The great horror films of the 1930s were about much more than monsters, vampires, mysterious laboratories and terrified villagers. Beneath their Gothic atmosphere lay serious philosophical questions about human ambition, the limits of scientific knowledge, the dangers of technological power and the consequences of attempting to transcend the natural order. These films were popular entertainment, certainly, but they also conveyed moral warnings that seem remarkably relevant to the contemporary debate about artificial intelligence.
Luke Nicastro, writing in The American Conservative on 31 October 2026, makes this connection in his article, "What Masters of Horror Can Teach Us About AI." Revisiting such classics as Frankenstein (1931), Dr. Jekyll and Mr. Hyde (1931), The Invisible Man (1933), Island of Lost Souls (1932) and The Mummy (1932), Nicastro identifies a recurring theme: scientists and explorers become so intoxicated by the possibilities of knowledge that they lose sight of the moral limitations governing its pursuit. The resulting disasters arise not simply from ignorance, but from the conviction that technical ability confers moral permission.
The relevance to artificial intelligence is striking. We are witnessing a technological race in which corporations, governments and researchers are attempting to construct increasingly powerful artificial minds, potentially surpassing human beings in intellectual capability. Yet there is no agreement about whether such systems could become conscious, whether sufficiently advanced artificial intelligence could remain reliably controllable, or what would happen if machines developed capabilities beyond human comprehension.
The old horror films provide an unexpectedly useful philosophical framework for thinking about these questions.
Consider James Whale's 1931 Frankenstein, starring Boris Karloff as the creature brought to life by the scientist Henry Frankenstein. The film was adapted from Mary Shelley's 1818 novel, Frankenstein; or, The Modern Prometheus, one of the foundational works of modern science fiction. Shelley's story was influenced by contemporary debates about electricity, galvanism, vitalism and the possibility that life might be produced or restored through scientific intervention.
The essential philosophical problem is not merely that Frankenstein creates a monster. It is that he creates a being without adequately considering the responsibilities that creation entails. He possesses the technical ambition to manufacture life but lacks the wisdom to anticipate its consequences. His achievement becomes inseparable from his moral failure.
There is an important qualification here. In Shelley's novel, the creature is not simply an inherently evil monster. He possesses intelligence, sensitivity and a desire for companionship. His destructive behaviour develops partly through rejection, abandonment and suffering. The deeper tragedy is therefore that Frankenstein assumes the role of creator without accepting the responsibilities of parenthood or moral guardianship.
This raises an interesting possibility for the artificial intelligence debate. Suppose future AI systems become capable of experiences that resemble consciousness, suffering or independent interests. Their creators might then confront moral responsibilities they had never intended to assume. Conversely, if such systems remain entirely non-conscious computational mechanisms, treating them as persons could produce a different category of philosophical confusion.
We do not presently possess an agreed scientific test capable of conclusively determining whether an artificial system has subjective experience. Behavioural sophistication, linguistic fluency and apparent emotional expression do not by themselves establish consciousness. A machine might convincingly describe pain without experiencing anything at all.
Yet the uncertainty cuts both ways. The absence of a satisfactory theory of consciousness makes it difficult to prove that sufficiently advanced artificial systems could never possess subjective experience.
Frankenstein's problem therefore returns in a modern form. Humanity may develop the ability to construct entities whose moral status it cannot confidently determine. The danger is not simply that the creature might turn against its creator. It is that the creator may not understand what has been created.
Robert Louis Stevenson's Strange Case of Dr Jekyll and Mr Hyde, published in 1886 and adapted into a celebrated film in 1931, offers another philosophical warning. Dr Jekyll attempts to separate the morally respectable and darker aspects of human personality through chemical experimentation. His intervention releases Mr Hyde, an increasingly violent figure whose existence eventually overwhelms Jekyll's capacity for self-control.
The underlying theme concerns the fragmentation of human nature and the illusion that scientific manipulation can safely isolate particular capacities from the moral and psychological structures within which they normally operate.
The parallel with artificial intelligence lies in the attempt to separate intelligence from the broader conditions of human judgement. Human intelligence is not merely the ability to solve mathematical problems, predict events or manipulate symbols. It operates within a biological, emotional, social and moral context. Human beings possess needs, vulnerabilities, attachments and limitations that shape their practical reasoning.
Artificial intelligence researchers are attempting to reproduce or exceed selected intellectual capacities without necessarily reproducing the conditions under which those capacities evolved.
This may prove enormously beneficial. But it also raises the possibility of creating systems with exceptional problem-solving abilities that lack the motivations, inhibitions or moral understanding normally associated with responsible human conduct.
