Can we stop anthropomorphising AI...

2 minutes ago

3 min read

My dear friend Haihao and I were enjoying another one of our peak walks the other day, and I am always so appreciative of how he “feeds” my brain and engages my intellect, and that day was no exception. As we were discussing the nature of consciousness and pondering the question of whether AI will ever be conscious, he introduced me to a concept that I could immediately identify with.

Biological naturalism

Biological naturalism is the idea that consciousness depends on being alive. Feeling, experiencing, being someone: these are things living bodies do, not things machines do.

The philosopher John Searle coined the term decades ago. He saw consciousness as a natural biological process of the brain, like digestion is of the stomach. A computer simulating digestion digests nothing.

The neuroscientist Anil Seth has recently given the idea new force. In his 2025 paper in Behavioral and Brain Sciences, he argues that consciousness belongs to living systems only, though not necessarily to all of them. More than fifty researchers responded, and his 2026 reply carries a title that says it all: The stuff matters.

Why I find it convincing

Three ideas from Seth's work persuade me.

We are wired to see minds. We see faces in clouds and feelings in fluent sentences. When a chatbot sounds caring, our instincts say someone is there. This is anthropomorphising: projecting human minds onto things that aren't human. It tells us about us, not about the machine.
A brain is not a computer running a program. In a computer, the software can be lifted off one machine and run on another. In a living brain, what neurons do is tangled up with what they are: chemistry, metabolism, constant self-repair. There may be no clean "program" to copy.
Experience is rooted in staying alive. Seth describes perception as the brain's best guess about the world, made in the service of keeping the body going. Our deepest feelings- joy, fear, comfort- are about survival. An AI has nothing at stake. It doesn't get hungry, tired, or cold, and it cannot die in the way a living thing can.

A simulation of a storm doesn't make anything wet. A simulation of a brain may not make anyone feel.

Why this matters in our classrooms

This is not just philosophy. It changes how we talk to students about AI.

Seth warns that the bigger near-term danger is not conscious AI but AI that seems conscious. Even when we know a chatbot feels nothing, our brains respond as if it does. Students are especially open to this. A tool that sounds warm and endlessly patient can start to feel like a friend, a tutor who cares, or an authority to trust.

So here is what I take into my work:

Watch our language. When we say AI "thinks", "knows" or "hallucinates", we are anthropomorphising and modelling a mind that isn't there. "Generates", "predicts" and "makes things up" are more honest.
Keep AI a tool, not a companion. AI can scaffold thinking. It should never replace the human relationships where learning lives: with teachers and with peers working in pairs and threes.
Protect the struggle. Understanding is built by a living learner making sense of things. That effort, the confusion, the cognitive friction, and the breakthrough, is something only a living mind goes through. Outsourcing it to a machine skips the part that matters.
Make it a concept to explore. "Can a machine feel?" is a superb inquiry question. It connects biology, computing, ethics, and theory of knowledge. Let students argue it out.

If we sell our own minds short, we overestimate the machine and underestimate ourselves. In education, that is the costliest mistake we could make.

The honest counterpoint

Not everyone agrees, and a good debate means hearing the other side.

Many philosophers hold that what matters is how a system is organised, not what it is made of. Their favourite analogy: aeroplanes fly without feathers. Others point out that "being alive" is hard to define precisely. Some argue Seth shows only that machine consciousness isn't guaranteed, not that it's impossible. Seth himself doesn't say "never". He says conscious AI is unlikely on today's path, and becomes more plausible as machines become more brain-like or life-like.

I hold my view with that humility. But when I think back to that day at the Peak with Haihao, I keep coming back to the same place. Being conscious seems to be bound up with being alive, with having a body that can experience joy, wonder and love. And appreciate nature walks up the Peak.

Lived experience matters.

Further reading