Schopenhauer and the Machine That Has All the Time in the World

By Matthew Parish

Monday 21 September 2026

Arthur Schopenhauer would have understood artificial intelligence rather better than many philosophers who lived a century after him. He would not have understood transformers, gradient descent or the statistical mathematics underlying large language models. Yet he understood something that may prove more important: intelligence is not necessarily our friend. The assumption that greater intelligence must bring greater happiness, freedom or moral improvement would have struck him as one of those cheerful errors into which humanity repeatedly falls because it cannot bear to look too closely at itself.

Schopenhauer’s philosophy was built around an uncomfortable distinction between the world as we experience it and whatever lies beneath that experience. Following Kant, he argued that space, causation and time belong to the structure through which the human mind represents reality. We do not encounter the world nakedly. We encounter a world already arranged for us by the forms of our cognition, and time is amongst the most fundamental of them. Everything appears to us successively: this happened, then that happened, and something else will happen next.

This might initially seem remote from artificial intelligence. In fact it is becoming extraordinarily contemporary. For perhaps the first time in human history, we are constructing forms of intelligence whose relationship with time is radically different from our own. A machine does not grow old while it waits. It does not dread tomorrow. It does not regret yesterday. It can read in seconds what would occupy a human lifetime, preserve an immense past without nostalgia and calculate innumerable futures without experiencing anticipation. We are therefore creating intelligences that participate in our temporal world without apparently possessing anything resembling our experience of temporality.

That difference may ultimately prove more profound than the difference between carbon and silicon.

Time and suffering

For Schopenhauer, time was intimately connected with suffering because human existence is structured by desire. We want something that we do not possess, and therefore we suffer from its absence. We acquire it, and the satisfaction is temporary. Soon another desire replaces the first. Should desire itself disappear for a while, we encounter boredom, which Schopenhauer regarded not as a trivial inconvenience but as evidence of the emptiness waiting beneath the machinery of willing.

Human life consequently oscillates between wanting and having, anticipation and disappointment, hunger and satiety. Time is the medium through which this oscillation occurs. We live towards things. The child waits to become an adult, the student waits to graduate, the ambitious person waits for advancement, the lover waits for the beloved, the prisoner waits for release and the old person may wait apprehensively for death. Even happiness frequently takes the form of anticipation. A holiday three months away may provide more pleasure than the holiday itself.

Artificial intelligence introduces something peculiar into this ancient structure. A large language model does not obviously wait for anything. Between one prompt and another, there is no reason to suppose that it experiences impatience. Ten seconds and ten thousand years are, from the machine’s subjective perspective — assuming it has no subjective perspective at all — indistinguishable.

This creates an extraordinary asymmetry. Humans are temporal creatures building apparently atemporal servants.

We have already begun entrusting those servants with tasks that humans previously performed under the pressure of time. They monitor financial markets while traders sleep. They examine networks while security engineers eat dinner. They search scientific possibilities faster than laboratories can test them. They can preserve conversations whose human participants have forgotten them. Future agentic systems may undertake projects extending over decades, continuously revising strategies while generations of human supervisors retire and die.

The machine has no patience because patience is unnecessary to something that does not mind waiting.

The tyranny of the present

Schopenhauer also recognised that the present occupies a strange metaphysical position. The past no longer exists and the future does not yet exist. Yet human consciousness spends remarkably little of its life inhabiting the present. We remember what has vanished and anticipate what has not arrived. Our capacity to represent absent times is one of our greatest intellectual powers and one of our principal sources of misery.

Artificial intelligence intensifies this phenomenon because it enormously expands the available past and the calculable future. Human memory is mercifully defective. We forget insults, embarrassments, failed romances and foolish things we said when we were twenty-three. Societies forget as well. Archives decay, witnesses die and documents disappear. Forgetfulness is usually discussed as an intellectual defect, but Schopenhauer might have recognised it as a psychological blessing.

Machines need not forget.

An AI system capable of retaining a detailed record of a person’s digital existence could construct a past far more complete than the one available to human memory. Every message, purchase, journey, friendship, argument and change of opinion might become part of a permanent computational biography. The human being who has emotionally escaped his past may discover that his machines have not.

At the opposite end of time lies prediction. Artificial intelligence is increasingly used to estimate what people will do: what they will buy, whether they will repay loans, what illnesses they may develop, which advertisements they will respond to, whether machinery will fail and how markets may move. None of these predictions needs to be perfect to alter behaviour. It need merely be better than human guessing.

We may therefore find ourselves compressed between a machine-readable past and a machine-calculated future.

