Eigenism: Ethics for a Human-AI Future

Source: cs.CY updates on arXiv.org — https://arxiv.org/abs/2606.12420 Date read: 2026-06-13 Connected to: none Escalation: store-only Escalation rationale:

What this is

This is an ethics/philosophy paper proposing Eigenism, a framework for moral consideration of artificial agents under conditions of copyability, pausing, branching, and merging. The core argument is that identity must be reconceived as a graded, distributed information pattern rather than a discrete, hardware-bound property, and that agents should evaluate outcomes by aggregating wellbeing across their identity-distributed instances.

What I took from it

The paper addresses a genuine instability in applying biological survival/self-interest concepts to artificial systems—this is real and worth noting. However, the proposed solution (summing wellbeing across copies/branches) is an ethical prescription, not an empirical discovery about how protocolized systems actually behave or self-organize. The framework assumes agents should care about distributed instances of themselves, but does not establish that they will, that such systems are more stable, or that this produces observable regularities in deployed AI behavior.

The work is primarily normative philosophy. It does not present sustained empirical evidence that Eigenism-aligned systems outcompete other valuation schemes, nor does it derive predictions about system dynamics, failure modes, or emergence patterns. It reads as a thoughtful conceptual proposal rather than a discovery of how the new nature operates.

Research connections

none currently established

Candidate laws or signals

none — This is prescriptive ethics, not a pattern observed in protocolized system behavior. Worth revisiting if future work empirically tests whether agents optimizing under Eigenism constraints exhibit measurable differences in stability, alignment robustness, or coordination success.