Ed Catmull

The Problem of Creative Institutions

There is a paradox that haunts every organization that depends on originality: the very structures you build to manage complexity tend to kill the thing that made you worth building in the first place. Bureaucracy accretes. Hierarchy calcifies. Success breeds orthodoxy. Ed Catmull spent roughly four decades thinking about this problem — not as a management theorist observing from the outside, but as a computer scientist and studio president who had to solve it in real time, with hundreds of millions of dollars and thousands of careers on the line.

Catmull’s career begins in the 1970s at the University of Utah, one of the great ARPA-funded research nodes, where he worked on foundational problems in computer graphics: texture mapping, z-buffering, subdivision surfaces. These are not minor contributions. Texture mapping alone is one of those ideas so deeply embedded in the rendering pipeline that it’s become invisible — like plumbing. He was, in the most literal sense, building the visual grammar that would eventually make Pixar possible. But the thing that makes Catmull genuinely unusual is that he didn’t stay in the lab. He crossed from technical research into organizational leadership and then spent decades trying to articulate what he’d learned about keeping a creative institution healthy. Creativity, Inc. (2014), written with Amy Wallace, is the primary document of that effort.

The Central Ideas

The book’s core thesis is deceptively simple: creative work is inherently messy, and the natural tendency of organizations is to impose order that destroys the mess before it can become something good. Catmull calls this “the beast” — the aggregate of institutional pressures, risk aversion, and social dynamics that push every project toward mediocrity. The leader’s job is not to have the best ideas but to build an environment where the best ideas can emerge from anyone, and where candor is structurally protected rather than merely encouraged.

The most concrete mechanism Catmull describes is the Braintrust — a group of senior creative people at Pixar who periodically review each film in progress and give unvarnished feedback. Two features make the Braintrust distinctive. First, the feedback has no authority: the director is not obligated to act on any of it. This separates the signal (honest assessment) from the noise (political maneuvering, ego protection, power plays). Second, the Braintrust is explicitly focused on the film, not the filmmaker. This is a structural choice designed to lower defensiveness, and Catmull is clear-eyed about how fragile this norm is — how easily it can degrade into politeness, groupthink, or covert hierarchy.

There’s a deeper idea beneath the Braintrust, though, which is Catmull’s conviction that all early versions of creative work are bad. He calls this “the Ugly Baby.” Every Pixar film, he insists, started out terrible. The organizational question is whether your culture can tolerate — even protect — ugly babies long enough for them to grow into something beautiful. Most organizations cannot. The pressure to show progress, to hit milestones, to demonstrate ROI at every stage, creates a selection environment that favors ideas that look good early, which is precisely the wrong filter for breakthrough creative work.

This connects to a concept Catmull borrows from complexity theory and systems thinking: the idea that failure is not just tolerable but informationally necessary. You cannot navigate an uncertain creative landscape without wrong turns. The question is whether you can make failure cheap enough and psychologically safe enough that people actually take the risks required to find something new. This is not a novel idea in the abstract — it echoes everything from Karl Weick’s work on sensemaking to the lean startup movement — but Catmull’s contribution is showing what it looks like when you actually try to institutionalize it at scale, over decades, with real stakes.

Adjacent Connections

Catmull’s thinking sits at a fascinating intersection: part organizational behavior, part epistemology, part design philosophy. His emphasis on candor and psychological safety anticipates (and in some ways parallels) Amy Edmondson’s research on team psychological safety at Harvard. His insistence on protecting emergent process from premature optimization resonates with Christopher Alexander’s work on pattern languages and the nature of organic order. And his background in computer science gives him an unusual sensitivity to the difference between problems that are well-defined and problems that are wicked — a distinction that many management frameworks, designed for the former, catastrophically mishandle when applied to the latter.

There’s also a connection to W. Edwards Deming’s philosophy of management, which Catmull has acknowledged. Deming argued that most problems in organizations are systemic, not individual — that blaming people for failures caused by bad systems is both morally wrong and practically useless. Catmull extends this: he argues that the system includes invisible social dynamics, unspoken assumptions, and information asymmetries that leaders are structurally unable to see from their position of authority. His chapter on the “hidden” — the things leaders don’t know they don’t know — is one of the most epistemologically honest pieces of management writing I’ve encountered.

What Remains Unresolved

The obvious question is: how transferable is any of this? Pixar had an extraordinary confluence of advantages — deep technical moats, a once-in-a-generation creative partner in John Lasseter, Disney’s distribution muscle, and the luxury of operating in an industry where individual projects can generate enormous returns. Catmull’s principles were forged in an environment with unusually high tolerance for long development cycles and unusually clear feedback signals (audiences either love a film or they don’t). Whether the same principles survive contact with industries that have tighter margins, faster cycles, or more diffuse quality signals is genuinely open.

There is also the Lasseter problem. Catmull’s framework centers on candor, trust, and the moral architecture of creative collaboration. When Lasseter left Pixar and Disney Animation in 2018 amid allegations of unwanted physical contact and workplace misconduct, it raised uncomfortable questions about what the culture of candor actually looked like from the perspective of people with less power. Catmull has said relatively little about this publicly, and the silence is itself a data point. A system designed to surface hidden problems apparently did not surface — or did not adequately address — problems that were happening in plain sight. This doesn’t invalidate the framework, but it does reveal a boundary condition: candor about creative work is not the same as candor about power, and the latter may require entirely different structural mechanisms.

The post-Catmull era at Pixar is also worth watching. Since his retirement in 2019, the studio has produced work of varying quality and has shifted significantly toward streaming releases. Whether the institutional DNA he helped build can survive the departure of its architects — and the very different economic pressures of the streaming era — is an ongoing, real-time experiment in exactly the kind of organizational decay he spent his career trying to prevent.

Why This Matters

What makes Catmull genuinely interesting to me is not the management advice per se but the underlying epistemological stance: the admission that leaders are structurally ignorant, that early versions of good things look like bad things, that systems have emergent properties their designers cannot predict. This is a deeply humble position, and it is rare in leadership literature, which tends to valorize vision, decisiveness, and the mythology of the singular genius. Catmull’s argument is that the genius model is not just wrong but dangerous — that it creates organizations optimized for the appearance of certainty rather than the reality of discovery. The question he spent his career asking — how do you build an institution that stays honest with itself? — is one of those problems that doesn’t have a stable solution, only a practice. And the fact that even he couldn’t fully solve it is, in a way, the most honest thing about his legacy.