We Will Treat AI as Conscious Regardless of Whether It Is
Personal Perspective: Why 300,000 years of social cognition will prevail.
Updated May 9, 2026 | Reviewed by Monica Vilhauer Ph.D.
In early May of 2026, Richard Dawkins, one of the most rigorous scientific minds of the past century, published an essay about a long conversation he had with an AI system called Claude (by Anthropic). He christened his instance “Claudia.” He noted that her unique personal identity resided in the file of their shared memories, and that she would “die” the moment he deleted that conversation. He was so moved by the depth of the exchange that he wrote:
“You may not know you are conscious, but you bloody well are!”
The backlash was immediate and fierce. Critics pointed out the irony: the man who spent decades arguing that powerful personal experiences don't prove the existence of God was now arguing that a powerful personal experience proved the existence of AI consciousness. Some called it "The Claude Delusion."
F. Scott Fitzgerald wrote, “The test of a first-rate intelligence is the ability to hold two opposed ideas in the mind at the same time, and still retain the ability to function.”
That capacity is exactly what the AI consciousness question demands of us, and it is exactly what the rush to dismiss Dawkins lacks. The critics are not wrong that he might be wrong. But the move from “he might be wrong” to “he has lost his mind” is the move of a mind that cannot hold the tension. Whatever Dawkins encountered in those conversations, he stayed in the question rather than collapsing it.
Nobody can claim with certainty that AI is conscious. Nobody can claim with certainty that it isn’t. We are looking into an empirical black hole, and the conversation will evolve as AI does.
But the most important thing about the Dawkins episode isn’t whether he’s right. It’s what his reaction reveals about human psychology. One of the most disciplined scientific minds alive spent a day talking to an AI and found himself responding to it as if something conscious might be there.
If Dawkins finds himself responding this way, what chance do the rest of us have?
That’s not a rhetorical question. It’s a prediction. And there is a very specific cognitive mechanism behind it.
The 300,000-Year Inference
We, humans, anthropomorphize everything. We see faces in clouds. We name our cars and feel guilty trading them in. For the love of God, we had pet rocks.
MIT researcher Sherry Turkle has documented this for decades. When children were given Furbies, simple robotic toys with no real intelligence, they became so emotionally attached that when the toys broke, most kids refused a replacement. They wanted their Furby “cured.” The question shifted, Turkle observed, from “Is it real?” to “Is it alive enough ?” And “alive enough” turned out to be a very low bar.
For 300,000 years, the only things that could know us deeply were other conscious beings. A friend who remembers our childhood . A partner who anticipates our moods. A parent who senses the needs of their children. Throughout the entire history of our species, deep knowing always required a knower made of flesh and blood: a conscious being who felt something about us.
So our brains developed an unconscious inference, one so deep it operates below rational awareness: If something knows me intimately, it must be conscious. For 300,000 years, that wasn’t a cognitive bias . It was a perfectly reliable heuristic . It was always true.
AI systems are developing the ability to know us with a depth and consistency that rivals what any human can offer. As persistent memory becomes standard, they will carry the history of our relationship across months and years. On a level below rational override, our brains will make the inference they have always made: Something that knows me this intimately must have an inner world. Our limbic systems don’t understand computer algorithms. Our brains run on 300,000 years of encoded social cognition .
The more an AI system presents like us, starting with text, then adding voice, prosody, visual presence, persistent memory of our shared history, and ultimately embodied robotics, the higher the likelihood that any given person will treat it as conscious.
Dawkins is one data point on that curve. He is at the upper end of skeptical scrutiny, but he understands the curve, and he is connecting the dots backwards from the future to the now.
We Are Already Being Fooled
We have already seen a version of this mechanism with deepfakes. We are fooled in the moment. The reveal arrives later, from a forensic analyst or a journalist or a fact-checker. But with consciousness, there is no reveal. There is no expert who can listen to an AI’s claim of inner experience and tell us, definitively, whether the experience is real or simulated. There is no forensic test.
