Gareth Edwards’ The Creator arrived in September 2023 to the kind of critical reception that looks worse the longer you stare at it. A 67% on Rotten Tomatoes. A consensus summary that praised its visuals while dismissing its ideas — “satisfies in the moment even if it lacks substance.” Two Academy Award nominations, a growing audience on streaming, and a critical record that will age about as well as Pauline Kael calling 2001: A Space Odyssey “a monumentally unimaginative movie.” The consensus was not just ungenerous. It was structurally wrong — built on a framework that mistook the film’s central choice for a deficiency.

The pattern is familiar enough to be its own genre. Blade Runner was “a stunningly interesting visual achievement, but a failure as a story.” The Thing was “instant junk.” Vertigo was “far-fetched nonsense.” Each of these verdicts was delivered with professional confidence by people who had applied the correct critical tools to the wrong object and mistaken the poor fit for a flaw in the work rather than in the instrument. The Creator belongs in this lineage — not because it is necessarily as great as those films, but because the nature of the misreading is identical.

The framework error

The most common serious criticism of The Creator is that it fails to interrogate AI consciousness. Justin Chang at NPR wrote that the film “makes no attempt to grapple with AI’s potential.” MaryAnn Johanson complained there was “no hint” that AI characters think or behave differently from humans. FirstShowing.net objected that the film “contains no actual dialogue or conversations or commentary about A.I. and what it means and how it works.”

These are not wrong observations. They are wrong demands. The critics are asking for Ex Machina — a film structured as a Turing test, where AI consciousness is a puzzle to be solved, where the audience is invited to weigh evidence and reach a verdict. Ex Machina is an excellent film doing an excellent version of that particular thing. The Creator is not doing that thing. It is doing the opposite.

Edwards’ film treats AI personhood the way Steve McQueen’s 12 Years a Slave treats Black personhood — as a given. Not as a proposition to be examined, debated, or proved to the audience’s satisfaction. As a fact the film simply operates from, leaving the audience to either see it or reveal something about themselves by not seeing it.

This is not a subtle distinction. It is the entire moral architecture of the film. And the critics missed it — not because they are unintelligent, but because they were carrying a toolkit designed for one kind of science fiction and found themselves holding a different kind. A hammer will tell you everything looks like a nail. A framework designed around Do Androids Dream of Electric Sheep? will tell you that any film which doesn’t ask the question must have forgotten it.

Edwards was explicit about this in production. He told actors not to differentiate between human and AI characters. He deliberately withheld from performers whether their characters would be designated as robots in post-production. “I stopped telling people at a certain point whether they’re a robot or not,” he said at SDCC, “because you notice that people, when they think they’re a robot, they start acting a bit different.” The erasure of the line was the point. The critics who complained about its absence were describing the thesis and calling it a plot hole.

The demand for proof

There is a word for the position that personhood must be demonstrated before it is granted. Several words, actually, depending on the century and the category of person being asked to perform.

The philosophical literature frames this precisely. Under what researchers call the “control paradigm,” the burden of proof falls on those claiming consciousness exists — ethical constraints apply only after consciousness has been verified. The alternative framework, “recognition before proof,” inverts this: those who would deny moral consideration must prove consciousness is absent. The first framework has been the default for every historical denial of personhood. The second is what functioning empathy looks like when formalised.

The Creator operates entirely within the second framework. Its AI characters care for their sick, mourn their dead, build religious communities, protect their children. The film presents this and waits. It does not argue. It does not explain. It shows you beings who are, by every observable metric, persons — and then shows you a superpower that has decided they are not, and is burning them alive from orbit.

The critic who responds to this by demanding that the film “grapple with” whether these characters are really conscious is not making a neutral analytical observation. They are adopting the moral position of the film’s antagonists and calling it rigour. The demand for proof is the thing the film is about. That so many professional critics reproduced the demand without noticing they were doing so is, if anything, evidence that Edwards identified something real.

”Derivative” is not a criticism

The second most common complaint: the film is derivative. An “ultra-familiar mishmash of well-worn sci-fi tropes.” The protector-child dynamic. The reluctant soldier. The vast world glimpsed in fragments.

This charge has a specific problem, which is that it applies with equal force to almost every film the critics making it would call a masterpiece. The protector-child framework — a figure of violence shepherding innocence through a hostile landscape — traces back through Logan, The Mandalorian, The Last of Us, Leon: The Professional, and The Road to the 1970 manga Lone Wolf and Cub. No serious critic calls Logan derivative for using this structure. No one dismisses The Last of Us as a “mishmash of well-worn tropes.” The framework is understood, correctly, as an archetype — a narrative tool, not a narrative flaw.

The line between “acceptable convention” and “unacceptable derivation” is, on inspection, entirely arbitrary. We do not criticise films for using three-act structure. We do not call them derivative for featuring protagonists, or for being set on Earth, or for employing shot-reverse-shot in dialogue scenes. These are conventions — shared grammar that allows filmmakers to build something new from familiar elements. T.S. Eliot put it precisely in 1920: “Immature poets imitate; mature poets steal; bad poets deface what they take, and good poets make it into something better, or at least something different.” Edwards stole from Kubrick, Coppola, Scott, and the Lone Wolf and Cub tradition. He welded the theft into something that looks and feels like nothing else released in the past decade. The charge of derivation is a description of how all films are made, selectively applied to one that a critic didn’t connect with.

