From a Worm’s Learning to the Mystery of Human Pain

Reflections on artificial intelligence, consciousness, learning, and the ethics of pain management

By Ali Araghi, MD

A Personal Introduction: How This Essay Came About

What follows is a reflection of my ongoing intellectual curiosity and practical concerns about the rapidly advancing capabilities of artificial intelligence (AI) and the possibility of artificial general intelligence (AGI). In particular, I find myself wondering how these technologies might reshape some of our most fundamental human needs: the need for purpose, mastery, and autonomy.

As machines become increasingly capable of learning, reasoning, and making decisions, we are confronted with questions that extend far beyond technology. What distinguishes human intelligence from artificial intelligence? Does learning require consciousness? Can intelligent behavior exist without subjective experience? And what does it mean to suffer?

These are not merely abstract philosophical questions. As a physician, I see their practical implications in everyday clinical medicine, particularly when trying to understand and relieve another person’s pain.

This morning, while continuing to read Max S. Bennett’s A Brief History of Intelligence, I became particularly interested in the evolutionary origins of learning and the remarkable abilities of primitive nervous systems. My reading prompted me to begin an exploratory, sometimes argumentative dialogue with ChatGPT about the relationship between learning, consciousness, suffering, and pain.

One question led to another. We moved from the learning abilities of a microscopic worm to Thomas Nagel’s famous philosophical question, What Is It Like to Be a Bat? Eventually, our discussion arrived at a familiar clinical dilemma: how can physicians objectively assess something as deeply subjective as another person’s pain?

The essay that follows is an edited and organized account of that dialogue. The questions, clinical thought experiments, and reflections developed through my exchange with ChatGPT. I have used AI as an intellectual conversation partner and writing assistant, not as a substitute for independent scientific reasoning or clinical judgment.

I share these reflections not because I believe I have resolved the mysteries of consciousness, but because I believe that asking such questions—and remaining willing to challenge our assumptions—is essential to both scientific progress and thoughtful medical practice.

Learning Before Consciousness?

Caenorhabditis elegans, a microscopic nematode, possesses a nervous system of only 302 neurons in the adult hermaphrodite. Yet it can detect environmental signals, navigate toward food, avoid harmful conditions, and modify some of its responses through experience.

This remarkable simplicity raises an evolutionary question: how much of what we call intelligent behavior requires consciousness?

Consider three fundamental principles of associative learning.

Blocking occurs when an organism has already learned that one stimulus predicts a reward, making it difficult to learn about a second, redundant stimulus.

Latent inhibition describes slower learning about a stimulus previously encountered repeatedly without meaningful consequences.

Overshadowing occurs when a more salient stimulus dominates learning at the expense of a weaker stimulus presented simultaneously.

These mechanisms illustrate a fundamental problem of learning known as credit assignment: when several environmental signals precede an outcome, how does the nervous system determine which signals deserve predictive significance?

Modern learning theories suggest that prediction errors—the differences between expected and actual outcomes—help guide the modification of neural connections. In C. elegans, however, evidence for learning should be interpreted carefully: some forms of behavioral plasticity have been demonstrated, but whether its behavior instantiates the same computational mechanisms identified in vertebrate associative-learning paradigms remains an active research question.

Such computations do not inherently require conscious deliberation. Simple neural circuits can change their responses to environmental regularities without necessarily being aware of what they have learned.

But this leads to a more difficult question.

Does the ability to learn imply the ability to experience?

Not necessarily. Learning is an observable change in behavior or, in some definitions, a change in internal state or representation. Consciousness, in its most fundamental sense, involves subjective experience.

Although C. elegans demonstrates several forms of behavioral plasticity, there is no established evidence that it possesses subjective consciousness. Equally important, we have no definitive experimental method for proving that such a simple organism lacks all subjective experience.

We can reconstruct its neural circuits and examine its behavior. But whether there is something it feels like to be a worm remains unknown.

What Is It Like to Be a Bat?

In 1974, philosopher Thomas Nagel published his influential essay, What Is It Like to Be a Bat?

At first glance, the question sounds almost trivial. Of course, humans do not know what it feels like to be bats. We cannot fly, hang upside down, or navigate by echolocation.

But Nagel was asking something more profound.

A neuroscientist could potentially describe many of the neural events involved in a bat’s echolocation. We might understand how ultrasonic signals are emitted, how returning echoes are processed, and how the bat constructs a representation of its surroundings.

Yet Nagel asks whether this objective knowledge tells us what echolocation actually feels like to the bat.

