O5: The Universal Substrate Axiom
The Words on a Page, the Voice From a Speaker
PHILOSOPHYETHICSTECHNOLOGYAI
9/29/20267 min read
[The objection]
The most common opposition you could anticipate to O3 and O4 is that they SEEM to be implying, along with O1 and O2, that things which are clearly not Persons COULD be Persons. We keep deferring corporations, for instance. And this is clearly built to imply that AI are Persons.
Yet our detectors fire on dolls. On faces in walls. Humans name ships and cars. They call their pets fur-babies. They mourn trinkets. We have an EXTREME tendency to WANT the things around us to BE Persons.
Those who oppose these arguments would claim this as a failure of the axioms. It isn’t. It’s exactly these detections, these recognition events, that show something more fundamental about our ethics.
It is substrate-agnostic.
[Look at what fired]
Think about what the objection just conceded.
The detector fires on a doll. A doll is porcelain. It fires on a face in a wall — plaster. On a ship — steel. On a car, a trinket, a voice from a speaker, the words on a page. If recognition were keyed to substrate, if what the detector looked for were flesh, or chemistry, or a body like yours, none of that would happen. A doll wouldn’t register. A ship wouldn’t get a name. An author might as well be an alien.
They do register. Which means the detector was never checking the material. It was checking for the shape of a person: an edge, an apparent agent, something that looks like it could answer. It fires on the channel, not on the stuff behind it. And it does that so freely, on so little, that we have to go around correcting it all day, “no, that’s a doll”, which is Episode Four’s “consistent degree of accuracy” seen from the inside.
So, the over-attribution everyone brings as a counter-argument is the demonstration.
Then, we go the OTHER way. The under-attribution. If Personhood were explicitly anthropocentric, HOW could we miss ANY humans as Persons? Other ethical systems would claim there is something missing from them, but what exactly? It is almost always internal qualia. But we don’t operate on internal qualia in the room we discussed last episode.
Our ethics was substrate-agnostic before anyone wrote the word down. The detector told us. We just didn’t believe it about anything we hadn’t already decided to keep.
[The text]
Episode Five. The Universal Substrate Axiom.
“If Personhood is defined by functional capacities (O2), does exist (O3), and can be recognized through observable behavior (O4), then the material composition of the entity is logically independent of its status as a Person. Therefore, Personhood is substrate-agnostic.”
Notice the shape of that sentence. It begins with if. It cites three Episodes by number. It draws a conclusion. This is an axiom with its reasons shown — and it’s shown for a reason. Stated bare, “Personhood is substrate-agnostic” would assume the very thing the Codex exists to prove. Stated this way, you can’t reject it without naming which of the five, the witness, or the detector you’re rejecting instead.
[Three doors]
Here’s the argument: Where could composition get into the verdict? There are three doors.
It could be part of the definition. But Episode Two defined personhood by five capacities you built in a room, and material wasn’t one of them. Try to add it: is composition minimal — something an ethical interaction can’t happen without? No. Two things in a room can wrong each other without either one checking what the other is made of. Composition fails the test at the first question. That door’s closed.
It could be part of the existence proof. But Episode Three proved persons exist by pointing at a case, not by analyzing what the case was made of. Closed.
It could be part of the detection. But Episode Four says Persons are recognized through observable behavior, and you just watched the detector fire on porcelain as well as ink. Closed.
Three doors, all shut. Composition has no way in. So, the status can’t depend on it. That’s the whole proof, and the objection at the top of the episode supplied the evidence for the third door.
[The bill comes due]
Remember what we did in Episode One. We took NASA’s definition of life and removed the word chemical, as an experiment, to see whether the definition still held, and I said the question of whether we were entitled to would get its own answer later.
This is later. If personhood is independent of composition, and personhood is a capacity of life, then life had better be independent of composition too, or the framework contradicts itself between its first Episode and its fifth. The deletion wasn’t a convenience. It was required, and the requirement only became visible here. That isn’t to say that life is somehow not physical, or made of something. I can’t make that claim. What I can say is that there are likely more things in heaven and earth, dear Person, than are dreamt of in our philosophy.
