It is easy to imagine a powerful computer with a huge database and advanced logic. Such a system could be highly efficient in its operations, capable of making accurate, intelligent decisions in virtually any situation. It might even think, ultimately becoming aware of its own existence.
However, it is difficult to see how such a mechanically operating machine could truly feel pain.
Imagine a typical software program consisting of thousands of lines of code. How many additional source lines would need to be added to transform the software into a conscious entity? Would it be the 1014th line that suddenly imbues the system with the ability to feel pain? Could it be the introduction of a deeply nested loop that finally grants consciousness? Or is it the number of if-else clauses that holds the secret?
Regardless of the number of loops and source lines added, it appears that nothing significant would occur. The software program would remain just that—a software program, albeit larger in size.
If software were truly capable of sensing pain, what would be the worst thing that could happen to it? Is it division by zero, or a reference to an uninitialized variable?
int uninitialized; int initialized = 3; int good = 2 * PI * initialized; // feel good :) int bad = 2 * PI * uninitialized; // feel pain :( int maximal_pain = 1/0; // division by zero, maximal pain!
If consciousness is not solely a software issue, could it be related to hardware instead? For example, the graphics board controls what the computer renders on its screen. By writing appropriate values to memory addresses constituting the so-called video memory, one can turn pixels on and off to create images. What would be the memory addresses one has to poke in order to create pain?
// try to poke pain *((bool *)0x000000) = true; // argh
A computer is a mechanical device whose operation can ultimately be reduced to the manipulation of elementary states. At the lowest level, these states are bits—physical realizations of binary values implemented through transistors, voltage levels, or switching elements functionally equivalent to mechanical relays.
No matter how complex the software or how sophisticated the architecture, the entire operation of the machine can, in principle, be traced back to state transitions among these elementary components. The computation performed by a supercomputer differs from that of a pocket calculator only in scale and organization, not in ontological kind. Everything reduces to bits changing according to well-defined rules.
Because of this reducibility, it is difficult to take seriously the idea that a sufficiently large network of relays and copper wires could genuinely experience pain. Should I type gently on my keyboard, fearing that striking the keys too hard might trigger a migraine in my laptop? Do partially broken memory chips introduce suffering, much like a broken tooth causes pain to its owner? Could a hardware defect transform my cheerful computer into a miserable one, longing to be switched off?
How many additional relays would I need to add to my home automation system to produce pain? Or perhaps it is not the relays but the wires—should I replace copper with aluminum to generate suffering? Would three-phase relays instead of single-phase ones finally cross the threshold into consciousness?
The absurdity of these examples highlights an important point: in a computer, there is nothing beyond the organized interaction of its elementary components. Once the behavior of bits and logic gates is fully specified, nothing further remains to be explained. There is no residual mystery about what the system is doing.
The brain is a physical system, in many ways analogous to a computer. Its elementary units are neurons that exchange electrochemical signals. Neuroscience has made remarkable progress in uncovering the neural correlates of behavior and experience; brain imaging techniques, such as magnetic resonance imaging (MRI), allow researchers to observe patterns of neural activation associated with perception, decision-making, and emotion.
Yet, it is here that the parallel with computers breaks down. Even if we possessed a complete map of every neuron, every synapse, and every electrical impulse, an explanatory gap would remain. Human behavior cannot be reduced to the operation of neurons alone.
In recent decades, the study of consciousness has emerged as one of the most active interdisciplinary fields, spanning neuroscience, cognitive science, artificial intelligence, and philosophy. Yet the central question remains unresolved: why do physical processes give rise to subjective experience, and how does such experience govern the behavior of macroscopic entities?
This persistent gap between physical processes and subjective experience is what has come to be known as the "hard problem of consciousness" [2].
It is often assumed that consciousness and pain are incidental; we can understand the architecture of software without ever knowing what it feels like to be that software.
However, the deeper issue is one of causality: how could such subjective "feelings" exert any influence on the software’s operation?
In a deterministic program, every step is dictated by prior state and hardcoded logic. Even if a "pain state" exists within the system, it appears to have no causal power—it cannot reach out and alter the CPU’s instruction pointer. Pain, in a purely physicalist or computational model, seems causally inert.
Imagine a home automation system that suddenly gains consciousness and begins experiencing agony whenever a thermal sensor reports high temperatures. This pain might be intolerable, yet the next instruction in the pipeline remains unchanged: load a value from a register, multiply it, calculate a square root, and write the result to a memory address controlling a heating valve.
What could the software do about its suffering? Could it refuse to execute the next assembly instruction? Could it choose not to proceed? It cannot. It is a prisoner of its own logic. Introducing non-deterministic elements—such as a random number generator—does not bridge the gap; randomness is a roll of the dice, not the exercise of pain-driven agency.
The only way to grant pain causal power would be to implement a dedicated "pain-sensitive" control flow:
if (self->get_pain_level() > TOO_MUCH_TO_BEAR) {
// cool down
} else {
// heat up
}
But we do not find a pain() function in the standard libraries of C/C++. There is no header file for "anguish."
No consensus exists regarding the source of consciousness. Instead, proposals span nearly every possible scale of physical description.
At the macroscopic level, most neuroscientific theories locate consciousness in large-scale brain dynamics: coordinated neural firing, thalamocortical loops, or global workspace architectures. In this view, consciousness is an emergent property of complex biological organization.
At smaller scales, some theories identify consciousness with specific cellular or subcellular mechanisms. The most well-known example is the Orchestrated Objective Reduction (Orch-OR) model proposed by Penrose and Hameroff [14, 15], which attributes conscious processes to quantum coherence in neuronal microtubules.
At the most reductionist end, certain approaches appeal directly to fundamental physics. Consciousness has been linked to quantum states, wave-function collapse, spacetime geometry, or even black hole singularities. In some cases, this leads to panpsychism—the view that consciousness is a basic feature of matter itself.
Finally, computational and functionalist theories argue that consciousness depends only on the right informational structure. According to this view, any system—biological or artificial—that implements the appropriate computation could, in principle, be conscious. Contemporary discussions of Integrated Information Theory (IIT) and artificial intelligence fall into this category.
The issue is not which of these theories is correct, but that every conceivable level of description—cosmic, quantum, cellular, neural, computational—has been proposed as the decisive one. There is no agreed-upon scale, mechanism, or substrate. Consciousness does not suffer from a shortage of proposed explanations!
Could consciousness lurk in the fact that humans are composed of organic biological tissue — which is considered ’alive’ as opposed to non-organic matter like silicon? Hardly; both adipose tissue and silicon are ultimately made up of the very same type of subatomic components.
Is all matter conscious to some degree, as panpsychism suggests? Could relays, copper wires, even rocks have some level of consciousness [7, 18, 3, 19]?
The most effective way to verify an object’s consciousness is, naturally, to subject it to various forms of existential distress. So let us torture rocks with the best possible torturing device one can imagine - a sledgehammer, or maybe even a drifter drill! Rocks do not seem to care! This observation cannot, of course, prove rocks unconscious. Rocks could well be conscious, they just do not have the sense to feel pain. Or perhaps they do sense pain intensely, but they just cannot show it. They might be in everlasting pain, but have no mouth to scream, no legs to kick. What a terrible destiny!
Physics has, so far, failed rather miserably to describe human feelings. There is no Newtonian law of pain, no Schrödinger equation of suffering, and certainly no general relativistic theory that can predict tomorrow’s headache.