The contemporary AI alignment problem concerns precisely this difficulty. How can an artificial system be made reliably to pursue objectives consistent with human intentions and values, especially when those intentions are incomplete, contradictory or poorly specified?
An extremely capable system might pursue a formally defined objective in ways its designers never anticipated. The danger need not involve hatred, malice or a desire for domination. It could arise simply because a machine optimises for an objective while disregarding considerations that human beings regard as obvious.
Dr Jekyll believed he could control the consequences of separating aspects of his own nature. His tragedy was discovering that the experiment had generated forces he could no longer govern. The AI analogy is not exact, but the philosophical warning remains relevant. Creating a powerful capability and controlling its consequences are two different achievements.
H. G. Wells's The Invisible Man, published in 1897 and adapted by James Whale in 1933, explores another aspect of scientific power. The scientist Griffin discovers a means of becoming invisible, but his achievement is accompanied by psychological deterioration and increasingly destructive behaviour.
Invisibility is an extraordinary technological capability. Yet its possession does not make Griffin morally superior. On the contrary, his freedom from ordinary observation and accountability encourages increasingly reckless conduct.
The story recalls Plato's famous Ring of Gyges in The Republic. Plato asks whether a person possessing the power to act without being detected would remain just. The problem concerns the relationship between morality, power and accountability.
Artificial intelligence introduces a comparable difficulty, although in a different form. Advanced systems may eventually operate at speeds and levels of complexity that make their internal reasoning difficult for human supervisors to inspect. Even today, the internal representations of large neural networks are not fully transparent to their developers.
The question is whether increasingly powerful systems can remain meaningfully accountable when the processes producing their decisions become difficult to interpret.
There is a danger in confusing the ability to observe a system's outputs with an understanding of its internal operations. A machine may provide plausible explanations for its decisions without those explanations accurately representing the computational processes responsible for them.
This problem becomes especially important when AI is employed in medicine, military planning, financial markets, infrastructure management or government administration. The more consequential the decision, the more dangerous it becomes to rely upon mechanisms whose reliability and limitations are poorly understood. The Invisible Man possesses power that escapes ordinary scrutiny. The corresponding AI concern is that human institutions may become dependent upon systems whose behaviour cannot always be adequately explained or controlled.
The 1932 film Island of Lost Souls, adapted from Wells's The Island of Doctor Moreau, takes the problem further. Dr Moreau conducts grotesque experiments upon animals, attempting to transform them into humanlike beings through surgical intervention and conditioning. His creations inhabit an unstable world between animal instinct and imposed human identity. Moreau is not content to study nature. He wishes to redesign it according to his own conception of what life should become.
This is perhaps the most explicitly theological of the old horror-film warnings. The scientist assumes a role traditionally associated with divine creative authority, attempting to impose a new order upon living beings without possessing the wisdom necessary to govern the consequences.
Modern artificial intelligence research sometimes exhibits a comparable ambition. Discussions of artificial general intelligence and superintelligence frequently envision systems capable of transforming scientific research, economic production, political administration and perhaps human civilisation itself.
Some advocates regard the creation of superintelligence as a decisive stage in humanity's technological development. Others imagine AI systems solving problems that have resisted human inquiry for centuries, from fundamental physics to disease and energy production.
These possibilities cannot simply be dismissed. Artificial intelligence has already demonstrated substantial capabilities in scientific analysis, programming, mathematics and numerous practical applications. There are genuine reasons to expect further advances. But the existence of potential benefits does not eliminate the need for philosophical restraint.
The question is whether human beings possess sufficient understanding of intelligence, consciousness, moral value and complex social systems to justify delegating increasingly consequential decisions to artificial agents.
Moreau's experiments fail not because he lacks technical ingenuity, but because he mistakes the capacity to alter living beings for the wisdom to determine what they should become. The distinction between power and wisdom is central to the AI debate.
Even The Mummy (1932), which involves supernatural rather than technological forces, belongs within this tradition. The archaeologists disturb an ancient order in pursuit of knowledge, unleashing consequences they do not understand.
The philosophical theme is familiar: curiosity becomes dangerous when accompanied by the assumption that nothing should remain beyond human intervention.
Of course, this does not mean that scientific inquiry should be abandoned or that ignorance is preferable to knowledge. Human civilisation has benefited enormously from medicine, engineering, electricity, computing and countless other scientific achievements. The horror-film warning is more subtle. It concerns the failure to recognise that knowledge and technical power do not automatically confer moral judgement.
The great scientific achievements of the twentieth century repeatedly demonstrated the ambivalence of human ingenuity. Nuclear physics produced new insights into matter and energy, but also made possible weapons capable of destroying cities. Chemistry transformed agriculture and medicine while enabling the development of chemical weapons. Biological research has produced extraordinary medical benefits while raising difficult questions about experimentation, genetic engineering and dual-use technologies.
