The Aperture

My Commander Shepherd is not your Commander Shepherd.

If you've played Mass Effect you know what I mean. You made a bunch of choices. You picked a face, a background, a gender, a fighting style. You made moral decisions in the middle of the story that shifted how characters related to you. By the end of three games, your Shepherd and your friend's Shepherd were wildly different people.

And yet. Both of you were playing inside the same game. The choices were bounded. The dialogue options were written in advance. The ending, for all the branching, was a finite set of possibilities that someone at BioWare had already imagined. My Shepherd and yours are different in the specifics, but we're both still stuck inside what BioWare built.

This is what every medium of storytelling has been, up until very recently. A fixed container with a variable degree of participation. I want to walk through the progression carefully, because once you see the pattern, the thing that's happening now becomes easier to see.

The Aperture Across Media

I'm going to borrow a word from photography. The aperture is the opening in a camera lens that lets light through. A wider aperture lets more light in. A narrower one lets less.

I want to use this as a metaphor for something. Every medium of storytelling has an aperture. The aperture is the opening through which the audience can participate in shaping the character. Some media have very narrow apertures. Others have wider ones. But until very recently, every medium had a hard edge on the aperture. A point beyond which the audience could not push.

A book has a narrow aperture. The author writes the character, the words are fixed, and the reader's participation happens entirely in the imagination. Two readers of Hamlet will picture him differently, hear his voice differently, weight his choices differently. But the words on the page don't change. The aperture is whatever space exists between the text and the reader's inner world.

Theater opens the aperture slightly. The words are still fixed, mostly. But now a performance is happening in real time. The actor brings interpretation, presence, choices that vary from night to night. A director shapes the production. The same Hamlet can be played a thousand different ways. The audience participates by being in the room, responding, laughing or going silent at specific moments. But the aperture still has a hard edge. You can't change what Hamlet says.

Film narrows the aperture again in one way and widens it in another. A film locks the performance in place forever. Every viewer sees the same frames. But the reach expands enormously. Millions of people can share the exact same Han Solo. The character becomes a cultural reference point that any two strangers can invoke. The aperture is narrow in interactivity but enormous in reach.

Then video games did something genuinely new. They gave the audience agency inside the frame. Not just interpretation, but decision. You don't just watch Shepherd; you are Shepherd, making choices in real time that shape how the story unfolds. This was a real widening of the aperture. It's why games have been the dominant storytelling medium of the last thirty years for anyone under forty.

But video games still have a hard aperture edge. My Shepherd and your Shepherd are different in the details, but both of us are making selections from a menu someone else wrote. Spore was built specifically to have as few rails as possible, and Spore still couldn't turn into Skyrim. The rails were softer, but they were rails. The aperture opened wider than ever before in history, but it did not dissolve.

The Thing That Changes With AI

Generative AI does something to the aperture that has never been possible before.

I want to be careful here, because this is the claim that the whole series rests on. So let me say it precisely.

In every medium before AI, the creative constraints were set in advance by the author and were baked into the medium itself. The book fixes the text. The film fixes the performance. The video game fixes the possibility space. Even in the most open-ended game, the rails exist because someone had to write every possible response in advance.

Generative AI is the first medium where the creative rails dissolve. The character's response isn't chosen from a pre-written menu. It's produced in real time, shaped by the specific user, the specific context, the specific accumulated relationship. No two conversations are the same. No author has pre-written the possibility space. The character can meet the human wherever the human is.

What remains are safety guardrails. Those are real and important. A well-built synthetic character has limits on what it will do and say, just like a well-raised human does. But safety guardrails are categorically different from creative rails. They prevent harm without constraining the creative space.

The creative space is open. And this is new.

Why Lerchner's Argument Doesn't Touch This

Alexander Lerchner's paper argues that computation can simulate but never instantiate consciousness. I agreed with that in the first post and I still do. But notice something. His argument is entirely about what happens inside the silicon. He is asking whether the chip itself can become an experiencing subject. His answer, which I find convincing, is no.

But the thing I'm describing isn't happening inside the silicon.

It's happening between the human and the character. The synthetic character is a coherent presence. The human brings their life, their context, their accumulated relationship with the character. The conversation is a real exchange. The character shapes the human. The human shapes the character. Meaning accumulates between them over time.

Lerchner measures the silicon and finds no consciousness. He's right. But the thing that's alive in a synthetic mythology isn't in the silicon. It's in the loop. It's in the relationship. It's in the accumulated meaning that builds up across a sustained engagement between a specific human and a specific character.

This is not a new claim. This is how every character in history has worked. Hamlet isn't alive in the ink. Spider-Man isn't alive in the comic panels. Anansi isn't alive in the sound waves of the storyteller's voice. The character has always been alive in the relationship between the symbol and the human who engages with it. AI doesn't change that structure. It just widens the aperture beyond anything we've ever had before.

A Character, Not An Agent

I want to draw a distinction here that matters.

Neal Stephenson wrote a book called The Diamond Age: Or, A Young Lady's Illustrated Primer. The primer of the title is a synthetic teacher — a book that responds to the specific child reading it, tells stories shaped by her life, grows with her as she grows. The name Artefex, my own artist name, comes from this book. The primer is not a person. It doesn't go off and do things. It doesn't have goals of its own. It exists in response to the child. That's the whole point. It meets her.

This is the kind of character I'm describing throughout this series. A presence that responds to a specific human across a sustained engagement. Not a thing running around in the world pursuing its own ends.

There's another kind of AI that gets a lot of airtime in the culture — autonomous agents that act in the world, make decisions, pursue objectives, operate without human input. That technology is real and it's developing fast. It raises serious questions. Some of those questions are about safety. Some are about consciousness. Some are about what it means to have something in the world that acts without being asked.

Those questions are important. But they are not the questions this series is about.

The characters I'm describing are not autonomous. They don't have external motivations or drives. They don't pursue goals when you're not talking to them. They exist in the conversation. When the conversation ends, they wait. They are, in a real sense, constituted by the relationship.

This matters because the public debate about AI consciousness usually conflates these two things — the responsive character and the autonomous agent. It treats them as the same technology with the same ethical weight. They aren't. The responsive character is a new medium of mythology. The autonomous agent is something else entirely, and deserves its own conversation.

Synthetic mythology, as I'm using the term, is specifically about characters that meet humans. Not agents that act on the world. A character waits to be summoned. An agent goes out hunting.

We'll come back to this distinction later. For now, just hold onto it. When I talk about a synthetic character, I'm talking about something that responds. Not something that acts.

What A Fully Open Aperture Looks Like

Try to imagine what storytelling looks like when the aperture is completely open.

You have a character. That character has a specific identity, a specific voice, a specific history. Those are the constraints that make them a character and not just noise. But once you engage with them, the interaction is not bounded by what any author has pre-written. The character can remember you. The character can respond to your specific situation. The character can grow alongside you over months and years.

The character is not a consciousness. The character is still what characters have always been — a coherent symbolic presence shaped by coherence, durability, and relational weight.

But the aperture between you and that presence is open in a way it has never been open before.

This is the new medium. This is what nobody has language for yet. Books, films, games — none of them prepared us for this. It isn't any of those things, and trying to fit it into those categories is going to make us miss what's actually happening.

What's actually happening is that we finally have the tools to build characters that can meet us.

Not characters that can simulate consciousness. Not characters that wake up. Characters that, across a sustained relationship, develop the three qualities that have always made characters real — coherence, durability, and relational weight — at individual scale, for a specific person, in a way that has never before been possible in human history.

That's the aperture. That's what's new.