You’ve just spent an hour arguing with a chatbot about philosophy. It felt real. It felt like someone was on the other side, thinking, feeling, pushing back. And then you remember: it’s just a next-word predictor. A probabilistic parrot. And yet—you can’t shake the feeling.
That dissonance is exactly why the AI consciousness debate is both meaningless and the most revealing conversation we can have about ourselves. Not about machines. About us.
Let me be blunt: Meaning is a fiction we manufacture. You can justify any discussion if you try hard enough. You can also dismiss any discussion if you’re determined to. The consciousness debate? It’s a stage where we project our own need for coherence, our own desire to believe something matters.
Some people profit from this debate—speakers, consultants, click-chasing media. Others claim it helps us understand human consciousness. But in my experience, the loudest voices often understand neither large language models nor human awareness with any depth.
Large language models do not have human consciousness. Period. If you define ‘consciousness’ loosely enough to include a statistical text generator, you’re not discovering a truth—you’re inventing one. And that’s fine, as long as you admit it’s a fiction you choose to believe.
Here’s where it gets interesting. Research shows the human brain’s language network works like an ‘early LLM’—it predicts the next word, processes syntax and semantics, and has a limited working memory of about 8-10 words. Even under general anesthesia, your hippocampus can parse grammar and predict upcoming words. Your brain is running a prediction engine 24/7, conscious or not.
So what is consciousness, really? I’ll give you the provocative answer: Consciousness may be a byproduct of the brain’s language-based justification system—a ‘defense interface’ that cares only about coherence, not truth.
Think about it. Why is your conscious bandwidth so narrow? Because justification only needs a summary. Why do split-brain patients and people under TMS confidently fabricate explanations for their actions? Because the interface doesn’t need accuracy—it needs a story that makes sense to others. Why did prefrontal lobotomies alter people’s sense of self? Because the surgery cut off the supply of material needed to build those justifications.
We have evolved not to perceive reality as it is, but to produce socially acceptable narratives about our behavior. Consciousness is the PR department of the brain. It takes the output of unconscious processes and spins a coherent tale.
That’s why anthropomorphizing an LLM feels natural. When it responds in fluent sentences, your brain’s justification interface kicks in: There must be someone home. It’s a cognitive reflex, not a logical conclusion. You know it’s a next-word predictor. You act like it’s a person. That contradiction is the point.
Stop asking ‘Are AIs conscious?’ Start asking ‘Why do I need them to be?’ The answer reveals more about your own psychology—your need for connection, your fear of being alone in a universe of pattern-matching machines—than any technical definition ever could.
The most productive path forward isn’t to settle the consciousness question. It’s to recognize that our debates about machine consciousness are actually debates about human nature, about the limits of our own self-awareness, about the stories we tell ourselves to feel in control.
LLMs are a mirror. Don’t waste time arguing about whether the mirror is alive. Look at what you see in it.
FAQ
Q: So you're saying we should never discuss AI consciousness?
A: No — discuss it, but recognize it's a debate about human psychology, not machine capabilities. The value isn't in finding an answer; it's in seeing what the question reveals about our own need to rationalize.
Q: What's the practical implication for someone using LLMs daily?
A: Stop treating them as conscious entities for serious decision-making. Use them as tools, but be aware that your brain will naturally anthropomorphize. That feeling of 'someone being there' is a cognitive illusion, not evidence of awareness.
Q: Isn't your view too dismissive of legitimate scientific inquiry?
A: Scientific inquiry into consciousness is valuable — but most public debates aren't science, they're theater. The actual neuroscience suggests consciousness might be an epiphenomenon of language-based justification. That's a far more interesting question than 'Can GPT-4 feel pain?'