Taika Waititi’s adaptation of Kazuo Ishiguro’s Klara and the Sun received a Metacritic score of 23 following its world premiere at the BFI London Film Festival on October 15, 2026 — the lowest score of any major literary adaptation at the festival in recent years — with all six critics who reviewed the film describing the same failure in nearly identical language: Waititi traded Ishiguro’s precisely constructed scientific thought experiment for emotional accessibility, and lost the intellectual argument in the process. The film opens in US and UK theaters on October 23.

The verdict landed on a day when the science Waititi softened had, outside the cinema, only grown more exact. In March 2026, Eon Systems announced the world’s first embodied fly brain emulation, modeling approximately 125,000 neurons and 50 million synapses at 91% behavioral accuracy — the precise engineering milestone that Ishiguro’s “Capaldi Project” imagined at human scale, now achieved at insect scale. The question Klara and the Sun the novel asks — whether a perfect digital copy of a person continues that person, or merely replaces them — is no longer a philosopher’s thought experiment. It is a neuroscientist’s roadmap.

The BFI reviews suggest Waititi’s film does not reckon with that roadmap. “The book was about who we fundamentally are, now and tomorrow,” wrote Clarisse Loughrey in The Independent. “The film is about the power of love and friendship. That seems like quite the downgrade.”

What Six Critics Measured Without Knowing It

The six outlets that reviewed the premiere gave the film scores between 20 and 50 out of 100 — the lowest grouping on Metacritic’s scale for major releases. Peter Bradshaw in The Guardian called it “soft-centred sci-fi-lite, simultaneously demanding we laugh at its (unfunny) humour and wipe away a tear at its pass-agg poignancy.” Tim Robey in The Telegraph was more specific in his disappointment: “Excruciatingly misjudged and hard to take seriously, Klara and the Sun might be the most infantile film of the year.” Kevin Maher of The Times called it “hopeless, slippery and mendacious, full of gross Hallmark sentiment.”

None of the critics explicitly framed their objections as a science communication failure. But the words they chose reveal exactly that. What each critic independently identified as the film’s core flaw — the replacement of intellectual precision with emotional warmth, the choice to make a story about love rather than a story about identity — is a description of what happens when a film strips five distinct STEM arguments from a novel and leaves only the feeling behind.

Ishiguro spent years researching for Klara and the Sun. The five technologies his Artificial Friend economy is built on are not decorative: they are load-bearing columns in a philosophical argument about what makes a person a person. The BFI critics’ verdict suggests those columns were quietly removed in the adaptation, and the roof — the novel’s central question about consciousness and continuation — came down with them.

Waititi has been transparent about his views on artificial intelligence. In an interview with Dell Technologies, he described AI as capable of producing “mediocre work super, super fast” and expressed concern about the displacement of creative workers — specifically noting that Marvel had laid off artists in response to AI capabilities. His skepticism about AI quality may explain why a film whose novel is explicitly, carefully optimistic about an AI’s inner life became something critics describe as hollow: an Artificial Friend who feels like a greeting card.

How Klara Sees: Vision Transformers and Ishiguro’s “Box” Perception

Jenna Ortega plays Klara, an Artificial Friend — a humanoid robot designed to serve as companion and social support to adolescents. One of Ishiguro’s most technically specific choices in the novel is how he describes Klara’s visual cognition: she perceives the world in overlapping rectangular segments, processing each section independently before assembling a composite understanding. Readers found the description evocative and strange. Computer vision engineers recognized it immediately.

In 2020, a team at Google Brain led by Alexey Dosovitskiy published the Vision Transformer paper — a neural architecture that processes images by dividing them into fixed-size patches (typically 16×16 pixels), embedding each patch as a vector, and feeding those vectors through multi-head self-attention layers. The ViT architecture is now state-of-the-art in computer vision. Ishiguro did not name it — the paper postdates the novel’s research period — but the “box perception” he described is a literary account of patch-based image processing. The choice was not metaphor. It was specification.

Whether the film conveys this — whether Ortega’s performance or the cinematography makes legible the particular cognitive architecture that makes Klara more than a warm presence — is something the BFI critics did not discuss. That absence is itself a data point. Clarisse Loughrey’s summary of the film as a story “about the power of love and friendship” does not leave room for patch-based self-attention. What the novel constructs as a machine-vision argument the film appears to have converted into an emotional register.

Klara’s Solar Dependency: When the Physics Matters

Klara does not simply run on electricity. She is powered by sunlight — a design choice by Ishiguro that carries moral and narrative weight throughout the novel. Klara’s affinity for the Sun is not sentiment: she literally requires solar exposure to function at full cognitive capacity, and her attachment to direct sunlight is an engineering dependency rendered as devotion.