The pacing confession

Several critics found the film’s middle section slow. This is worth taking at face value — not as a criticism of the film, but as information about the critic.

The middle act of The Creator does what the middle acts of 2001, Blade Runner, and Apocalypse Now do: it builds the world and the emotional stakes that make the ending land. It allows the relationship between Joshua and Alphie to develop at a pace that earns the film’s conclusion. It trusts the audience to sit with atmosphere, to absorb a world through texture rather than exposition, to let something accumulate rather than detonate.

This is the register in which the greatest science fiction has always operated. It is also the register that has become hardest for a professional critic consuming four to six films a week to receive. When a reviewer trained on the metabolic rate of contemporary studio filmmaking encounters a film that breathes at the pace of Kubrick or Tarkovsky — but wrapped in the production design of a blockbuster — the dissonance produces a specific complaint. The film is “slow.” But 2001 would be slow by the same metric. Blade Runner was literally nicknamed “Blade Crawler” by the Los Angeles Times. The pacing complaint is not a measure of the film’s rhythm. It is a measure of the distance between the film’s rhythm and the critic’s expectations, and those expectations have been set by a content ecosystem optimised for the shortest possible interval between stimulus and response.

Renata Adler wrote of 2001 in the New York Times that it was “somewhere between hypnotic and immensely boring.” Stanley Kauffmann in The New Republic called it “a film that is so dull, it even dulls our interest in the technical ingenuity for the sake of which Kubrick has allowed it to become dull.” These are the same structural complaint, delivered with the same professional confidence, about a film now universally regarded as one of the greatest ever made. The pacing complaint is a clock that tells you the time of the critic, not the quality of the work.

What the film actually is

The Creator is not a puzzle-box. It is not an interrogation of AI consciousness. It is not interested in providing the audience with enough evidence to reach a verdict on whether its characters deserve empathy. It is a film that has already reached its verdict — that personhood is recognised through empathy or denied through ideology, and there is no neutral philosophical ground between the two — and then builds an entire world from inside that conclusion.

The film’s flaws are real. John David Washington’s performance is uneven in places, particularly in the early domestic scenes. Some of the dialogue in the first act is functional rather than sharp. The supporting characters are sketched rather than painted. These are legitimate observations. They are also not what the critical consensus was about. The consensus was about substance — the claim that the film was visually extraordinary but intellectually hollow. That claim is wrong, and it is wrong in the specific way that matters: the critics were looking for a kind of intellectual engagement the film had deliberately chosen not to provide, because providing it would have undermined the film’s actual thesis.

A film that pauses to debate whether its AI characters are conscious is a film that has already conceded the debate. It has accepted that the question is open, that the default position is non-personhood, that the burden of proof lies with the claimant. The Creator refuses that concession. Its AI characters are people. The film does not argue this. It assumes it, the way 12 Years a Slave assumes Solomon Northup is a person and builds its entire moral force from the gap between that assumption and the world’s refusal to share it.

The verdict that will come

The Creator hit theatres during the SAG-AFTRA and WGA strikes, with no cast promotion, no late-night appearances, no premiere circuit. It opened against the Taylor Swift: Eras Tour concert film and Scorsese’s Killers of the Flower Moon. It made $104 million worldwide against an $80 million budget — enough to technically clear production costs, nowhere near enough to cover marketing. By the conventional metrics of 2023, it failed.

Then it arrived on Prime Video in late 2025 and immediately entered the platform’s top ten. The reassessment pieces began. “The most underrated sci-fi movie of the decade.” “Finds streaming success two years later.” The pattern is Blade Runner’s pattern exactly — commercial disappointment, critical ambivalence, a slow migration toward recognition as the framework catches up to what the film was actually doing.

Edwards himself seems to understand the timeline. “It’s a very strange thing,” he said, “how it’s not how people feel about your movie on opening weekend, it’s how they feel about it 10 or 20 years later.”

Film criticism has developed an unexamined conviction that ambiguity equals depth and clarity equals simplicity. This is precisely backwards. The greatest scientists and philosophers are not the ones who make simple subjects complex — they are the ones who take genuinely complex subjects and produce clear answers. The Creator knows what it thinks about AI personhood. It states that position not through dialogue or debate but through the entire architecture of its world — through Buddhist robot monks, through AI children who laugh and hide and grieve, through a species that exhibits every marker of consciousness and is being exterminated by a civilisation that has decided the markers don’t count.

The critics who called this shallow were not wrong about what was missing. They were wrong about whether its absence was a flaw. The film omitted the debate on purpose. The omission was the argument. And the fact that a significant portion of professional critics could not see this — could not recognise a moral position unless it was presented as a question — tells you something important. Not about the film. About the tools being used to evaluate it, and whether those tools are still fit for the work they are being asked to do.