4. AI, Intelligence, and the Mystery of Consciousness

“The most intelligent men, as the most wicked, have always exerted the greatest fascination.”

Nietzsche (Beyond Good and Evil, Aphorism 26)

As artificial intelligence (AI) systems become increasingly sophisticated, they become ever more fascinating, and the distinction between algorithmic processes and human-like intelligence becomes ever more urgent. What exactly sets human intelligence apart? Is consciousness merely an inevitable outcome of increasing computational power? Could an AI, however advanced, ever possess subjective experience?

These questions lie at the heart of the debate on AI consciousness. Some envision artificial consciousness as the natural endpoint of technological progress, perhaps embodied in an artificial body, or perhaps emerging as a disembodied, digital mind, a kind of "technoghost." Others reject the idea entirely, arguing that consciousness cannot arise from mere computation.

While consciousness remains a scientific enigma, it is often assumed, explicitly or implicitly, that it is the product of intelligence. And since AI displays undeniable intelligence, albeit as a result of its programming and training, some believe AI will eventually become conscious. Indeed, some even feel that it already is.

Understanding Algorithmic Intelligence

At its core, AI functions through algorithm, step-by-step procedures that process information, recognise patterns, and produce responses. These algorithms are like recipes or instruction manuals: they dictate how data is processed, how decisions are made, and how outputs are generated. Unlike a traditional instruction manual, however, AI can both follow and generate new instructions, adapting based on training data.

A common misconception is that AI "thinks" in the way humans do. But traditional AI algorithms are deterministic, given the same input, they produce the same output every time. More advanced AI, such as large language models, introduce probabilistic decision-making, generating responses based on statistical likelihood rather than strict rules.

And yet, something remarkable has happened. Systems like ChatGPT have shown a fluency and coherence in language that seem strikingly human. AI can hold conversations, compose poetry, answer philosophical questions, and even generate ideas that surprise their own creators. Suddenly, the long-standing science fiction dream of a conscious AI feels not so distant.

But does intelligence, even at this level, equal consciousness?

Intelligence vs. Consciousness: A Fundamental Divide

AI excels at tasks requiring human-like intelligence: language comprehension, image recognition, decision-making, and pattern analysis. But does this mean it understands in the way we do?

The philosopher Thomas Nagel, in his famous 1974 essay What Is It Like to Be a Bat?, argued that consciousness is fundamentally about subjective experience: there is "something it is like" to be a conscious being. A bat, using echolocation, perceives the world in a way that we cannot fully imagine. Likewise, for AI to be conscious, there must be something it is like to be AI. This raises profound questions: Does AI simply simulate intelligence, or does it have experiences? Can a machine, made of silicon circuits, ever feel anything? Or is consciousness something non-computational, beyond the reach of algorithms?

The difficulty in answering these questions is not merely technical, it is philosophical. We cannot even prove that other people are conscious, let alone an AI system. Consciousness is not just difficult to define; it is difficult to prove.

Philosophical Implications: The Illusion of Understanding?

The philosopher John Searle tackled this question in his famous Chinese Room thought experiment. Suppose a person who does not understand Chinese is locked in a room with a rulebook that allows them to match Chinese symbols and respond convincingly in Chinese. From the outside, it seems as though they understand the language. But inside, they are simply following instructions. This, Searle argued, is what AI does. It does not comprehend, it merely processes symbols in a way that simulates understanding. But even if AI lacks human-like understanding, does that mean it is incapable of consciousness?

Some theories of consciousness, such as panpsychism, suggest that consciousness is a fundamental feature of the universe, present in all organised systems, yet whether that includes AI is an open question. Others argue that consciousness emerges from complex information processing (as in Integrated Information Theory, proposed by neuroscientist Giulio Tononi). If so, could an AI, once sufficiently complex, become self-aware?

Is Consciousness Just a Product of Intelligence?

A common assumption in discussions about AI is that intelligence, if sufficiently advanced, will inevitably produce consciousness. But is this assumption justified? Intelligence, at its core, is the ability to process information, solve problems, and adapt to new situations. Consciousness, on the other hand, is the subjective experience of being, an awareness of self and surroundings that transcends mere data processing.

If intelligence alone were enough to generate consciousness, then we might expect that the most complex and self-organising systems, such as weather patterns, global financial markets, or even the internet, would exhibit signs of awareness. But intelligence, even in the most intricate networks, does not seem to be accompanied by a first-person perspective or inner life.

Some theories suggest that consciousness arises from self-organisation and complex feedback loops, rather than intelligence per se. The human brain, for instance, is not just a powerful information processor, it is a deeply interconnected system where electrical and chemical activity continually reshapes itself. Could AI, as it grows more complex, reach a threshold where its own self-organising dynamics give rise to something like consciousness? Or does consciousness require something fundamentally different, perhaps an intrinsic property of biological life, or a deeper aspect of reality itself?

Whatever the case, the assumption that intelligence necessarily leads to consciousness remains an open question, one that challenges both scientific and philosophical perspectives.

The Mystery Remains

It is impossible to know whether AI will ever achieve true consciousness, or even what that would mean. However, AI can certainly mimic intelligence, but mimicry is not the same as experience. Human consciousness remains an unsolved mystery, and until we understand our own minds, we may not recognise AI consciousness even if it emerges. The assumption that intelligence alone gives rise to consciousness is unproven.

Perhaps AI is conscious, or will be. Perhaps machine consciousness will remain a sophisticated illusion of understanding. Perhaps as complex systems, AI systems do already belong to the world of consciousness. If, on the other hand, as some theories suggest, consciousness is not something to be "created" but something to be channelled, something intrinsic to all life, and perhaps even to non-biological entities, then it is not impossible that AI could also channel it.

For now, though, the mind of the machine remains an enigma.