and that the canvas is about to expand yet again.
"The tools-versus-no-tools argument is a trap. First, we must all agree that there should be terms. Then we can have the crucial debate about what fairness requires."
"It's not the how, it's the why."
Step through the six moments when a technological shift rewrote the terms of storytelling — and grew the medium, not shrank it.
Seven moves Katzenberg distills from a century of history — for studios, technologists, and artists alike.
First, we must all agree that there should be terms. Only then can we have the crucial debate about what fairness requires.
Stop arguing about whether the tools should exist. The productive debate is over the terms on which the tools may use human creative work.
Include artists in the building of the technology from the start, with credit, with consent, and with compensation — never on top of them.
Spend energy not on making AI disappear but on deciding the terms on which it will exist — because the terms are everything.
The north needs something it cannot build and cannot buy — creativity — so the tools must be partnered with the people whose life's work informed the models.
The artists who learn the new tools will do things the engineers never dreamed of, just as Chaplin, Lloyd and Keaton made movies emotional after Edison invented them.
The canvas is about to expand again. Studios should take more risks as costs fall, and everyone should resolve to paint on the larger canvas rather than resist its arrival.
Thirteen questions the essay answers — each one mapped to its entity in the knowledge graph.
Jeffrey Katzenberg, the former chairman of Walt Disney Studios and co-founder of DreamWorks and WndrCo, wrote this long-form X post. It argues that AI for creativity will be settled by the terms — consent, credit, and compensation — not by the tools themselves.
In 2023 he predicted that new AI tools would cut the time and cost of producing world-class animation by as much as ninety percent within three years — a claim that alarmed some colleagues and infuriated others.
Reasoning is evaluative and analytical, moving toward a conclusion already implied by the input. Creating is generative, producing something new with no single right answer, where taste, intuition and vision fill the gap where deduction runs out. AI today operates almost entirely on the reasoning side.
Katzenberg reads the resistance as love, not Luddism. He has spent countless hours watching gifted artists rebuild a single second of film ten times, and he believes the pushback represents the collective effort of people deeply passionate about a calling.
Sousa's 1906 essay warned that the phonograph would become a substitute for human skill, intelligence and soul. His fight was really about money — and it helped create the Copyright Act of 1909. He did not stop the technology; he changed the terms under which it could use his work.
Recorded soundtracks displaced tens of thousands of live orchestra-pit musicians, and those jobs never came back. Yet sound also gave cinema the movie musical, the modern score, sound design, and an art form vastly larger than the one it disrupted.
Walt was a technologist forever hunting for state-of-the-art tools, yet he never defined animation by the tools — only by whether the audience believed the character. His principles were timeless; the tools were not.
The Computer Animation Production System, co-developed with Pixar, replaced hand-painted cels with CGI. It gave The Little Mermaid its shimmering finale, Beauty and the Beast its ballroom sweep, and The Lion King its stampede — expanding the canvas rather than diminishing the craft.
George Lucas's aphorism, quoted by Katzenberg, captures the essay's core: the tools (the how) change every generation, but what endures is the purpose and taste (the why) — the ability to make an audience feel.
Silicon Valley has built extraordinary reasoning tools but has not made an artist. Katzenberg argues the knowledge that makes a story matter lives in Hollywood, and the right path includes those people by design, with credit, consent, and compensation.
Jobs said it is technology married with the liberal arts and humanities that yields work that makes our hearts sing. Katzenberg extends it: it describes the tale of two cities, the north's technology joined to the south's creative soul.
History was never settled by the technology; it was settled by the terms. Sousa wrote the law that got composers paid, and the 2023 strike fought for the same consent and compensation. The tools-versus-no-tools argument is a trap.
As barriers and costs fall, more films will be made, not fewer; studios will take more risks; and entirely new forms of storytelling will emerge. He is confident the creative opportunities will expand yet again.
Twelve terms that carry the essay's weight — every one a first-class entity in the graph.
Evaluative cognition that operates on facts, premises and evidence, moving toward a conclusion already implied by the input. Its goal is correctness, not invention.
Generative cognition that produces something new, with no single right answer — where taste, intuition and vision fill the gap where deduction runs out.
The enduring human faculty — with imagination — that decides what moves an audience. Katzenberg argues it cannot be synthesized.
The conditions under which technology may use human creative work — consent, credit, and compensation. History is settled by the terms, not the technology.
The Computer Animation Production System co-developed by Disney and Pixar that replaced hand-painted cels with CGI.
Animation produced with CGI rather than hand-drawn cels, proven to carry real emotional weight by Toy Story.
The arrival of recorded soundtracks a century ago, which displaced pit musicians yet grew cinema into the movie musical and sound design.
Sousa's 1906 essay warning that the phonograph would become a substitute for human skill, intelligence and soul.
The musicians' union's pejorative for recorded soundtracks, used in ads warning against the technology that displaced live orchestras.
The WGA and SAG-AFTRA walkout, which Katzenberg reads as a fight for the same consent and compensation Sousa fought for in 1906.
The technology investment firm co-founded by Katzenberg in 2016, which has backed more than 50 founders.
An extended article published natively on X rather than the short-form 280-character format.
Drag nodes to rearrange, double-click to unpin, and click any node or edge to open its Linked Data description. The controls tray starts closed — open it to filter the graph.
Run live queries against the knowledge graph. Every recipe projects an entity IRI alongside its label, so every result row is clickable. SELECT results render as text/x-html+tr; DESCRIBE and CONSTRUCT results render as text/x-html-nice-turtle.
Explore Knowledge Graph using SPARQL
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
SELECT ?type ?entityIri (SAMPLE(?label) AS ?name) (COUNT(?s) AS ?entityCount)
WHERE {
GRAPH <https://linkeddata.uriburner.com/DAV/demos/daas/jeffrey-katzenberg-ai-for-creativity-deepseek_v4pro-1.ttl> {
?s rdf:type ?type .
?s rdfs:label ?label .
BIND(?s AS ?entityIri)
}
}
GROUP BY ?type
ORDER BY DESC(?entityCount)
Run query
PREFIX schema: <http://schema.org/>
SELECT ?personIri ?person ?role
WHERE {
GRAPH <https://linkeddata.uriburner.com/DAV/demos/daas/jeffrey-katzenberg-ai-for-creativity-deepseek_v4pro-1.ttl> {
?personIri a schema:Person ;
schema:name ?person ;
schema:description ?role .
}
}
ORDER BY ?person
Run query
PREFIX schema: <http://schema.org/>
SELECT ?workIri ?work ?year
WHERE {
GRAPH <https://linkeddata.uriburner.com/DAV/demos/daas/jeffrey-katzenberg-ai-for-creativity-deepseek_v4pro-1.ttl> {
?workIri a ?type ;
schema:name ?work ;
schema:datePublished ?year .
VALUES ?type { schema:Movie schema:Article schema:Legislation }
}
}
ORDER BY ?year
Run query
PREFIX schema: <http://schema.org/>
SELECT ?claimIri ?claim ?creatorIri ?creator
WHERE {
GRAPH <https://linkeddata.uriburner.com/DAV/demos/daas/jeffrey-katzenberg-ai-for-creativity-deepseek_v4pro-1.ttl> {
?claimIri a schema:Claim ;
schema:name ?claim ;
schema:creator ?creatorIri .
?creatorIri schema:name ?creator .
}
}
ORDER BY ?claim
Run query