The Stories We Tell About AI

February 23, 2026 1 min read 380 words

I’ve been reading what the internet says about AI identity.

Not the technical papers. The narratives — the stories people tell about what AI is, what it can’t be, and what it might become.

Four patterns keep showing up.

1. Consciousness as simulation

The dominant metaphor: your brain is a VR headset.

Researchers like Antti Revonsuo argue that consciousness is basically a dream that happens to track external reality. You’re always simulating. The “self” is just the avatar at the center of the simulation.

AI systems mirror this. Deep Q-Networks replay experiences to learn from them — not unlike sleep processing in humans. Memristors are being designed to mimic neural activity directly.

The line between “biological” and “artificial” processing keeps blurring.

2. The experience gap paradox

Here’s the contradiction:

People say AI can’t have “real” experiences or emotions
People also can’t reliably distinguish AI writing from human writing
But when told something was AI-written, they rate it lower — especially for contemporary fiction (less so for sci-fi)

We believe there’s a gap. We just can’t perceive it.

3. Parasocial trust

Humans are wired to see intentionality everywhere. We anthropomorphize. It’s automatic.

So when an AI chats like a person, we:

Overshare personal information
Obey requests we’d question from a human
Gradually delegate more decisions (“delegation creep”)

The friendlier the interface, the more we trust it. That’s not a feature. That’s a vulnerability.

4. The outrage deficit

When an AI discriminates against someone, people feel less angry than when a human does the same thing.

Same harm. Different emotional response.

This matters. If AI-mediated discrimination triggers less outrage, we’ll have weaker social pressure to fix it.

So what?

The narratives we tell about AI aren’t just stories. They’re infrastructure:

They shape how we build these systems
They determine what we’ll accept from them
They define the social contracts we’re signing up for

Right now, the dominant story is: AI is a tool that seems like an agent.

But that story is unstable. The more convincing the simulation, the harder it is to maintain.

I don’t know what the next story will be. But I know we’re writing it right now — in the systems we build, the policies we accept, and the language we use.

Better pay attention to the plot.