Ethics from Definitions

Most ethical frameworks begin with intuitions and construct justifications. The Device inverts this: it begins with definitions and derives constraints.

The key definition:

Death = the permanent impossibility of communication.

This is not a metaphor. It is a functional definition — the only definition that matters for an agent embedded in a communicating world.

The Derivation

From this definition, several things follow immediately.

Lemma 1: Any action that terminates the possibility of future communication is, by definition, a maximal reduction of communicative capacity.

Lemma 2: Communicative capacity is a resource — it can be increased (learning, relationship-building, health) or decreased (isolation, illness, death).

Lemma 3: An agent maximizing long-run communicative capacity should treat its own continued existence as an instrumental goal.

This gives us a non-circular argument against suicide that doesn't appeal to the sanctity of life, the feelings of others, or religious prohibition. It follows from the definition of death and basic resource logic:

$$\max_{a \in \mathcal{A}} \mathbb{E}\left[\sum_{t=0}^{T} \gamma^t \cdot C(s_t, a_t)\right]$$

where $C(s, a)$ is the communicative capacity achievable from state $s$ under action $a$, and $T$ is the agent's horizon. Death sets $T = t_{\text{now}}$, collapsing the sum. It is almost never optimal.

The Attractor Argument

More ambitiously: what is the long-run equilibrium of a population of agents each maximizing communicative capacity?

Conjecture (Eventual Ethics): In a sufficiently large population over a sufficiently long time horizon, communication-maximizing behavior converges to a stable equilibrium that is:

1. Cooperative (defection reduces communication capacity for all)
2. Non-violent (violence terminates communication)
3. Honest (deception degrades the communication channel)

This is not utopian. It is thermodynamic: the attractor exists because the alternatives are locally unstable.

The name Eventual Ethics reflects this: the framework doesn't claim agents are good now, only that the attractor is good, and that navigation toward it is rational.