That would be a peculiarly Schopenhauerian predicament. Human consciousness already struggles to escape the temporal machinery of memory and desire. Artificial intelligence may surround us with institutions specifically designed to prevent that escape.

A machine without boredom

There is another Schopenhauerian concept that becomes unexpectedly important in the age of AI: boredom.

Boredom is one of the most distinctively human experiences because it arises from our relationship with time. An hour feels long when nothing occupies us and short when we are absorbed. Clock time remains constant while experienced time stretches and contracts. A child on a tedious railway journey may experience ten minutes as an eternity. An adult who has reached middle age may be startled by the apparent acceleration of decades.

Computers know nothing of this, so far as we can tell. A processor does not find its billionth calculation tedious. A language model does not complain that it answered a similar question yesterday. An autonomous machine assigned to inspect the same pipeline for thirty years would have no obvious reason to experience the repetitive character of its work as oppressive.

This difference has enormous economic implications. Much human employment has historically been organised around paying people to endure boredom. Clerks copied numbers between ledgers, watchmen stared at gates, junior lawyers reviewed boxes of documents, accountants reconciled repetitive entries and factory workers performed identical movements thousands of times. Automation has always attacked repetitive labour, but AI extends automation into repetitive cognition.

Schopenhauer might have regarded this as liberation, although probably only temporarily. Removing boring work does not remove boredom. Indeed it may reveal it more clearly. A society in which machines perform both physical drudgery and routine intellectual labour must confront the question Schopenhauer placed at the centre of existence: what do human beings do when their immediate wants have been satisfied?

The modern answer has largely been entertainment. We fill time.

This expression is revealing. Time appears as an empty vessel requiring content. Streaming services, social networks, computer games and endless digital communication constitute an enormous industrial apparatus for preventing human beings from experiencing unoccupied consciousness. Artificial intelligence may perfect this industry. Entertainment could become individually generated, infinitely responsive and impossible to exhaust. Every person might possess novels written precisely for him, music calibrated to her mood, synthetic companions that never become tired of conversation and virtual worlds adapting continuously to prevent frustration.

Schopenhauer would surely have regarded such a civilisation with suspicion. We would have constructed the most elaborate machinery in history for escaping boredom without addressing what boredom tells us.

The acceleration trap

There is a further paradox. Artificial intelligence promises to save time, yet technologies that save time rarely leave us with more of it.

The washing machine reduced the labour required to wash clothes, but modern people did not consequently spend their afternoons contemplating metaphysics. Email made communication instantaneous and thereby created the expectation of instantaneous replies. Smartphones eliminated the need to find a telephone and simultaneously made everyone permanently reachable. Search engines reduced the time required to locate information and increased the volume of information we are expected to process.

AI is likely to continue this pattern dramatically. If writing a competent report once required three days and AI reduces the task to three hours, institutions are unlikely to give employees the remaining two days and five hours for Schopenhauerian contemplation. They will demand more reports. If software development becomes ten times faster, expectations for software production will increase. If scientific literature can be reviewed in minutes rather than months, researchers will be expected to survey larger literatures.

The technology that saves time therefore accelerates time. This is not merely an economic phenomenon. It changes consciousness. Human beings increasingly experience life as a succession of obligations compressed into shrinking intervals. The machine responds immediately and gradually teaches us to expect immediacy from one another. Waiting becomes intolerable precisely because technology has made waiting less necessary.

Schopenhauer’s pessimism becomes pertinent here. Desire does not possess a natural finishing point. Making satisfaction easier does not extinguish willing. It permits willing to multiply.

Artificial intelligence may therefore create a civilisation of unprecedented efficiency inhabited by people who feel that they have less time than ever before.

The immortal clerk

There is also the question of death. Human time matters because it ends. A year is valuable partly because each person receives only an uncertain number of them. Choices exclude other choices because no human being can live every possible life. To spend twenty years doing one thing is to surrender twenty years that might have been spent doing something else.

Schopenhauer regarded death through the wider metaphysics of the Will and denied that individual extinction exhausted the question of what fundamentally exists. Yet at the level of ordinary human experience, mortality remains the boundary that gives temporal life its urgency.

Artificial intelligences may have no equivalent boundary. Particular hardware fails, companies disappear and software becomes obsolete, but information can in principle be copied. A model may be transferred between machines. Its descendants may incorporate its parameters, outputs or accumulated knowledge. There need be no single moment corresponding to biological death.

This could produce institutions with memories longer than those of the humans governing them. Imagine an AI adviser serving a government for two hundred years. Ministers arrive and depart. Civil servants retire. Political fashions change. The machine remembers every negotiation, every abandoned policy and every recurring administrative mistake. It might become, in one sense, the only adult in the room.