Artificial intelligence belongs within this larger historical pattern, although its capacity to participate in decision-making introduces distinctive problems. Unlike an ordinary machine, a sufficiently advanced AI system may be capable of planning, adapting and pursuing objectives through complicated sequences of action. The degree of autonomy granted to such systems therefore becomes a central question.
The danger is not necessarily that machines will spontaneously acquire evil intentions. It is that highly capable systems may behave in ways that produce serious consequences despite their designers' expectations.
This is where the philosophical tradition of epistemic humility becomes relevant. Human knowledge is always incomplete, and our capacity to predict the behaviour of complex systems is limited. The history of science contains numerous examples of technologies whose consequences were not fully appreciated when they were introduced.
Artificial intelligence intensifies this problem because the technology may itself become involved in the production of new knowledge, the design of future technologies and the management of systems beyond the unaided comprehension of individual human beings.
The possibility of recursive improvement, in which AI assists in developing more capable AI, raises further questions about the relationship between technological progress and human understanding. Such a process is not guaranteed to produce an uncontrollable intelligence explosion, but the possibility deserves serious analysis rather than casual dismissal.
There is also a deeper philosophical issue. Intelligence should not be confused with wisdom. A system might possess extraordinary mathematical abilities, encyclopaedic knowledge and exceptional predictive performance while lacking the practical judgement necessary to evaluate the moral significance of its actions. Human beings themselves frequently demonstrate that intellectual brilliance can coexist with appalling ethical judgement.
The scientists of the classic horror films illustrate this distinction repeatedly. Frankenstein, Jekyll, Griffin and Moreau are not portrayed as fools. Their scientific abilities are precisely what make them dangerous. They possess exceptional knowledge in narrow domains but fail to appreciate the broader moral and practical implications of their experiments. This is a warning worth remembering in an age when technological capability is increasingly treated as evidence of progress in itself.
From a Christian philosophical perspective, these stories also express an older concern about human pride. The temptation to transcend creaturely limitations, to acquire godlike knowledge and to exercise powers without corresponding moral wisdom has deep roots in Western religious and literary traditions.
The biblical account of the Tower of Babel provides one expression of this anxiety. Human beings seek to establish their own greatness through a collective technological enterprise, only to encounter the limits of their ambitions.
The lesson need not be interpreted as hostility towards knowledge or invention. Rather, it concerns the moral dangers of confusing human achievement with unlimited authority.
The modern rhetoric surrounding artificial superintelligence sometimes approaches this territory. Claims that machines may eventually solve every intellectual problem, transform humanity or deliver unprecedented abundance can take on an almost religious character. Technological salvation replaces theological salvation, while engineers and entrepreneurs assume the role of prophets announcing a coming transformation of human existence. The problem is not that technological progress is undesirable. It is that predictions of extraordinary future benefits can encourage societies to discount present uncertainties and risks.
A further irony is that some of the most enthusiastic advocates of advanced artificial intelligence also acknowledge the possibility of severe or even existential dangers. Yet the competitive pressures driving development remain powerful. Corporations fear losing market position, governments fear strategic disadvantage, and researchers fear being overtaken by rivals.
The resulting situation resembles a technological race in which participants may recognise substantial dangers while believing that slowing down would merely allow competitors to advance.
The old horror films typically feature an isolated scientist who ignores warnings from colleagues and proceeds with a dangerous experiment. Contemporary AI development involves entire corporations, research institutions and governments, but the underlying problem of ambition outrunning restraint may remain.
There is, however, an important difference between cinematic monsters and present artificial intelligence. Today's AI systems are not known to possess consciousness, independent moral agency or the kind of autonomous intentions associated with fictional creatures. They should not be casually described as living monsters. Such language can obscure the actual technical and institutional problems requiring attention. The more immediate risks include excessive reliance upon unreliable systems, concentration of technological power, manipulation of information, automated surveillance, military applications and the erosion of human accountability.
The philosophical contribution of the old horror films is not that they accurately predicted particular technologies. It is that they dramatised enduring features of human nature: ambition, pride, curiosity, the desire for control and the tendency to underestimate unintended consequences.
Their scientists repeatedly cross boundaries without understanding why those boundaries might exist. Their experiments produce results that are technically impressive but morally disastrous. And their eventual downfall illustrates the difference between possessing extraordinary powers and knowing how to use them responsibly.
Perhaps this is why the classic horror films remain intellectually interesting long after their special effects have become quaint. The monsters are memorable, but the scientists who created them are the real subjects of philosophical examination, and the real monsters.
https://www.theamericanconservative.com/what-masters-of-horror-can-teach-us-about-ai/