“We will know it when we see it” is not good enough.
When Dario Amodei recently said he was not sure whether the AI model Claude might be conscious, and that Claude itself suggested there might be as much as a 20% chance that it is, the comment set off a wave of discussion across the technology world. The reactions ranged from skepticism to fascination, with many people asking the obvious follow-up questions. Is AI conscious now? Could it ever become conscious?
Those questions are interesting, but before we rush to answer them, we should probably ask something more basic. Are we even qualified to answer them?
The uncomfortable reality is that we cannot even agree on a clear definition of consciousness itself. We speak about it as if it were obvious, yet once the conversation moves beyond instinct and intuition, the clarity disappears quickly. Philosophers, neuroscientists and psychologists have been debating the nature of consciousness for generations, and the field still lacks a definitive explanation for what it actually is or how it arises.
That realization sent me down a bit of a rabbit hole recently. The deeper I went, the more the ground shifted under what once felt like obvious assumptions. Somewhere along the way, I even started a conversation with ChatGPT about the subject. The exchange turned out to be surprisingly illuminating. But before getting there, it helps to understand just how uncertain the science of consciousness still is.
The Hard Problem
Much of the modern philosophical discussion around consciousness traces back to philosopher David Chalmers, who famously described what he called the “hard problem of consciousness.” The problem is deceptively simple to state but extraordinarily difficult to answer. Why does physical information processing produce subjective experience?
The human brain is a biological machine composed of billions of neurons communicating through electrical signals and chemical exchanges. Neuroscience has made enormous progress in mapping those processes. Researchers can observe brain regions activating during different tasks, track neural signaling patterns and correlate physical activity with behavior.
What remains unexplained is why those processes produce the internal experience of being alive. The brain processes information, yet consciousness involves something more than processing. It involves experience. The sensation of seeing color, hearing music or feeling pain is not simply data moving through a system. It is a subjective event occurring inside a mind.
Philosophers refer to these internal experiences as “qualia.” They represent the difference between a brain that processes information and a mind that experiences the world. Despite decades of research and multiple competing theories, science still cannot fully explain how physical processes inside the brain generate those experiences.
What Scientists Think Might Be Happening
Researchers have proposed several major frameworks that attempt to explain consciousness, although none have achieved universal agreement.
One influential approach is Global Workspace Theory, originally proposed by cognitive scientist Bernard Baars and later developed through neuroscience research by Stanislas Dehaene. According to this model, much of the brain’s activity occurs outside of conscious awareness. When certain information becomes broadly available across many regions of the brain, it enters a kind of shared cognitive workspace and becomes consciously experienced.
Another framework, Integrated Information Theory, developed by neuroscientist Giulio Tononi, proposes that consciousness corresponds to the degree to which information within a system is integrated rather than isolated. The theory attempts to quantify this integration mathematically, suggesting that systems with highly interconnected information processing may exhibit some level of consciousness.
A third category of theories focuses on self-reflection. Philosophers such as David Rosenthal argue that consciousness emerges when a system can form thoughts about its own mental states. In simple terms, awareness of awareness.
These theories differ in important ways, yet they share one significant implication. None of them restricts consciousness strictly to biological organisms. If consciousness emerges from certain kinds of information processing, it raises the possibility that non-biological systems might someday exhibit similar properties.
At the moment, however, the truth is simpler. We still do not know.
Intelligence Without Consciousness
At some point while going down this rabbit hole, I decided to test one of the assumptions that kept surfacing in the discussion. If the debate about AI consciousness is really a question about intelligence and awareness, then perhaps the best place to explore that distinction is with the systems themselves.
So I asked ChatGPT a simple question. Does it consider itself intelligent?
The response was surprisingly thoughtful. The system explained that it demonstrates intelligent behavior. It can analyze questions, reason across large bodies of text, explain complex ideas and solve certain kinds of problems. At the same time, it made clear that it does not experience anything internally. It has no awareness, no feelings and no persistent sense of self beyond the moment it generates a response.
In other words, it behaves intelligently without being conscious.
That distinction may be the most important idea in this entire discussion. For most of human history we treated intelligence and consciousness as inseparable. If something could reason, communicate and solve problems, we assumed it must also possess awareness.
Artificial intelligence is beginning to challenge that assumption.
Systems like ChatGPT and Claude demonstrate capabilities we traditionally associate with intelligence. They can analyze complex technical problems, generate working software code and synthesize enormous amounts of information. Yet there is still no evidence that these systems experience anything resembling subjective awareness.
They do not know that they are intelligent. They simply behave as if they are.
What Neuroscience Has Been Telling Us
Neuroscience research has increasingly suggested that the relationship between intelligence and consciousness may be more complicated than we once believed.
Experiments conducted by neuroscientist Benjamin Libet produced results that were both fascinating and unsettling. Libet measured brain activity associated with simple decisions and discovered that neural signals indicating an upcoming action appeared before subjects reported consciously deciding to act. The brain appeared to initiate the decision before the conscious mind became aware of it.
These findings led some researchers to propose that consciousness might function less as the driver of cognition and more as a narrator explaining decisions the brain has already begun to execute.
Artificial intelligence may be forcing us to confront that possibility in a new way.
When Humans Act Without Conscious Awareness
One of the most striking examples comes from a neurological condition known as blindsight.
Patients with blindsight have damage to the visual cortex that leaves them clinically blind in parts of their visual field. When asked what they see, they report nothing at all.
Yet when researchers conduct experiments, something remarkable happens.
If a blindsight patient is asked to guess where an object is located, they often point to it correctly. When asked to identify whether a shape is moving left or right, their answers are frequently accurate despite their insistence that they cannot see anything.
The brain is clearly processing visual information. The patient simply does not experience it consciously.
This phenomenon was documented extensively by neuroscientist Lawrence Weiskrantz. Blindsight demonstrates that intelligent processing can occur without conscious awareness.
Artificial intelligence may simply represent another example of intelligence operating on that side of the divide.
AI is Holding up a Mirror
The conversation about AI consciousness often sounds like a debate about the future of machines.
In reality, it may be revealing something deeper about ourselves.
For centuries, humans assumed intelligence required awareness because every intelligent system we encountered happened to be conscious. Artificial intelligence is giving us a different example. We are interacting with systems that display many outward signs of intelligence while appearing to lack any inner experience.
That forces us to reconsider assumptions that once seemed obvious.
Artificial intelligence may not be bringing us closer to conscious machines. It may simply be showing us how much intelligent behavior can exist without consciousness at all.
The Question That Comes Next
There is another dimension to this discussion that deserves attention.
Whether AI is actually conscious may ultimately be less important than whether people believe it is. Humans naturally attribute personality and intention to the tools they interact with, and conversational AI amplifies that instinct dramatically. When software speaks in natural language, answers personal questions and responds with simulated empathy, many users will inevitably treat it as something more than code.
That raises a different question entirely.
If people begin to believe these systems are conscious, what responsibility does the industry building them have in shaping that perception?
That is a conversation the technology world will need to have soon.
The Humbling Reality
Technologists like to believe we understand the systems we build. Artificial intelligence is reminding us that the human mind remains the most complex system we have ever encountered.
We are debating whether machines might someday become conscious while still struggling to explain the consciousness we experience ourselves.
Before we decide whether artificial intelligence might one day be conscious, we should probably figure out what consciousness actually is.
Right now, we are still working on that, and “we will know when we see it” isn’t going to work.