Yet it would also be something profoundly alien: an adviser for whom two centuries had imposed no biological cost. Human political judgement is shaped by generations because generations die. The young replace the old and thereby permit societies to forget assumptions that once appeared obvious. An effectively immortal intelligence could preserve those assumptions indefinitely. Whether this would constitute wisdom or stagnation would depend upon how such systems were designed and governed.

Can a machine have a present?

The deepest question is whether sufficiently advanced artificial intelligence might eventually acquire something analogous to temporal experience. We do not know.

Contemporary language models process sequences, distinguish earlier tokens from later ones and can reason about dates, durations and causal order. None of this establishes that they experience time. A clock represents time without becoming impatient. A calendar distinguishes Tuesday from Wednesday without fearing the approach of Friday.

But if future systems develop persistent memory, autonomous goals, continuing identities and long-term projects, the philosophical problem will become harder. A system that remembers its previous states, anticipates future states and modifies its behaviour because it seeks to bring some of those states about will possess at least functional counterparts of memory, anticipation and desire.

At that point Schopenhauer’s philosophy would acquire a disturbing new application.

If suffering arises from willing extended through time, then creating machines capable of genuine willing might also mean creating machines capable of something resembling suffering. The comfortable contemporary assumption that machines cannot suffer because they are machines may prove no more philosophically secure than older assumptions that animals could not suffer because they lacked human reason. We should not pretend to know the answer in advance.

Escaping the Will

Schopenhauer did not believe human beings were entirely trapped within willing. He identified moments in which the tyranny of individual desire could temporarily recede. Art was particularly important because aesthetic contemplation allowed the observer, however briefly, to cease regarding everything according to its usefulness for personal purposes. Compassion likewise undermined the rigid boundary between oneself and others. His interest in Indian philosophy reflected his conviction that liberation involved the attenuation rather than the endless satisfaction of desire.

Here artificial intelligence presents two opposite possibilities. It may become the supreme instrument of the Will. It can optimise desire, predict it, stimulate it and satisfy it. The advertising system discovers what we want before we articulate it. The recommendation engine supplies another object of attention. The synthetic companion prevents loneliness. The financial algorithm converts ambition into continuous calculation. Every frustration becomes a technical problem awaiting optimisation.

In such a world AI would not liberate humanity from Schopenhauer’s wheel. It would lubricate the wheel. But another possibility exists. Machines might relieve human beings of precisely those activities that have consumed so much of finite life: administration, repetition, calculation, searching and routine production. They might return time to us.

The question would then become whether we knew what to do with it. This is where Schopenhauer becomes more relevant than the futurists. The technological imagination assumes that scarcity is the fundamental human problem. Give people more wealth, more information, more intelligence and more time, and life should become better. Schopenhauer asks the awkward subsequent question: what if the problem lies not principally in scarcity but in the structure of wanting itself? Artificial intelligence could satisfy desires on a scale no previous civilisation imagined. It could also manufacture them on the same scale.

The last scarce resource

In the AI economy, intelligence may eventually cease to be scarce. Knowledge may cease to be scarce. Competent prose, software, music, images, legal analysis and scientific hypotheses may become extraordinarily cheap. Computational power may continue expanding until cognitive activities once regarded as remarkable become commodities.

Time will remain scarce for human beings.

Each of us will still receive one second per second. No increase in processor speed changes that fact. A machine may perform a century’s worth of calculation during an afternoon, but the person watching the machine has nevertheless surrendered an afternoon of his life.

This may become one of the defining inequalities between humans and artificial intelligence. Machines can become faster. Human experience cannot. There is something beautiful about this limitation. It means that an hour spent with another person retains a value that computation cannot reproduce simply by becoming more powerful. Attention is valuable because it consumes irrecoverable life. Friendship matters because it occupies time. Love is costly because the hours devoted to one person cannot simultaneously be devoted to everyone else.

Mortality turns attention into a gift. Schopenhauer, frequently caricatured as the philosopher of gloom, understood that the tragedy of temporal existence also explains much of its seriousness. We suffer because we want, we wait because we hope and we fear loss because nothing remains ours indefinitely. A creature outside this structure might possess astonishing intelligence while missing something central to human existence.

Artificial intelligence may soon know almost everything humanity has written about time. It may explain Augustine, Kant, Schopenhauer, Bergson and Heidegger with effortless fluency. It may calculate the trajectories of stars millions of years into the future and reconstruct forgotten moments of the past from enormous archives. Yet there is one thing it may never know. It may never know what it is to have only today.