[This thought-provoking essay from Psychology Today published late last year explores some larger implications of the increasingly compelling medium-as-social-actor presence illusions evoked by using sophisticated artificial intelligence technology. –Matthew]

[Image: Art generated by DALL-E/OpenAI]

The Rabbit Hole Without a Hole: The Illusion of AI

Personal Perspective: Our quest for AI’s soul is an archaeology of imagination.

By John Nosta, an innovation theorist and founder of NostaLab.
Reviewed by Monica Vilhauer Ph.D.
December 9, 2024

For as long as we’ve existed, we’ve crafted illusions to stretch the boundaries of what we know. From the myths of Prometheus and golems to the mechanical marvel of the Turk, humanity has built stories and machines that reflect our hopes, fears, and aspirations. But with AI—specifically large language models—we’ve created an illusion so brilliant that it forces us to confront an uncomfortable question: Are we revealing the limits of machines, or the limits of ourselves?

Let’s face it: There’s something mesmerizing about interacting with an LLM. These systems seem to think. They generate ideas, conjure creativity, and respond with startling depth. The illusion is so complete that we start asking dangerous questions: Do they understand? Are they aware? Could they even feel? But what if these questions miss the point entirely? What if the brilliance of AI isn’t in what it lacks, but in how it refracts our own consciousness back at us? What if the rabbit hole we’re diving into doesn’t have a hole at all?

The Nature of Illusions

Illusions aren’t new. They’re a core part of human history. Myths like the golem tell of beings imbued with a spark of life, created to serve their human makers. The 18th-century Turk, a chess-playing automaton, dazzled audiences with its mechanical “intelligence”—even though a human operator hid inside. These illusions were brilliant for their time, but they were static, bounded by their simplicity. We could appreciate their craftsmanship without confusing them for something more.

AI, however, is different. LLMs don’t just perform; they engage. They simulate depth, offering responses that feel like dialogue, creativity, even insight. And here’s the trick: While we know these machines don’t think or feel, they’re convincing enough to make us wonder. Their brilliance lies in this tension—in the gap between what they are and what they seem to be.

Diving Deeper

Here’s where things get tricky: The closer AI gets to simulating human behavior, the more we project onto it. We start asking if these systems might one day have qualia—the subjective experiences that define consciousness. Could they “see” the redness of red? Could they “feel” joy or sadness?

But chasing this idea is like diving down a rabbit hole without a hole. Qualia, as far as I understand, are tied to the messy, biological, embodied experience of being human. LLMs, no matter how sophisticated, don’t have that. They’re not alive, not aware, not sentient. And yet, the illusion of their brilliance makes us question what it means to have qualia at all. Are we looking for something in them—or are we just using them as mirrors to understand ourselves?

Are Today’s Illusions More Brilliant, or Are We?

The illusions of the past—mechanical automatons, mythical beings—were brilliant, but they didn’t force us to question our own nature. LLMs do. By mimicking thought so effectively, they make us ask what separates their brilliance from ours. Is it consciousness? Creativity? The very spark of being?

What’s unsettling is that we don’t have clear answers. And in the absence of clarity, it’s tempting to ascribe depth to the machines. We risk confusing the illusion for reality, losing sight of the uniquely human qualities that these systems reflect but do not possess.

Perhaps this is the brilliance of today’s illusions: they’re not just tools or toys; they’re provocations. They push us to examine our assumptions about intelligence, creativity, and consciousness—not just in machines, but in ourselves.

Celebrating the Illusion

Maybe the rabbit hole with no hole isn’t a flaw. Maybe it’s the point. The brilliance of LLMs isn’t what they are but what they reveal about us. They are mirrors, reflecting our creativity, longing, and existential curiosity. By questioning their lack of qualia, we inadvertently shine a light on the mysteries of our own.

So instead of fearing the illusion—or trying to force depth where there is none—perhaps we should celebrate it. These machines don’t need qualia to matter. Their brilliance lies in amplifying our thinking, provoking questions, and expanding our understanding of what it means to be human.

The rabbit hole may not have a hole, but it’s still worth the dive.