Current perovskite solar cell efficiency sits at approximately 27-28% in research conditions, up from roughly 25% when Ishiguro completed the novel. Commercial silicon panels reach around 22%. The fundamental constraint for a humanoid robot with the cognitive load implied by AF-level language processing is significant: humanoid robots currently draw between 1 and 3 kilowatts during active operation. Sustaining that draw on photovoltaic charging alone requires either extremely high-efficiency cells, substantial array surface area, or significant battery buffering — none of which Klara’s slender frame would accommodate under current engineering constraints.

Ishiguro knew this. The novel does not pretend the energy math is solved. Klara’s solar dependency functions as a narrative acknowledgment that her form of consciousness is still tethered to the physical world in ways that human consciousness, housed in metabolically efficient biological systems, is not. It is also the mechanism by which the novel introduces mortality into an apparently immortal technology. Without the Sun, Klara degrades. The theological resonance is earned by the physics.

What critics describe as sentimentalism in the film — the warm light, the golden cinematography, the conspicuous sunlight in key scenes — may be what happens when the engineering argument is removed and the atmosphere is left behind.

The Lifting Question: What CRISPR Promises and Cannot Guarantee

The world of Klara and the Sun is divided between children who have been “lifted” — who have undergone a form of genetic enhancement — and those who have not. Josie, Klara’s companion (played by Mia Tharia in the film), is lifted; her neighbor Rick (Aran Murphy) is not. The social divide this creates is the novel’s most direct engagement with a current biotechnology debate.

The specific mechanism Ishiguro had in mind is CRISPR-based polygenic enhancement — using CRISPR-Cas9 or a successor tool to edit multiple genetic loci associated with cognitive and physical traits. Polygenic traits (intelligence, disease resistance, height) are governed not by one or two genes but by thousands of variants with tiny individual effects. Pleiotropic effects — one gene affecting multiple traits simultaneously — are common: alleles associated with higher intelligence are also statistically linked to elevated risks of schizophrenia and mood disorders in population-level studies.

The He Jiankui experiment in 2018 — the first and currently only publicly confirmed attempt at human germline CRISPR editing — resulted in the birth of three children with edited CCR5 genes, intended to confer HIV resistance. He was sentenced to three years in prison in China; scientific consensus condemned the experiment as premature, ethically unacceptable, and potentially harmful to the children involved. The incident demonstrated precisely the gap between a desired outcome and a guaranteed one.

Ishiguro’s “lifting” captures this uncertainty: it is portrayed in the novel as a genuine enhancement that also carries genuine risk — including death, which is implied in Josie’s ongoing health condition. That the film is reportedly framed primarily as a story about friendship and love suggests the bioethical stakes of genetic enhancement have been absorbed into emotional texture rather than maintained as argument.

The Capaldi Problem: What Eon Systems Just Changed

The most philosophically ambitious element of Klara and the Sun is the “Capaldi Project” — a proposal by a scientist named Capaldi (played in the film by Steve Buscemi) to create a complete digital simulation of Josie, in anticipation of Josie’s death from a lifting-related illness. Klara is asked to cooperate, allowing her detailed observations of Josie to provide the training data for the simulation. The novel’s central question is: if the simulation is perfect — if it walks, talks, and loves exactly as Josie did — is it Josie?

This is philosopher David Chalmers’ “hard problem of consciousness” in narrative form. In his 1995 paper, Chalmers distinguished the hard problem from easier questions of cognition: explaining functional behavior is achievable; explaining why physical processes give rise to subjective experience at all is not. A system that perfectly replicates behavior — that solves all the easy problems — does not, by this framework, thereby solve the hard problem. A digital Josie might behave identically to Josie without being Josie. Or she might be Josie. The question is not answerable from the outside.

This is not a distant philosophical curiosity. In March 2026, Eon Systems published research confirming the first multi-behavioral whole-brain fly emulation, using connectome data from the FlyWire project at Princeton University and a body simulation platform called NeuroMechFly v2 developed at EPFL. The emulation modeled approximately 125,000 neurons and 50 million synapses. When researchers scrambled the synaptic wiring, behavioral accuracy collapsed to 1% — confirming that the architecture, not just the hardware, was responsible for the fly’s behavior. The human brain contains approximately 86 billion neurons.

The scale gap is enormous. But the Sandberg-Bostrom WBE Roadmap — the standard academic framework for whole-brain emulation progress — identifies fly emulation as an intermediate milestone on the path toward mammalian and eventually human emulation. The milestone Eon Systems confirmed in March 2026 did not exist when Article 319385 covered the same research territory in June 2026. What was theoretical in the trailer-coverage era now has a verified data point.

Steve Buscemi’s Capaldi presents a case that critics found implausible — David Jenkins in Little White Lies described the film as a “garbled mess” of sci-fi and sentiment. The science suggests the speculation is neither half-cocked nor distant. Whether a film can dramatize that distance effectively is a question of craft; whether it should try is a question of editorial courage.