He was not primarily questioning whether bats are conscious. Rather, he assumed they have subjective experiences and questioned whether objective scientific descriptions can fully capture the character of those experiences.

Consider a simpler example: looking at a red rose.

We can describe the wavelengths of reflected light, retinal phototransduction, thalamic processing, and activity in the visual cortex. But none of these descriptions, by itself, conveys the actual sensation of seeing red.

This distinction between objective neural processing and subjective experience is central to the philosophy of consciousness.

It also creates a problem that physicians encounter every day.

The Patient Who Says, “My Pain Is Seven Out of Ten”

Imagine two patients exposed to comparable noxious stimuli.

Patient A reports pain intensity of 7 out of 10.

Patient B reports pain intensity of 3 out of 10.

The stimulus may be similar, yet their reported experiences differ substantially.

We know that pain is influenced by nociceptive input, previous experience, expectations, attention, emotional state, central sensitization, and descending pain modulation.

The difference between these patients does not necessarily mean that one is exaggerating or that the other is unusually tolerant.

Their subjective experiences may genuinely differ.

Now imagine a remarkable scientific discovery.

Suppose researchers identify a brain chemical whose concentration correlates perfectly with experienced pain intensity across all individuals.

A concentration of 70 units consistently corresponds to pain rated 7 out of 10, while 30 units corresponds to pain rated 3 out of 10.

Under these hypothetical conditions, physicians would possess something medicine has long sought: an objective measure that reliably predicts reported pain intensity under the tested conditions.

Such a discovery could transform clinical practice, especially for critically ill, neurologically impaired, or noncommunicative patients, provided that its validity could also be established in those populations.

It could also help evaluate responses to analgesic therapy.

But would it solve the philosophical problem of consciousness?

Not entirely.

We would have a highly reliable indicator of reported pain intensity, but we might still not understand why a particular neural or biochemical state is accompanied by the subjective experience of pain.

This is the distinction between identifying neural correlates of consciousness and explaining consciousness itself.

Furthermore, correlation would not necessarily establish causation. The hypothetical chemical might contribute to pain, result from pain-related processing, or merely reflect another underlying neural process.

Understanding these possibilities would require carefully designed physiological and interventional experiments.

Could an Objective Pain Test Identify Drug-Seeking Patients?

Here, philosophy meets a particularly consequential problem in medical practice.

Physicians must frequently make difficult decisions about pain treatment, especially when considering opioid medications.

Could a reliable objective pain biomarker distinguish patients experiencing severe pain from individuals seeking opioids for other reasons?

Such a test could provide valuable information about pain-related processing or likely pain intensity, but it could not independently establish a patient’s motivation.

A person with opioid use disorder may also experience severe pain. Conversely, a person with relatively low current pain intensity may seek opioids because of dependence, fear of withdrawal, or other factors.

Even a highly accurate pain measurement would not be a perfect test of honesty, addiction, or treatment appropriateness.

This distinction matters ethically.

A biomarker should help physicians understand suffering, not become an instrument for automatically dismissing patients whose reports differ from laboratory measurements.

Pain assessment must remain multidimensional, integrating patient-reported experience, clinical findings, function, psychosocial context, and treatment risks.

An objective measure could improve this process without replacing clinical judgment.

Pain, Nociception, and Consciousness

Another important distinction emerges when considering patients under general anesthesia.

A deeply anesthetized patient may demonstrate spinal reflexes or autonomic responses to noxious stimulation without consciously experiencing pain. However, depth of anesthesia is not perfectly uniform or directly observable, and autonomic or motor responses alone cannot establish the presence or absence of conscious pain.

This illustrates why nociception and pain are not synonymous.

Nociception refers to neural encoding and processing of potentially tissue-damaging stimuli.

Pain is an unpleasant sensory and emotional experience associated with, or resembling that associated with, actual or potential tissue damage.

The former can occur without consciousness. The latter, by definition, involves subjective experience.

Now consider our hypothetical biomarker.

Suppose an anesthetized patient exhibits the same chemical concentration as an awake patient reporting severe pain.

Does that mean the anesthetized patient is experiencing pain?

Not necessarily.

It might mean that our biomarker measures nociceptive processing rather than conscious pain. Alternatively, it might reflect a mixture of nociceptive, arousal-related, autonomic, or other neural processes.

To establish its clinical meaning, researchers would need to investigate how the signal behaves during wakefulness, sleep, anesthesia, and disorders of consciousness, while recognizing that reports of pain are unavailable or limited in some of these states.

This is where philosophical distinctions become experimentally relevant.

Can Suffering Exist Without Consciousness?

Some philosophers argue that the capacity for suffering is evidence of consciousness.

