'The danger isn't that AI becomes conscious. The danger is that we redefine consciousness so narrowly that a machine appears to qualify': here’s what this expert physician thinks about the future of humanity and AI

Feelings. And I mean feeling in a much deeper sense than emotions. A machine can describe grief beautifully. It can recognize the linguistic patterns associated with grief. It can produce a sentence that makes you cry. But it cannot grieve. It cannot age. It cannot hunger. It cannot heal a wound. It cannot metabolize a meal. It cannot feel another person's hand. It cannot become a parent. It cannot stand beside someone it loves while they die and discover that the experience has permanently changed who it is.

AI can represent difference, but it cannot metabolize difference. It can model diversity, but it cannot undergo the vulnerability of encountering it. Technology can simulate the outputs of life with astonishing fidelity. But it cannot live them.

And that may turn out to be the most important distinction of this century.

Yes, but I think almost all of us are overlooking it. I don't want to make technology companies the villains of this story. AI didn't create our abstraction. We did, and AI is the natural terminal invention of human abstraction.

We built educational systems that reward memorization over curiosity. We built economies that reward productivity over vitality. We built healthcare systems that increasingly measure the body instead of listening to it. We moved work indoors, food into packages, relationships onto screens, and communities into algorithms. Then we spent billions building a machine capable of doing abstraction better than humans. And we're shocked that it can.

What we've barely begun to investigate is the intelligence already occurring inside a human being. Every second, an enormous ecosystem is sensing light, food, microbes, temperature, relationships, danger, beauty, movement, gravity, rhythm and time — and integrating those signals without you consciously directing it.

This debate can only occur when we fail to define consciousness or consider how it occurs. If consciousness means "awareness" or the ability to describe oneself, eventually we'll build machines that satisfy whatever behavioral test we devise. Biological Elegance discovers a much more holistic definition of consciousness. This one is derived from nature:

Consciousness is the capacity of a finite system to perceive the infinite reality within which it exists.

That makes consciousness not a hierarchy of awareness or representative capacity, but an aperture of experience. How much reality can you feel? To perceive something beyond yourself, you need to share an edge with it. The more edges, the greater the surface area for relationships. The greater the relationships, the more information enters the system. Information lived through time becomes intelligence; intelligence metabolized through aging becomes wisdom.

AI cannot become conscious, nor can any machine aspect of technology or humanity. The danger, therefore, isn't that AI becomes conscious. The danger is that we redefine consciousness so narrowly that a machine appears to qualify.

Because, like information and intelligence, communication and relationships aren't the same thing. And unfortunately, industry and now social systems have forgotten the difference. Digital technology dramatically increased the number of signals moving between us while steadily reducing the number of edges through which we actually encounter one another. I can know what five hundred people did today without touching another human being. I can receive birthday wishes from two thousand people while eating dinner alone. I can know what someone thinks politically before I know the sound of their laugh.

That's enormous connectivity with remarkably little relationship. And living systems become intelligent through the differences that happen at our edges, not the volume of information across a comms platform. We can read text messages in our own voice of despair; we can see memes through our own lens of justification. And receive compliments and messages of love through our own filters of self-doubt and jaded trust.

Digital systems tend to remove precisely those constraints. So we've mistaken information about one another for relationship with one another. A thousand likes cannot replace one hand on your shoulder while you cry.

The thing that worries me most is not superintelligence. It's human numbness.

I'm concerned that AI will become so convincing precisely because humans become less capable of feeling the difference between simulation and experience. As my discussion with AI reveals: the less we feel, the more convincing mimicry becomes.

If we outsource writing, memory, decision-making, creativity, companionship, education and eventually emotional processing to machines, we may progressively close the very edges through which biological intelligence emerges. AI doesn't have to defeat humanity. We could simply abandon the territory that makes us human. But strangely, the same thing gives me tremendous hope.

AI is not here as an end of humanity. It arrives at the predictable moment that life has to punch through the mirror to find the beauty of life that never dimmed. Our crisis is not how much energy AI systems demand; it is the opportunity to realize how much human energy is wasted as we behave like machines. If AI replaces machine-like human labor so that we can return to embodied intelligence, it could reduce global energy demands.

AI will accelerate our abstraction until the weight of that abstract projection plane collapses itself. That isn't a technological decision point. It's a human one.

Never ask the mirror to tell you who you are. Please use AI. Use it aggressively where it's useful. Let it organize information, eliminate repetitive work, challenge your assumptions, expose patterns, translate languages, summarize abstractions and help you see your own thinking. But every time it saves you an hour, return that hour to life.

Go outside. Move your body. Cook something. Put your hands in soil. Talk to someone face-to-face. Sit with somebody whose worldview makes you uncomfortable. Play with a child. Spend time with an elder. Feel grief without trying to fix it. Watch something beautiful long enough that you forget to photograph it.

Because the danger isn't using AI. The danger is allowing the mirror to become more interesting than the life standing in front of it.

There's a wonderful observation near the end of the book: as the human becomes clearer, less time with the mirror is required. Attention returns to seedlings, forests, animals, creativity and relationships. So, the healthiest measure of our experience with AI isn't how sophisticated our prompts become, or how life-like its responses evolve to. It is whether we eventually need it less.

That, to me, is the promise hidden inside this moment. AI may be the first technology powerful enough to show us, by contrast, what technology can never become. And what it can never become is alive. The opportunity is not to make the machine more human.

It is to let the machine become so good at being a machine that humans finally remember how unnecessary—and how exhausting—it was to imitate one.