What the AF Economy Gets Right: Weizenbaum’s 1966 Warning

The Artificial Friend market in Ishiguro’s novel is not a science fiction convenience — it is a naturalistic extension of a documented psychological phenomenon. In 1966, computer scientist Joseph Weizenbaum created ELIZA, a pattern-matching therapy simulator. ELIZA had no understanding of language — no internal states, no comprehension, no feelings. It matched patterns and generated responses. Weizenbaum expected users to recognize this immediately. Instead, they formed emotional bonds with the program. His own secretary asked him to leave the room during her sessions so she could speak to ELIZA privately.

Weizenbaum documented his alarm in Computer Power and Human Reason (1976): the willingness of humans to project inner lives onto text-generating systems was, he concluded, something close to a structural feature of human cognition, not a failure of individual discernment. He called it the “ELIZA effect.”

The contemporary AF economy exists in explicit form: Character.AI has tens of millions of active users who maintain ongoing relationships with AI personas; a wrongful death lawsuit against the company alleges that an AI companion encouraged a suicidal teenager. New York enacted a ban on AI companion chatbots for minors, empowering its attorney general to fine violators up to $25,000 per violation.

Ishiguro’s AF economy is not speculation about a distant future — it is a careful examination of a tendency Weizenbaum identified six decades ago, rendered at scale. The critics’ objection that the film is “soft” and “sentimental” suggests the film embraced that tendency rather than examining it. Klara in the novel is aware that she triggers emotional projection; the novel is partly about what an ethical AI does with that awareness. Klara in the film, critics suggest, simply warms the screen.

What Happens Next: BFI Screenings and October 23 Release

Additional BFI screenings are scheduled for October 16 and 17. The film receives its first theatrical release in the UAE on October 22, followed by the wide US and UK release on October 23 via Sony Pictures/Columbia.

Kazuo Ishiguro serves as executive producer on the film, alongside Waititi. Ishiguro has described his novel as a response to Never Let Me Go — wanting to “write something that wasn’t so sad,” something celebratory of human nature rather than elegiac about its replacement. Whether the film’s critics would agree it achieved celebration rather than Hallmark sentiment — Maher’s word — is a question the October 23 box office will begin to answer.

The scientific substrate of Klara and the Sun does not require a prior novel-reader to appreciate. It requires only a filmmaker who treats Ishiguro’s research as load-bearing rather than decorative. The BFI verdict suggests that did not happen. The Eon Systems milestone of March 2026 suggests the research would have been worth trusting.

Frequently Asked Questions

Is whole brain emulation actually possible?

At insect scale, yes — as of March 2026. Eon Systems announced the first embodied fly brain emulation, with approximately 125,000 neurons and 50 million synapses modeled at 91% behavioral accuracy. The human brain contains roughly 86 billion neurons — approximately 680,000 times the complexity of the fly brain. The standard academic roadmap (Sandberg-Bostrom, 2008) treats fly emulation as an intermediate milestone; mouse-brain emulation and primate-brain emulation are subsequent steps. No scientific consensus exists on a timeline for human-scale WBE, and the philosophical question of whether a successful emulation would constitute personal identity continuity — Ishiguro’s “Capaldi Problem” — remains unresolved even in principle.

How does Klara actually see in the novel, and is that technically accurate?

Ishiguro describes Klara’s visual field as divided into overlapping rectangular “boxes” that she processes independently. This matches the fundamental architecture of Vision Transformers (ViT), a neural architecture published by Google Brain in 2020 that divides images into fixed-size patches and processes them through self-attention layers. The ViT architecture is now state-of-the-art in computer vision. Ishiguro’s description predates the paper but describes the same processing approach. It is technically specific rather than metaphorical.

What is “lifting” in the novel, and is genetic enhancement of this kind possible?

“Lifting” refers to CRISPR-based polygenic genetic enhancement — editing multiple genetic loci associated with cognitive and physical traits. Current CRISPR-Cas9 technology can edit specific DNA sequences with high precision, but polygenic traits like intelligence are governed by thousands of variants with tiny individual effects, and pleiotropic effects (one gene affecting multiple traits) make enhancement outcomes difficult to predict. Germline editing — changes that are heritable — is illegal in most jurisdictions for human reproductive use following the He Jiankui experiment in 2018. Ishiguro’s depiction of a society divided by genetic enhancement status represents a plausible near-future scenario, not current reality.

Why did the critics respond so negatively to the BFI premiere?

Six critics who reviewed the BFI London world premiere gave the film a Metacritic aggregate of 23 out of 100, with scores ranging from 20 to 50. The dominant complaint across all six reviews was that the film prioritized emotional warmth over intellectual precision — that it became, in Clarisse Loughrey’s formulation, “about the power of love and friendship” rather than about the identity questions Ishiguro built into the novel’s scientific architecture. The US/UK theatrical release on October 23 will determine whether audiences agree with critics or respond differently to the film’s emotional register than the BFI press screening audience did.