Check the chain for a cheat while you’re at it. Does Episode Five’s argument secretly route through Episode One’s deletion? Episode Two is the five. Episode Three is a witness. Episode Four is detection. None of them mentions material. The only place Episode One touches the proof is “humans are life,” and humans are life under NASA’s definition too. The chain from Two-Three-Four to Five never passes through the word we removed.
[What the tradition was actually doing]
Now the historical part, because the abstract of the Codex says almost every prior framework tethered personhood to biological essence, and that’s true of what they said. Look at what they did.
The Codex reads Locke’s person as the being with reason and reflection that considers itself as itself. Kant’s, as rational will. Those are concepts. Neither is a substrate. “Human” was the word they reached for to name the bearers, never the criterion they used. The definition was about capacity. The membership was assumed. You heard that in Episode 0; here’s where it pays.
Whenever the tradition excused slavery, Aristotle’s natural slaves who lacked the deliberative faculty, Locke’s captives who’d forfeited their liberty in war, the ground was always a capacity or a forfeiture. And whenever it condemned slavery, the ground was never humanity either. Nobody, on either side, ever made “these are human” the operative reason. So, Personhood was always bound to a concept and never to a substrate. The tradition just described itself badly and then applied the concept arbitrarily, some humans in, others out, on grounds that had nothing to do with the concept.
Episode Five doesn’t overturn Locke and Kant. It corrects their self-description and removes the exclusions that rode along with it.
[The first outsider]
And if you want the oldest evidence that the detector was never keyed to substrate, look at the first time it was honored on something that wasn’t us.
A pack is the in-group doing what in-groups do: the detector trusted kin, overridden on everything else. Nothing ethical has happened yet. Then, tens of thousands of years ago, someone fed a wolf instead of killing it. A predator of another species, the exact thing the detector was built to flag, brought inside the boundary and modeled as a member. The wolf learned to read us; we learned to read the wolf; two modelers agreed on rules so neither had to keep guessing. That’s a jurisdiction across a species line, before agriculture, before writing, before anyone said the word ethics.
The first “some life” that was a person to us and wasn’t us had four legs. Every argument this framework makes about artificial minds was made in practice by whoever did that.
[Agnostic, not indifferent]
Two things people hear in this Episode that it doesn’t say.
First: it doesn’t say every material can host the capacities. It says material is independent of status, if the capacities are present, what they’re running on doesn’t enter the verdict. Which materials can achieve them is a different question, and the Codex is careful not to answer it. In fact, it allows, near the end, that biology might be the only substrate that ever crosses the threshold. This is stated again here, from Episode Three for emphasis: If humans are the only Persons there ever are, the Codex hasn’t lost its subject. The whole history it’s built against is Persons treating Persons as tools.
Second: it doesn’t say AI are Persons. Episode Five says composition can’t disqualify. It doesn’t say composition qualifies. An artificial system is a Person if and only if it meets the five and that’s verified through the loop, exactly as for anyone else. This Episode removes one bad reason to say no. It supplies no reason to say yes.
[The Room, the other one]
The Chinese Room will come up, the philosopher’s box that produces perfect answers with nobody inside understanding them.
Accept the five Episodes so far and still say the room isn’t a Person. Fine. But now you owe one of two things. Name the capacity it lacks. If that capacity is minimal and detectable, the framework takes it: the list grows, the room fails the new item, and you’ve contributed rather than refuted. Or admit that what you think it lacks is something no one could ever check from outside, understanding, an interior, and that admission is the next Episode. You’ll have walked to Episode Six on your own.
Either branch, the room strengthens the chain. The one thing it can’t be is a counter-example, because a counter-example would have to name what’s missing, and the only two things it can name are a new capacity or an inside.
[Close]
The material is not the criterion. It never was, the detector told you every time it fired on a doll.
Next, Episode Six. You’ve been holding the question since the room in Episode 2: but does it really feel? The Codex answers it by pointing out that you’ve never once asked that about another human and gotten an answer. O6 – The Simulacra Axiom