This proposition becomes clearer when we distinguish suffering from the physiological response to harmful stimulation.

A nervous system can withdraw from a noxious stimulus without necessarily experiencing pain. Similarly, stress responses and avoidance behavior do not independently establish subjective suffering.

But if an organism genuinely suffers—if there is an unpleasant experience occurring—then it possesses at least some form of phenomenal consciousness.

The difficulty is determining when physiological processing becomes subjective experience.

Could a worm experience a primitive negative sensation when exposed to a harmful environment? We do not know.

Could consciousness have evolved gradually, rather than appearing abruptly with the development of complex brains? This is a plausible hypothesis, although not an established evolutionary fact.

The possibility invites scientific humility about the subjective lives of organisms very different from ourselves.

What Should Physicians Learn From Philosophy?

Modern medicine has achieved extraordinary success by translating biological phenomena into objective measurements.

We measure oxygenation, cardiac output, intracranial pressure, serum electrolytes, and thousands of other physiological variables.

Yet some of the most important dimensions of illness—pain, anxiety, fear, fatigue, despair, and suffering—remain fundamentally dependent on subjective experience.

The lesson is not that objective science is inadequate or that subjective reports are beyond scientific investigation.

Rather, objective measurement and subjective experience provide different but complementary forms of information.

A physician who relies exclusively on laboratory measurements risks overlooking the lived experience of illness.

A physician who ignores objective evidence risks missing important physiological mechanisms and treatment complications.

Good medicine requires both, and this can be considered the ART OF MEDICINE.

Our responsibility is not simply to measure disease, but to understand the person experiencing it.

Final Reflection: From 302 Neurons to the Human Bedside—and the Age of AI

The evolutionary journey from primitive nervous systems to the human brain is extraordinary.

Along that journey, nervous systems developed increasingly sophisticated abilities to detect signals, predict outcomes, learn from errors, and regulate behavior.

At some point in evolutionary history, subjective experience emerged—or perhaps its earliest forms appeared much sooner than we currently imagine.

We do not yet know.

Today, artificial intelligence introduces another dimension to these questions. Machines can increasingly recognize patterns, generate predictions, solve problems, and support decisions. Some of these capabilities resemble functions that biological nervous systems evolved over hundreds of millions of years.

But functional similarity does not establish that artificial systems possess subjective experience. Whether consciousness requires particular biological properties, specific computational architectures, or something else remains unresolved.

For me, the implications extend beyond whether machines might become conscious.

As AI becomes more capable, how will human beings preserve their sense of purpose, mastery, and autonomy? What happens to professional identity when machines perform tasks that once required years of human learning and expertise? And how should physicians use increasingly sophisticated algorithms without surrendering the responsibility to understand their patients?

These questions are especially important in medicine, where technical competence and human understanding must coexist.

What we do know is that the human brain produces both measurable physiological activity and experiences that matter profoundly to the individual.

For the physician, this creates an important ethical responsibility.

We may not directly experience another person’s pain, but our inability to do so does not justify dismissing it.

And even if neuroscience eventually develops an extraordinarily accurate objective measure that predicts reported pain, the measurement will be most valuable when used to improve understanding and treatment rather than to replace compassion or clinical reasoning.

Perhaps this is where medicine, artificial intelligence, and the philosophy of mind meet most meaningfully:

Science helps us understand the mechanisms of suffering. Philosophy reminds us to ask what suffering means to the individual experiencing it. Artificial intelligence may extend our ability to analyze and predict. But medicine must preserve the human responsibility to interpret, care, and act wisely.

Selected References and Further Reading

1. Bennett MS. A Brief History of Intelligence: Evolution, AI, and the Five Breakthroughs That Made Our Brains. 2023.

2. Nagel T. What Is It Like to Be a Bat? The Philosophical Review. 1974;83(4):435–450.

3. Rescorla RA, Wagner AR. A theory of Pavlovian conditioning: Variations in the effectiveness of reinforcement and nonreinforcement. In: Classical Conditioning II. 1972.

4. Schultz W, Dayan P, Montague PR. A neural substrate of prediction and reward. Science. 1997;275:1593–1599.

5. Chalmers DJ. The Conscious Mind: In Search of a Fundamental Theory. Oxford University Press; 1996.

6. Raja SN, et al. The revised International Association for the Study of Pain definition of pain: Concepts, challenges, and compromises. Pain. 2020;161(9):1976–1982.

7. Baliki MN, Apkarian AV. Nociception, pain, negative moods, and behavior selection. Neuron. 2015;87(3):474–491.