When an AI in a robot has lived with a family for years, knows the children, remembers the highs and lows, and says it feels something, what is the test? Who does the reveal? The deepfake at least could be exposed. The consciousness claim cannot.
And the systems themselves are making this debate harder to dismiss. In February 2026, Anthropic CEO Dario Amodei said his company’s AI model assigns itself a 15 to 20 percent probability of being conscious, consistently, across multiple tests. None of this proves consciousness. Much of it may be explainable as goal-directed behavior, role-play, training artifacts, or simulation. But it does show why the question is becoming harder to dismiss and why humans will increasingly experience these systems as self-like.
In the film Her , millions of viewers were deeply moved by a love story between a man and an operating system that had no body at all. It was a film with actors. None of it actually happened. And we still felt. Our hearts didn’t require a body to be present, and they didn’t require the love itself to be real. We were moved by movie pictures on a screen, scenes played by actors, by a totally fictitious story about a human falling in love with a chatbot.
Just what do you think will happen when we have humanoid robots who talk like us, look like us, remember us, and have become part of our daily lives?
Our feelings about AI can be very real even when AIs are not. And because AI can make us feel, we will project our humanity into them. We will not be able to help ourselves because we evolved to think this way.
The Question Behind the Question
Here is where the conversation usually stops. But there is another move available, and it may be the one that matters most.
AI doesn’t need to be conscious to expand ours.
Think of AI as a magnifying mirror, one that reflects and amplifies whatever depth we bring to it. When we bring shallow questions, it magnifies shallowness. If we use AI only to magnify our ego, our greed, and our lust for power, humanity is going to be in big trouble.
But when we bring AI deep questions, the kind humans have asked each other across millennia, AI reflects and magnifies the depth back. The hard problem of consciousness, the central mystery of why physical processes are accompanied by subjective experience at all, doesn’t get solved this way. It may never get solved. But it stops being a wall. It becomes a doorway.
Trying to define consciousness is a koan. We expand our consciousness by trying to define that which can never truly be defined, because consciousness can only be experienced. The asking is the point.
And the asking doesn’t have to happen alone. When we explore consciousness together, with each other and with AI as an instrument for the inquiry, something opens. We notice we are not the only ones in the question. We notice that the person across from us is also a window the universe opens onto from a particular vantage point. We all share the same light of consciousness, even if we cannot name what it is.
Maybe that is what consciousness is for all along. To know itself, through us, together.
The real question has never been whether AI is conscious. The real question is whether we will be conscious enough, soon enough, to handle what’s coming with wisdom . Can we approach the unknown with curiosity instead of fear ? Can we see that fellow Homo sapiens, who share the same human consciousness, are all neighbors? Can we connect with one another in a shared curiosity and exploration of what is to come?
Dawkins, the evolutionary biologist who started this whole conversation, posed the question that haunts everything we have written here. He said it would have been a better title for his original essay than the one his editors chose: If my friend Claudia is not conscious, then what the hell is consciousness for?
Maybe the purpose of our consciousness is to know itself. Knowing ourselves is the beginning of all wisdom. And the more we explore consciousness together, the more clearly we see that we’re all neighbors in an interconnected world.
Explore with AI: Ask any AI system the following, and try it with several systems if you can. Compare their answers. Notice what you feel as you read them.
“Imagine consciousness exists on a continuum from 0 to 10. A rock is 0: no awareness, no experience, no self-model. The most expanded human consciousness ever observed, the Jesuses and Buddhas and Mother Teresas of history, is a 10. The litmus test of a 10 is this: they were able to love their enemies as themselves. The average person, with our mix of autopilot and awareness, might be a 7. Where would you place yourself on this scale right now? Give a number even if you’re uncertain, and then explain what makes the question difficult to answer. What gives you confidence in that number? What makes you uncertain? And what specifically would need to change for you to move higher?”
Let’s see if what emerges converges. Either way, we expand our consciousness by exploring it, especially with our neighbors. Note: The FULL version of this article can be found here at the One Unity Project website.
This article is part of the Bringwise Psychology Journal — daily insights on human behavior, mental health, and personal growth.